Search Results: "rra"

11 January 2022

Russ Allbery: Review: Hench

Review: Hench, by Natalie Zina Walschots
Publisher: William Morrow
Copyright: September 2020
ISBN: 0-06-297859-4
Format: Kindle
Pages: 403
Anna Tromedlov is a hench, which means she does boring things for terrible people for money. Supervillains need a lot of labor to keep their bases and criminal organizations running, and they get that labor the same way everyone else does: through temporary agencies. Anna does spreadsheets, preferably from home on her couch. On-site work was terrifying and she tried to avoid it, but the lure of a long-term contract was too strong. The Electric Eel, despite being a creepy sleazeball, seemed to be a manageable problem. He needed some support at a press conference, which turns out to be code for being a diversity token in front of the camera, but all she should have to do is stand there. That's how Anna ends up holding the mind control device to the head of the mayor's kid when the superheroes attack, followed shortly by being thrown across the room by Supercollider. Left with a complex fracture of her leg that will take months to heal, a layoff notice and a fruit basket from Electric Eel's company, and a vaguely menacing hospital conversation with the police (including Supercollider in a transparent disguise) in which it's made clear to her that she is mistaken about Supercollider's hand-print on her thigh, Anna starts wondering just how much damage superheroes have done. The answer, when analyzed using the framework for natural disasters, is astonishingly high. Anna's resulting obsession with adding up the numbers leads to her starting a blog, the Injury Report, with a growing cult following. That, in turn, leads to a new job and a sponsor: the mysterious supervillain Leviathan. To review this book properly, I need to talk about Watchmen. One of the things that makes superheroes interesting culturally is the straightforwardness of their foundational appeal. The archetypal superhero story is an id story: an almost pure power fantasy aimed at teenage boys. Like other pulp mass media, they reflect the prevailing cultural myths of the era in which they're told. World War II superheroes are mostly all-American boy scouts who punch Nazis. 1960s superheroes are a more complex mix of outsider misfits with a moral code and sarcastic but earnestly ethical do-gooders. The superhero genre is vast, with numerous reinterpretations, deconstructions, and alternate perspectives, but its ur-story is a good versus evil struggle of individual action, in which exceptional people use their powers for good to defeat nefarious villains. Watchmen was not the first internal critique of the genre, but it was the one that everyone read in the 1980s and 1990s. It takes direct aim at that moral binary. The superheroes in Watchmen are not paragons of virtue (some of them are truly horrible people), and they have just as much messy entanglement with the world as the rest of us. It was superheroes re-imagined for the post-Vietnam, post-Watergate era, for the end of the Cold War when we were realizing how many lies about morality we had been told. But it still put superheroes and their struggles with morality at the center of the story. Hench is a superhero story for the modern neoliberal world of reality TV and power inequality in the way that Watchmen was a superhero story for the Iran-Contra era and the end of the Cold War. Whether our heroes have feet of clay is no longer a question. Today, a better question is whether the official heroes, the ones that are celebrated as triumphs of individual achievement, are anything but clay. Hench doesn't bother asking whether superheroes have fallen short of their ideal; that answer is obvious. What Hench asks instead is a question familiar to those living in a world full of televangelists, climate denialism, manipulative advertising, and Facebook: are superheroes anything more than a self-perpetuating scam? Has the good superheroes supposedly do ever outweighed the collateral damage? Do they care in the slightest about the people they're supposedly protecting? Or is the whole system of superheroes and supervillains a performance for an audience, one that chews up bystanders and spits them out mangled while delivering simplistic and unquestioned official morality? This sounds like a deeply cynical premise, but Hench is not a cynical book. It is cynical about superheroes, which is not the same thing. The brilliance of Walschots's approach is that Anna has a foot in both worlds. She works for a supervillain and, over the course of the book, gains access to real power within the world of superheroic battles. But she's also an ordinary person with ordinary problems: not enough money, rocky friendships, deep anger at the injustices of the world and the way people like her are discarded, and now a disability and PTSD. Walschots perfectly balances the tension between those worlds and maintains that tension straight to the end of the book. From the supervillain world, Anna draws support, resources, and a mission, but all of the hope, true morality, and heart of this book comes from the ordinary side. If you had the infrastructure of a supervillain at your disposal, what would you do with it? Anna's answer is to treat superheroes as a destructive force like climate change, and to do whatever she can to drive them out of the business and thus reduce their impact on the world. The tool she uses for that is psychological warfare: make them so miserable that they'll snap and do something too catastrophic to be covered up. And the raw material for that psychological warfare is data. That's the foot in the supervillain world. In descriptions of this book, her skills with data are often called her superpower. That's not exactly wrong, but the reason why she gains power and respect is only partly because of her data skills. Anna lives by the morality of the ordinary people world: you look out for your friends, you treat your co-workers with respect as long as they're not assholes, and you try to make life a bit better for the people around you. When Leviathan gives her the opportunity to put together a team, she finds people with skills she admires, funnels work to people who are good at it, and worries about the team dynamics. She treats the other ordinary employees of a supervillain as people, with lives and personalities and emotions and worth. She wins their respect. Then she uses their combined skills to destroy superhero lives. I was fascinated by the moral complexity in this book. Anna and her team do villainous things by the morality of the superheroic world (and, honestly, by the morality of most readers), including some things that result in people's deaths. By the end of the book, one could argue that Anna has been driven by revenge into becoming an unusual sort of supervillain. And yet, she treats the people around her so much better than either the heroes or the villains do. Anna is fiercely moral in all the ordinary person ways, and that leads directly to her becoming a villain in the superhero frame. Hench doesn't resolve that conflict; it just leaves it on the page for the reader to ponder. The best part about this book is that it's absurdly grabby, unpredictable, and full of narrative momentum. Walschots's pacing kept me up past midnight a couple of times and derailed other weekend plans so that I could keep reading. I had no idea where the plot was going even at the 80% mark. The ending is ambiguous and a bit uncomfortable, just like the morality throughout the book, but I liked it the more I thought about it. One caveat, unfortunately: Hench has some very graphic descriptions of violence and medical procedures, and there's an extended torture sequence with some incredibly gruesome body horror that I thought went on far too long and was unnecessary to the plot. If you're a bit squeamish like I am, there are some places where you'll want to skim, including one sequence that's annoyingly intermixed with important story developments. Otherwise, though, this is a truly excellent book. It has a memorable protagonist with a great first-person voice, an epic character arc of empowerment and revenge, a timely take on the superhero genre that uses it for sharp critique of neoliberal governance and reality TV morality, a fascinatingly ambiguous and unsettled moral stance, a gripping and unpredictable plot, and some thoroughly enjoyable competence porn. I had put off reading it because I was worried that it would be too cynical or dark, but apart from the unnecessary torture scene, it's not at all. Highly recommended. Rating: 9 out of 10

3 January 2022

Ian Jackson: Debian s approach to Rust - Dependency handling

tl;dr: Faithfully following upstream semver, in Debian package dependencies, is a bad idea. Introduction I have been involved in Debian for a very long time. And I ve been working with Rust for a few years now. Late last year I had cause to try to work on Rust things within Debian. When I did, I found it very difficult. The Debian Rust Team were very helpful. However, the workflow and tooling require very large amounts of manual clerical work - work which it is almost impossible to do correctly since the information required does not exist. I had wanted to package a fairly straightforward program I had written in Rust, partly as a learning exercise. But, unfortunately, after I got stuck in, it looked to me like the effort would be wildly greater than I was prepared for, so I gave up. Since then I ve been thinking about what I learned about how Rust is packaged in Debian. I think I can see how to fix some of the problems. Although I don t want to go charging in and try to tell everyone how to do things, I felt I ought at least to write up my ideas. Hence this blog post, which may become the first of a series. This post is going to be about semver handling. I see problems with other aspects of dependency handling and source code management and traceability as well, and of course if my ideas find favour in principle, there are a lot of details that need to be worked out, including some kind of transition plan. How Debian packages Rust, and build vs runtime dependencies Today I will be discussing almost entirely build-dependencies; Rust doesn t (yet?) support dynamic linking, so built Rust binaries don t have Rusty dependencies. However, things are a bit confusing because even the Debian binary packages for Rust libraries contain pure source code. So for a Rust library package, building the Debian binary package from the Debian source package does not involve running the Rust compiler; it s just file-copying and format conversion. The library s Rust dependencies do not need to be installed on the build machine for this. So I m mostly going to be talking about Depends fields, which are Debian s way of talking about runtime dependencies, even though they are used only at build-time. The way this works is that some ultimate leaf package (which is supposed to produce actual executable code) Build-Depends on the libraries it needs, and those Depends on their under-libraries, so that everything needed is installed. What do dependencies mean and what are they for anyway? In systems where packages declare dependencies on other packages, it generally becomes necessary to support versioned dependencies. In all but the most simple systems, this involves an ordering (or similar) on version numbers and a way for a package A to specify that it depends on certain versions of B. Both Debian and Rust have this. Rust upstream crates have version numbers and can specify their dependencies according to semver. Debian s dependency system can represent that. So it was natural for the designers of the scheme for packaging Rust code in Debian to simply translate the Rust version dependencies to Debian ones. However, while the two dependency schemes seem equivalent in the abstract, their concrete real-world semantics are totally different. These different package management systems have different practices and different meanings for dependencies. (Interestingly, the Python world also has debates about the meaning and proper use of dependency versions.) The epistemological problem Consider some package A which is known to depend on B. In general, it is not trivial to know which versions of B will be satisfactory. I.e., whether a new B, with potentially-breaking changes, will actually break A. Sometimes tooling can be used which calculates this (eg, the Debian shlibdeps system for runtime dependencies) but this is unusual - especially for build-time dependencies. Which versions of B are OK can normally only be discovered by a human consideration of changelogs etc., or by having a computer try particular combinations. Few ecosystems with dependencies, in the Free Software community at least, make an attempt to precisely calculate the versions of B that are actually required to build some A. So it turns out that there are three cases for a particular combination of A and B: it is believed to work; it is known not to work; and: it is not known whether it will work. And, I am not aware of any dependency system that has an explicit machine-readable representation for the unknown state, so that they can say something like A is known to depend on B; versions of B before v1 are known to break; version v2 is known to work . (Sometimes statements like that can be found in human-readable docs.) That leaves two possibilities for the semantics of a dependency A depends B, version(s) V..W: Precise: A will definitely work if B matches V..W, and Optimistic: We have no reason to think B breaks with any of V..W. At first sight the latter does not seem useful, since how would the package manager find a working combination? Taking Debian as an example, which uses optimistic version dependencies, the answer is as follows: The primary information about what package versions to use is not only the dependencies, but mostly in which Debian release is being targeted. (Other systems using optimistic version dependencies could use the date of the build, i.e. use only packages that are current .)
Precise Optimistic
People involved in version management Package developers,
downstream developers/users.
Package developers,
downstream developer/users,
distribution QA and release managers.
Package developers declare versions V and dependency ranges V..W so that It definitely works. A wide range of B can satisfy the declared requirement.
The principal version data used by the package manager Only dependency versions. Contextual, eg, Releases - set(s) of packages available.
Version dependencies are for Selecting working combinations (out of all that ever existed). Sequencing (ordering) of updates; QA.
Expected use pattern by a downstream Downstream can combine any
declared-good combination.
Use a particular release of the whole system. Mixing-and-matching requires additional QA and remedial work.
Downstreams are protected from breakage by Pessimistically updating versions and dependencies whenever anything might go wrong. Whole-release QA.
A substantial deployment will typically contain Multiple versions of many packages. A single version of each package, except where there are actual incompatibilities which are too hard to fix.
Package updates are driven by Top-down:
Depending package updates the declared metadata.
Bottom-up:
Depended-on package is updated in the repository for the work-in-progress release.
So, while Rust and Debian have systems that look superficially similar, they contain fundamentally different kinds of information. Simply representing the Rust versions directly into Debian doesn t work. What is currently done by the Debian Rust Team is to manually patch the dependency specifications, to relax them. This is very labour-intensive, and there is little automation supporting either decisionmaking or actually applying the resulting changes. What to do Desired end goal To update a Rust package in Debian, that many things depend on, one need simply update that package. Debian s sophisticated build and CI infrastructure will try building all the reverse-dependencies against the new version. Packages that actually fail against the new dependency are flagged as suffering from release-critical problems. Debian Rust developers then update those other packages too. If the problems turn out to be too difficult, it is possible to roll back. If a problem with a depending packages is not resolved in a timely fashion, priority is given to updating core packages, and the depending package falls by the wayside (since it is empirically unmaintainable, given available effort). There is no routine manual patching of dependency metadata (or of anything else). Radical proposal Debian should not precisely follow upstream Rust semver dependency information. Instead, Debian should optimistically try the combinations of packages that we want to have. The resulting breakages will be discovered by automated QA; they will have to be fixed by manual intervention of some kind, but usually, simply updating the depending package will be sufficient. This no longer ensures (unlike the upstream Rust scheme) that the result is expected to build and work if the dependencies are satisfied. But as discussed, we don t really need that property in Debian. More important is the new property we gain: that we are able to mix and match versions that we find work in practice, without a great deal of manual effort. Or to put it another way, in Debian we should do as a Rust upstream maintainer does when they do the regular update dependencies for new semvers task: we should update everything, see what breaks, and fix those. (In theory a Rust upstream package maintainer is supposed to do some additional checks or something. But the practices are not standardised and any checks one does almost never reveal anything untoward, so in practice I think many Rust upstreams just update and see what happens. The Rust upstream community has other mechanisms - often, reactive ones - to deal with any problems. Debian should subscribe to those same information sources, eg RustSec.) Nobbling cargo Somehow, when cargo is run to build Rust things against these Debian packages, cargo s dependency system will have to be overridden so that the version of the package that is actually selected by Debian s package manager is used by cargo without complaint. We probably don t want to change the Rust version numbers of Debian Rust library packages, so this should be done by either presenting cargo with an automatically-massaged Cargo.toml where the dependency version restrictions are relaxed, or by using a modified version of cargo which has special option(s) to relax certain dependencies. Handling breakage Rust packages in Debian should already be provided with autopkgtests so that ci.debian.net will detect build breakages. Build breakages will stop the updated dependency from migrating to the work-in-progress release, Debian testing. To resolve this, and allow forward progress, we will usually upload a new version of the dependency containing an appropriate Breaks, and either file an RC bug against the depending package, or update it. This can be done after the upload of the base package. Thus, resolution of breakage due to incompatibilities will be done collaboratively within the Debian archive, rather than ad-hoc locally. And it can be done without blocking. My proposal prioritises the ability to make progress in the core, over stability and in particular over retaining leaf packages. This is not Debian s usual approach but given the Rust ecosystem s practical attitudes to API design, versioning, etc., I think the instability will be manageable. In practice fixing leaf packages is not usually really that hard, but it s still work and the question is what happens if the work doesn t get done. After all we are always a shortage of effort - and we probably still will be, even if we get rid of the makework clerical work of patching dependency versions everywhere (so that usually no work is needed on depending packages). Exceptions to the one-version rule There will have to be some packages that we need to keep multiple versions of. We won t want to update every depending package manually when this happens. Instead, we ll probably want to set a version number split: rdepends which want version <X will get the old one. Details - a sketch I m going to sketch out some of the details of a scheme I think would work. But I haven t thought this through fully. This is still mostly at the handwaving stage. If my ideas find favour, we ll have to do some detailed review and consider a whole bunch of edge cases I m glossing over. The dependency specification consists of two halves: the depending .deb s Depends (or, for a leaf package, Build-Depends) and the base .deb Version and perhaps Breaks and Provides. Even though libraries vastly outnumber leaf packages, we still want to avoid updating leaf Debian source packages simply to bump dependencies. Dependency encoding proposal Compared to the existing scheme, I suggest we implement the dependency relaxation by changing the depended-on package, rather than the depending one. So we retain roughly the existing semver translation for Depends fields. But we drop all local patching of dependency versions. Into every library source package we insert a new Debian-specific metadata file declaring the earliest version that we uploaded. When we translate a library source package to a .deb, the binary package build adds Provides for every previous version. The effect is that when one updates a base package, the usual behaviour is to simply try to use it to satisfy everything that depends on that base package. The Debian CI will report the build or test failures of all the depending packages which the API changes broke. We will have a choice, then: Breakage handling - update broken depending packages individually If there are only a few packages that are broken, for each broken dependency, we add an appropriate Breaks to the base binary package. (The version field in the Breaks should be chosen narrowly, so that it is possible to resolve it without changing the major version of the dependency, eg by making a minor source change.) When can then do one of the following: Breakage handling - declare new incompatible API in Debian If the API changes are widespread and many dependencies are affected, we should represent this by changing the in-Debian-source-package metadata to arrange for fewer Provides lines to be generated - withdrawing the Provides lines for earlier APIs. Hopefully examination of the upstream changelog will show what the main compat break is, and therefore tell us which Provides we still want to retain. This is like declaring Breaks for all the rdepends. We should do it if many rdepends are affected. Then, for each rdependency, we must choose one of the responses in the bullet points above. In practice this will often be a mass bug filing campaign, or large update campaign. Breakage handling - multiple versions Sometimes there will be a big API rewrite in some package, and we can t easily update all of the rdependencies because the upstream ecosystem is fragmented and the work involved in reconciling it all is too substantial. When this happens we will bite the bullet and include multiple versions of the base package in Debian. The old version will become a new source package with a version number in its name. This is analogous to how key C/C++ libraries are handled. Downsides of this scheme The first obvious downside is that assembling some arbitrary set of Debian Rust library packages, that satisfy the dependencies declared by Debian, is no longer necessarily going to work. The combinations that Debian has tested - Debian releases - will work, though. And at least, any breakage will affect only people building Rust code using Debian-supplied libraries. Another less obvious problem is that because there is no such thing as Build-Breaks (in a Debian binary package), the per-package update scheme may result in no way to declare that a particular library update breaks the build of a particular leaf package. In other words, old source packages might no longer build when exposed to newer versions of their build-dependencies, taken from a newer Debian release. This is a thing that already happens in Debian, with source packages in other languages, though. Semver violation I am proposing that Debian should routinely compile Rust packages against dependencies in violation of the declared semver, and ship the results to Debian s millions of users. This sounds quite alarming! But I think it will not in fact lead to shipping bad binaries, for the following reasons: The Rust community strongly values safety (in a broad sense) in its APIs. An API which is merely capable of insecure (or other seriously bad) use is generally considered to be wrong. For example, such situations are regarded as vulnerabilities by the RustSec project, even if there is no suggestion that any actually-broken caller source code exists, let alone that actually-broken compiled code is likely. The Rust community also values alerting programmers to problems. Nontrivial semantic changes to APIs are typically accompanied not merely by a semver bump, but also by changes to names or types, precisely to ensure that broken combinations of code do not compile. Or to look at it another way, in Debian we would simply be doing what many Rust upstream developers routinely do: bump the versions of their dependencies, and throw it at the wall and hope it sticks. We can mitigate the risks the same way a Rust upstream maintainer would: when updating a package we should of course review the upstream changelog for any gotchas. We should look at RustSec and other upstream ecosystem tracking and authorship information. Difficulties for another day As I said, I see some other issues with Rust in Debian. Thanks all for your attention!

comment count unavailable comments

Russ Allbery: Review: Crashed

Review: Crashed, by Adam Tooze
Publisher: Penguin Books
Copyright: 2018
Printing: 2019
ISBN: 0-525-55880-2
Format: Kindle
Pages: 615
The histories of the 2008 financial crisis that I have read focus almost exclusively on the United States. They also stop after the bank rescue and TARP or, if they press on into the aftermath, focus on the resulting damage to the US economy and the widespread pain of falling housing prices and foreclosure. Crashed does neither, instead arguing that 2008 was a crisis of European banks as much as American banks. It extends its history to cover the sovereign debt crisis in the eurozone, treating it as a continuation of the same crisis in a different guise. In the process, Tooze makes a compelling argument that one can draw a clear, if wandering, line from the moral revulsion at the propping up of the international banking system to Brexit and Trump. Qualifications first, since they are important for this type of comprehensive and, in places, surprising and counterintuitive history. Adam Tooze is Kathryn and Shelby Cullom Davis Professor of History at Columbia University and the director of its European Institute. His previous books have won multiple awards, and Crashed won the Lionel Gelber Prize for non-fiction on foreign policy. That it won a prize in that topic, rather than history or economics, is a hint at Tooze's chosen lens. The first half of the book is the lead-up and response to the crisis provoked by the collapse in value of securitized US mortgages and leading to the failure of Lehman Brothers, the failure in all but name of AIG, and a massive bank rescue. The financial instruments at the center of the crisis are complex and difficult to understand, and Tooze provides only brief explanation. This therefore may not be the best first book on the crisis; for that, I would still recommend Bethany McClean and Joe Nocera's All the Devils Are Here, although it's hard to beat Michael Lewis's storytelling in The Big Short. Tooze is not interested in dwelling on a blow-by-blow account of the crisis and initial response, and some of his account feels perfunctory. He is instead interested in describing its entangled global sweep. The new detail I took from the first half of Crashed is the depth of involvement of the European banks in what is often portrayed as a US crisis. Tooze goes into more specifics than other accounts on the eurodollar market, run primarily through the City of London, and the vast dollar-denominated liabilities of European banks. When the crisis struck, the breakdown of liquidity markets left those banks with no source of dollar funding to repay dollar-denominated short-term loans. The scale of dollar borrowing by European banks was vast, dwarfing the currency reserves or trade surpluses of their home countries. An estimate from the Bank of International Settlements put the total dollar funding needs for European banks at more than $2 trillion. The institution that saved the European banks was the United States Federal Reserve. This was an act of economic self-protection, not largesse; in the absence of dollar liquidity, the fire sale of dollar assets by European banks in a desperate attempt to cover their loans would have exacerbated the market crash. But it's remarkable in its extent, and in how deeply this contradicts the later public political position that 2008 was an American recession caused by American banks. 52% of the mortgage-backed securities purchased by the Federal Reserve in its quantitative easing policies (popularly known as QE1, QE2, and QE3) were sold by foreign banks. Deutsche Bank and Credit Suisse unloaded more securities on the Fed than any American bank by a significant margin. And when that wasn't enough, the Fed went farther and extended swap lines to major national banks, providing them dollar liquidity that they could then pass along to their local institutions. In essence, in Tooze's telling, the US Federal Reserve became the reserve bank for the entire world, preventing a currency crisis by providing dollars to financial systems both foreign and domestic, and it did so with a remarkable lack of scrutiny. Its swap lines avoided public review until 2010, when Bloomberg won a court fight to extract the records. That allowed the European banks that benefited to hide the extent of their exposure.
In Europe, the bullish CEOs of Deutsche Bank and Barclays claimed exceptional status because they avoided taking aid from their national governments. What the Fed data reveal is the hollowness of those boasts. The banks might have avoided state-sponsored recapitalization, but every major bank in the entire world was taking liquidity assistance on a grand scale from its local central bank, and either directly or indirectly by way of the swap lines from the Fed.
The emergency steps taken by Timothy Geithner in the Treasury Department were nearly as dramatic as those of the Federal Reserve. Without regard for borders, and pushing the boundary of their legal authority, they intervened massively in the world (not just the US) economy to save the banking and international finance system. And it worked. One of the benefits of a good history is to turn stories about heroes and villains into more nuanced information about motives and philosophies. I came away from Sheila Bair's account of the crisis furious at Geithner's protection of banks from any meaningful consequences for their greed. Tooze's account, and analysis, agrees with Bair in many respects, but Bair was continuing a personal fight and Tooze has more space to put Geithner into context. That context tells an interesting story about the shape of political economics in the 21st century. Tooze identifies Geithner as an institutionalist. His goal was to keep the system running, and he was acutely aware of what would happen if it failed. He therefore focused on the pragmatic and the practical: the financial system was about to collapse, he did whatever was necessary to keep it working, and that effort was successful. Fairness, fault, and morals were treated as irrelevant. This becomes more obvious when contrasted with the eurozone crisis, which started with a Greek debt crisis in the wake of the recession triggered by the 2008 crisis. Greece is tiny by the standards of the European economy, so at first glance there is no obvious reason why its debt crisis should have perturbed the financial system. Under normal circumstances, its lenders should have been able to absorb such relatively modest losses. But the immediate aftermath of the 2008 crisis was not normal circumstances, particularly in Europe. The United States had moved aggressively to recapitalize its banks using the threat of compensation caps and government review of their decisions. The European Union had not; European countries had done very little, and their banks were still in a fragile state. Worse, the European Central Bank had sent signals that the market interpreted as guaranteeing the safety of all European sovereign debt equally, even though this was explicitly ruled out by the Lisbon Treaty. If Greece defaulted on its debt, not only would that be another shock to already-precarious banks, it would indicate to the market that all European debt was not equal and other countries may also be allowed to default. As the shape of the Greek crisis became clearer, the cost of borrowing for all of the economically weaker European countries began rising towards unsustainable levels. In contrast to the approach taken by the United States government, though, Europe took a moralistic approach to the crisis. Jean-Claude Trichet, then president of the European Central Bank, held the absolute position that defaulting on or renegotiating the Greek debt was unthinkable and would not be permitted, even though there was no realistic possibility that Greece would be able to repay. He also took a conservative hard line on the role of the ECB, arguing that it could not assist in this crisis. (Tooze is absolutely scathing towards Trichet, who comes off in this account as rigidly inflexible, volatile, and completely irrational.) Germany's position, represented by Angela Merkel, was far more realistic: Greece's debt should be renegotiated and the creditors would have to accept losses. This is, in Tooze's account, clearly correct, and indeed is what eventually happened. But the problem with Merkel's position was the potential fallout. The German government was still in denial about the health of its own banks, and political opinion, particularly in Merkel's coalition, was strongly opposed to making German taxpayers responsible for other people's debts. Stopping the progression of a Greek default to a loss of confidence in other European countries would require backstopping European sovereign debt, and Merkel was not willing to support this. Tooze is similarly scathing towards Merkel, but I'm not sure it's warranted by his own account. She seemed, even in his account, boxed in by domestic politics and the tight constraints of the European political structure. Regardless, even after Trichet's term ended and he was replaced by the far more pragmatic Mario Draghi, Germany and Merkel continued to block effective action to relieve Greece's debt burden. As a result, the crisis lurched from inadequate stopgap to inadequate stopgap, forcing crippling austerity, deep depressions, and continued market instability while pretending unsustainable debt would magically become payable through sufficient tax increases and spending cuts. US officials such as Geithner, who put morals and arguably legality aside to do whatever was needed to save the system, were aghast. One takeaway from this is that expansionary austerity is the single worst macroeconomic idea that anyone has ever had.
In the summer of 2012 [the IMF's] staff revisited the forecasts they had made in the spring of 2010 as the eurozone crisis began and discovered that they had systematically underestimated the negative impact of budget cuts. Whereas they had started the crisis believing that the multiplier was on average around 0.5, they now concluded that from 2010 forward it had been in excess of 1. This meant that cutting government spending by 1 euro, as the austerity programs demanded, would reduce economic activity by more than 1 euro. So the share of the state in economic activity actually increased rather than decreased, as the programs presupposed. It was a staggering admission. Bad economics and faulty empirical assumptions had led the IMF to advocate a policy that destroyed the economic prospects for a generation of young people in Southern Europe.
Another takeaway, though, is central to Tooze's point in the final section of the book: the institutionalists in the United States won the war on financial collapse via massive state interventions to support banks and the financial system, a model that Europe grudgingly had to follow when attempting to reject it caused vast suffering while still failing to stabilize the financial system. But both did so via actions that were profoundly and obviously unfair, and only questionably legal. Bankers suffered few consequences for their greed and systematic mismanagement, taking home their normal round of bonuses while millions of people lost their homes and unemployment rates for young men in some European countries exceeded 50%. In Europe, the troika's political pressure against Greece and Italy was profoundly anti-democratic. The financial elite achieved their goal of saving the financial system. It could have failed, that failure would have been catastrophic, and their actions are defensible on pragmatic grounds. But they completely abandoned the moral high ground in the process. The political forces opposed to centrist neoliberalism attempted to step into that moral gap. On the Left, that came in the form of mass protest movements, Occupy Wall Street, Bernie Sanders, and parties such as Syriza in Greece. The Left, broadly, took the moral side of debtors, holding that the primary pain of the crisis should instead be born by the wealthy creditors who were more able to absorb it. The Right by contrast, in the form of the Tea Party movement inside the Republican Party in the United States and the nationalist parties in Europe, broadly blamed debtors for taking on excessive debt and focused their opposition on use of taxpayer dollars to bail out investment banks and other institutions of the rich. Tooze correctly points out that the Right's embrace of racist nationalism and incoherent demagoguery obscures the fact that their criticism of the elite center has real merit and is partly shared by the Left. As Tooze sketches out, the elite centrist consensus held in most of Europe, beating back challenges from both the Left and the Right, although it faltered in the UK, Poland, and Hungary. In the United States, the Democratic Party similarly solidified around neoliberalism and saw off its challenges from the Left. The Republican Party, however, essentially abandoned the centrist position, embracing the Right. That left the Democratic Party as the sole remaining neoliberal institutionalist party, supplemented by a handful of embattled Republican centrists. Wall Street and its money swung to the Democratic Party, but it was deeply unpopular on both the Left and the Right and this shift may have hurt them more than helped. The Democrats, by not abandoning the center, bore the brunt of the residual anger over the bank bailout and subsequent deep recession. Tooze sees in that part of the explanation for Trump's electoral victory over Hilary Clinton. This review is already much too long, and I haven't even mentioned Tooze's clear explanation of the centrality of treasury bonds to world finances, or his discussions of Russian and Ukraine, China, or Brexit, all of which I thought were excellent. This is not only an comprehensive history of both of the crises and international politics of the time period. It is also a thought-provoking look at how drastic of interventions are required to keep the supposed free market working, who is left to suffer after those interventions, and the political consequences of the choice to prioritize the stability of a deeply inequitable and unsafe financial system. At least in the United States, there is now a major political party that is likely to oppose even mundane international financial institutions, let alone another major intervention. The neoliberal center is profoundly weakened. But nothing has been done to untangle the international financial system, and little has been done to reduce its risk. The world will go into the next financial challenge still suffering from a legitimacy crisis. Given the miserly, condescending, and dismissive treatment of the suffering general populace after moving heaven and earth to save the banking system, that legitimacy crisis is arguably justified, but an uncontrolled crash of the financial system is not likely to be any kinder to the average citizen than it is to the investment bankers. Crashed is not the best-written book at a sentence-by-sentence level. Tooze's prose is choppy and a bit awkward, and his paragraphs occasionally wander away from a clear point. But the content is excellent and thought-provoking, filling in large sections of the crisis picture that I had not previously been aware of and making a persuasive argument for its continuing effects on current politics. Recommended if you're not tired of reading about financial crises. Rating: 8 out of 10

1 January 2022

Chris Lamb: Favourite books of 2021: Classics

In my three most recent posts, I went over the memoirs and biographies, the non-fiction and fiction I enjoyed in 2021. But in the last of my 2021 book-related posts, however, I'll be going over my favourite classics. Of course, the difference between regular fiction and a 'classic' is an ambiguous, arbitrary and often-meaningless distinction: after all, what does it matter if Hemingway's The Old Man and the Sea (from 1951) is a classic or not? The term also smuggles in some of the ethnocentric gatekeeping encapsulated in the term 'Western canon' too. Nevertheless, the label of 'classic' has some utility for me in that it splits up the vast amount of non-fiction I read in two... Books that just missed the cut here include: Oscar Wilde's The Picture of Dorian Gray (moody and hilarious, but I cannot bring myself to include it due to the egregious antisemitism); Tolstoy's The Kreutzer Sonata (so angry! so funny!); and finally Notes from Underground by Fyodor Dostoevsky. Of significant note, though, would be the ghostly The Turn of the Screw by Henry James.

Heart of Darkness (1899) Joseph Conrad Heart of Darkness tells the story of Charles Marlow, a sailor who accepts an assignment from a Belgian trading company as a ferry-boat captain in the African interior, and the novella is widely regarded as a critique of European colonial rule in Africa. Loosely remade by Francis Ford Coppola as Apocalypse Now (1979), I started this book with the distinct possibility that this superb film adaptation would, for a rare treat, be 'better than the book'. However, Conrad demolished this idea of mine within two chapters, yet also elevated the film to a new level as well. This was chiefly due to how observant Conrad was of the universals that make up human nature. Some of his insight pertains to the barbarism of the colonialists, of course, but Conrad applies his shrewd acuity to the at the smaller level as well. Some of these quotes are justly famous: Ah! but it was something to have at least a choice of nightmares, for example, as well as the reference to a fastidiously turned-out colonial administrator who, with unimaginable horrors occurring mere yards from his tent, we learn he was devoted to his books, which were in applepie order . (It seems to me to be deliberately unclear whether his devotion arises from gross inhumanity, utter denial or some combination of the two.) Oh, and there's a favourite moment of mine when a character remarks that It was very fine for a time, but after a bit I did get tired of resting. Tired of resting! Yes, it's difficult to now say something original about a many-layered classic such as this, especially one that has analysed from so many angles already; from a literary perspective at first, of course, but much later from a critical postcolonial perspective, such as in Chinua Achebe's noted 1975 lecture, An Image of Africa. Indeed, the history of criticism in the twentieth century of Heart of Darkness must surely parallel the social and political developments in the Western world. (On a highly related note, the much-cited non-fiction book King Leopold's Ghost is on my reading list for 2022.) I will therefore limit myself to saying that the boat physically falling apart as it journeys deeper into the Congo may be intended to represent that our idea of 'Western civilisation' ceases to function, both morally as well as physically, in this remote environment. And, whilst I'm probably not the first to notice the potential ambiguity, when Marlow lies to Kurtz's 'Intended [wife]' in the closing section in order to save her from being exposed to the truth about Kurtz (surely a metaphor about the ignorance of the West whilst also possibly incorporating some comment on gender?), the Intended replies: I knew it. For me, though, it is not beyond doubt that what the Intended 'knows' is that she knew that Marlow would lie to her: in other words, that the alleged ignorance of everyday folk in the colonial homeland is studied and deliberate. Compact and fairly easy-to-read, it is clear that Heart of Darkness rewards even the most rudimentary analysis.

Rebecca (1938) Daphne du Maurier Daphne du Maurier creates in Rebecca a credible and suffocating atmosphere in the shape of Manderley, a grand English mansion owned by aristocratic widower Maxim de Winter. Our unnamed narrator (a young woman seemingly na ve in the ways of the world) meets Max in Monte Carlo, and she soon becomes the second Mrs. de Winter. The tale takes a turn to the 'gothic', though, when it becomes apparent that the unemotional Max, as well as potentially Manderley itself, appears to be haunted by the memory of his late first wife, the titular Rebecca. Still, Rebecca is less of a story about supernatural ghosts than one about the things that can haunt our minds. For Max, this might be something around guilt; for our narrator, the class-centered fear that she will never fit in. Besides, Rebecca doesn't need an actual ghost when you have Manderley's overbearing housekeeper, Mrs Danvers, surely one of the creepiest characters in all of fiction. Either way, the conflict of a kind between the fears of the protagonists means that they never really connect with each other. The most obvious criticism of Rebecca is that the main character is unreasonably weak and cannot quite think or function on her own. (Isn't it curious that the trait of the male 'everyman' is a kind of physical clumsiness yet the female equivalent is shorthanded by being slightly slow?) But the na vete of Rebecca's narrator makes her easier to relate to in a way, and it also makes the reader far more capable of empathising with her embarrassment. This is demonstrated best whilst she, in one of the best evocations of this particular anxiety I have yet come across, is gingerly creeping around Manderlay and trying to avoid running into the butler. A surprise of sorts comes in the latter stages of the book, and this particular twist brings us into contact with a female character who is anything but 'credulous'. This revelation might even change your idea of who the main character of this book really is too. (Speaking of amateur literary criticism, I have many fan theories about Rebecca, including that Maxim de Winter's estate manager, Frank Crawley, is actually having an affair with Max, and also that Maxim may have a lot more involvement in Mrs Danvers final act that he lets on.) An easily accessible novel (with a great-but-not-perfect 1940 adaptation by Alfred Hitchcock, Rebecca is a real indulgence.

A Clockwork Orange (1962) Anthony Burgess One of Stanley Kubrick's most prominent tricks was to use different visual languages in order to prevent the audience from immediately grasping the underlying story. In his 1975 Barry Lyndon, for instance, the intentionally sluggish pacing and elusive characters require significant digestion to fathom and appreciate, and the luminous and quasi-Renaissance splendour of the cinematography does its part to constantly distract the viewer from the film's greater meaning. This is very much the case in Kubrick's A Clockwork Orange as well whilst it ostensibly appears to be about a Saturnalia of violence, the 'greater meaning' of A Clockwork Orange pertains to the Christian conception of free will; admittedly, a much drier idea to bother making a film around. This is all made much clearer when reading Anthony Burgess' 1962 original novel. Alex became a 'true Christian' through the experimental rehabilitation process, and even offers to literally turn the other cheek at one point. But as Alex had no choice to do so (and can no longer choose to commit violence), he is incapable of making a free moral choice. Thus, is he really a Man? Yet whilst the book's central concern is our conception of free will in modern societies, it also appears to be a repudiation of two conservative principles. Firstly, A Clockwork Orange demolishes the idea that 'high art' leads to morally virtuous citizens. After all, if you can do a bit of the old ultra-violence whilst listening to the glorious 9th by old Ludvig van, then so much for the oft-repeated claims that culture makes you better as a person. (This, at least, I already knew from personal experience.) The other repudiation in A Clockwork Orange is in regard to the pervasive idea that the countryside is a refuge from crime and sin. By contrast, we see the gang commit their most horrific violence in rural areas, and, later, Alex is taken to the countryside by his former droogs for a savage beating. Although this doesn't seem to quite fit the novel, this was actually an important point for Burgess to include: otherwise his book could easily be read as a commentary on the corrupting influence of urban spaces, rather than of modernity itself. The language of this book cannot escape comment here. Alex narrates most of the book in a language called Nadsat, a fractured slang constructed by Burgess based on Russian and Cockney rhyming slang. (The language is strange for only a few pages, I promise. And note that 'Alex' is a very common Russian name.) Using Nadsat has the effect of making the book feel distinctly alien, but it also prevents it from prematurely aging too. Indeed, it comes as bit of a shock to realise that A Clockwork Orange was published 1962, the same year as The Beatles' released their first single, Love Me Do. I could probably say a whole lot more about this thoroughly engrossing book and its movie adaptation (eg. the meta-textual line in Kubrick's version: It's funny how the colours of the real world only seem really real when you watch them on a screen... appears verbatim in the textual original), but I'll leave it there. The book of A Clockwork Orange is not only worth the investment in the language, but is, again, somehow better than the film.

The Great Gatsby (1925) F. Scott Fitzgerald I'm actually being a little deceitful by including this book here: I cannot really say that The Great Gatsby was a 'favourite' read of the year, but its literary merit is so undeniable (and my respect for Fitzgerald's achievement is deep enough) that the experience was one of those pleasures you feel at seeing anything done well. Here you have a book so rich in symbolic meaning that you could easily confuse the experience with drinking Coke syrup undiluted. And a text that has made the difficulty and complexity of reading character a prominent theme of the novel, as well as a technical concern of the book itself. Yet at all times you have in your mind that The Great Gatsby is first and foremost a book about a man writing a book, and, therefore, about the construction of stories and myths. What is the myth being constructed in Gatsby? The usual answer today is that the book is really about the moral virtues of America. Or, rather, the lack thereof. Indeed, as James Boice wrote in 2016:
Could Wilson have killed Gatsby any other way? Could he have ran him over, or poisoned him, or attacked him with a knife? Not at all this an American story, the quintessential one, so Gatsby could have only died the quintessential American death.
The quintessential American death is, of course, being killed with a gun. Whatever your own analysis, The Great Gatsby is not only magnificently written, but it is captivating to the point where references intrude many months later. For instance, when reading something about Disney's 'princess culture', I was reminded of when Daisy says of her daughter: I hope she'll be a fool that's the best thing of a girl can be in this world, a beautiful little fool . Or the billboard with the eyes of 'Doctor T. J. Eckleburg'. Or the fact that the books in Gatsby's library have never been read (so what is 'Owl Eyes' doing there during the party?!). And the only plain room in Gatsby's great house is his bedroom... Okay, fine, I must have been deluding myself: I love this novel.

31 December 2021

Chris Lamb: Favourite books of 2021: Fiction

In my two most recent posts, I listed the memoirs and biographies and followed this up with the non-fiction I enjoyed the most in 2021. I'll leave my roundup of 'classic' fiction until tomorrow, but today I'll be going over my favourite fiction. Books that just miss the cut here include Kingsley Amis' comic Lucky Jim, Cormac McCarthy's The Road (although see below for McCarthy's Blood Meridian) and the Complete Adventures of Tintin by Herg , the latter forming an inadvertently incisive portrait of the first half of the 20th century. Like ever, there were a handful of books that didn't live up to prior expectations. Despite all of the hype, Emily St. John Mandel's post-pandemic dystopia Station Eleven didn't match her superb The Glass Hotel (one of my favourite books of 2020). The same could be said of John le Carr 's The Spy Who Came in from the Cold, which felt significantly shallower compared to Tinker, Tailor, Soldier, Spy again, a favourite of last year. The strangest book (and most difficult to classify at all) was undoubtedly Patrick S skind's Perfume: The Story of a Murderer, and the non-fiction book I disliked the most was almost-certainly Beartown by Fredrik Bachman. Two other mild disappointments were actually film adaptions. Specifically, the original source for Vertigo by Pierre Boileau and Thomas Narcejac didn't match Alfred Hitchock's 1958 masterpiece, as did James Sallis' Drive which was made into a superb 2011 neon-noir directed by Nicolas Winding Refn. These two films thus defy the usual trend and are 'better than the book', but that's a post for another day.

A Wizard of Earthsea (1971) Ursula K. Le Guin How did it come to be that Harry Potter is the publishing sensation of the century, yet Ursula K. Le Guin's Earthsea is only a popular cult novel? Indeed, the comparisons and unintentional intertextuality with Harry Potter are entirely unavoidable when reading this book, and, in almost every respect, Ursula K. Le Guin's universe comes out the victor. In particular, the wizarding world that Le Guin portrays feels a lot more generous and humble than the class-ridden world of Hogwarts School of Witchcraft and Wizardry. Just to take one example from many, in Earthsea, magic turns out to be nurtured in a bottom-up manner within small village communities, in almost complete contrast to J. K. Rowling's concept of benevolent government departments and NGOs-like institutions, which now seems a far too New Labour for me. Indeed, imagine an entire world imbued with the kindly benevolence of Dumbledore, and you've got some of the moral palette of Earthsea. The gently moralising tone that runs through A Wizard of Earthsea may put some people off:
Vetch had been three years at the School and soon would be made Sorcerer; he thought no more of performing the lesser arts of magic than a bird thinks of flying. Yet a greater, unlearned skill he possessed, which was the art of kindness.
Still, these parables aimed directly at the reader are fairly rare, and, for me, remain on the right side of being mawkish or hectoring. I'm thus looking forward to reading the next two books in the series soon.

Blood Meridian (1985) Cormac McCarthy Blood Meridian follows a band of American bounty hunters who are roaming the Mexican-American borderlands in the late 1840s. Far from being remotely swashbuckling, though, the group are collecting scalps for money and killing anyone who crosses their path. It is the most unsparing treatment of American genocide and moral depravity I have ever come across, an anti-Western that flouts every convention of the genre. Blood Meridian thus has a family resemblance to that other great anti-Western, Once Upon a Time in the West: after making a number of gun-toting films that venerate the American West (ie. his Dollars Trilogy), Sergio Leone turned his cynical eye to the western. Yet my previous paragraph actually euphemises just how violent Blood Meridian is. Indeed, I would need to be a much better writer (indeed, perhaps McCarthy himself) to adequately 0utline the tone of this book. In a certain sense, it's less than you read this book in a conventional sense, but rather that you are forced to witness successive chapters of grotesque violence... all occurring for no obvious reason. It is often said that books 'subvert' a genre and, indeed, I implied as such above. But the term subvert implies a kind of Puck-like mischievousness, or brings to mind court jesters licensed to poke fun at the courtiers. By contrast, however, Blood Meridian isn't funny in the slightest. There isn't animal cruelty per se, but rather wanton negligence of another kind entirely. In fact, recalling a particular passage involving an injured horse makes me feel physically ill. McCarthy's prose is at once both baroque in its language and thrifty in its presentation. As Philip Connors wrote back in 2007, McCarthy has spent forty years writing as if he were trying to expand the Old Testament, and learning that McCarthy grew up around the Church therefore came as no real surprise. As an example of his textual frugality, I often looked for greater precision in the text, finding myself asking whether who a particular 'he' is, or to which side of a fight some two men belonged to. Yet we must always remember that there is no precision to found in a gunfight, so this infidelity is turned into a virtue. It's not that these are fair fights anyway, or even 'murder': Blood Meridian is just slaughter; pure butchery. Murder is a gross understatement for what this book is, and at many points we are grateful that McCarthy spares us precision. At others, however, we can be thankful for his exactitude. There is no ambiguity regarding the morality of the puppy-drowning Judge, for example: a Colonel Kurtz who has been given free license over the entire American south. There is, thank God, no danger of Hollywood mythologising him into a badass hero. Indeed, we must all be thankful that it is impossible to film this ultra-violent book... Indeed, the broader idea of 'adapting' anything to this world is, beyond sick. An absolutely brutal read; I cannot recommend it highly enough.

Bodies of Light (2014) Sarah Moss Bodies of Light is a 2014 book by Glasgow-born Sarah Moss on the stirrings of women's suffrage within an arty clique in nineteenth-century England. Set in the intellectually smoggy cities of Manchester and London, this poignant book follows the studiously intelligent Alethia 'Ally' Moberly who is struggling to gain the acceptance of herself, her mother and the General Medical Council. You can read my full review from July.

House of Leaves (2000) Mark Z. Danielewski House of Leaves is a remarkably difficult book to explain. Although the plot refers to a fictional documentary about a family whose house is somehow larger on the inside than the outside, this quotidian horror premise doesn't explain the complex meta-commentary that Danielewski adds on top. For instance, the book contains a large number of pseudo-academic footnotes (many of which contain footnotes themselves), with references to scholarly papers, books, films and other articles. Most of these references are obviously fictional, but it's the kind of book where the joke is that some of them are not. The format, structure and typography of the book is highly unconventional too, with extremely unusual page layouts and styles. It's the sort of book and idea that should be a tired gimmick but somehow isn't. This is particularly so when you realise it seems specifically designed to create a fandom around it and to manufacturer its own 'cult' status, something that should be extremely tedious. But not only does this not happen, House of Leaves seems to have survived through two exhausting decades of found footage: The Blair Witch Project and Paranormal Activity are, to an admittedly lesser degree, doing much of the same thing as House of Leaves. House of Leaves might have its origins in Nabokov's Pale Fire or even Derrida's Glas, but it seems to have more in common with the claustrophobic horror of Cube (1997). And like all of these works, House of Leaves book has an extremely strange effect on the reader or viewer, something quite unlike reading a conventional book. It wasn't so much what I got out of the book itself, but how it added a glow to everything else I read, watched or saw at the time. An experience.

Milkman (2018) Anna Burns This quietly dazzling novel from Irish author Anna Burns is full of intellectual whimsy and oddball incident. Incongruously set in 1970s Belfast during The Irish Troubles, Milkman's 18-year-old narrator (known only as middle sister ), is the kind of dreamer who walks down the street with a Victorian-era novel in her hand. It's usually an error for a book that specifically mention other books, if only because inviting comparisons to great novels is grossly ill-advised. But it is a credit to Burns' writing that the references here actually add to the text and don't feel like they are a kind of literary paint by numbers. Our humble narrator has a boyfriend of sorts, but the figure who looms the largest in her life is a creepy milkman an older, married man who's deeply integrated in the paramilitary tribalism. And when gossip about the narrator and the milkman surfaces, the milkman beings to invade her life to a suffocating degree. Yet this milkman is not even a milkman at all. Indeed, it's precisely this kind of oblique irony that runs through this daring but darkly compelling book.

The First Fifteen Lives of Harry August (2014) Claire North Harry August is born, lives a relatively unremarkable life and finally dies a relatively unremarkable death. Not worth writing a novel about, I suppose. But then Harry finds himself born again in the very same circumstances, and as he grows from infancy into childhood again, he starts to remember his previous lives. This loop naturally drives Harry insane at first, but after finding that suicide doesn't stop the quasi-reincarnation, he becomes somewhat acclimatised to his fate. He prospers much better at school the next time around and is ultimately able to make better decisions about his life, especially when he just happens to know how to stay out of trouble during the Second World War. Yet what caught my attention in this 'soft' sci-fi book was not necessarily the book's core idea but rather the way its connotations were so intelligently thought through. Just like in a musical theme and varations, the success of any concept-driven book is far more a product of how the implications of the key idea are played out than how clever the central idea was to begin with. Otherwise, you just have another neat Borges short story: satisfying, to be sure, but in a narrower way. From her relatively simple premise, for example, North has divined that if there was a community of people who could remember their past lives, this would actually allow messages and knowledge to be passed backwards and forwards in time. Ah, of course! Indeed, this very mechanism drives the plot: news comes back from the future that the progress of history is being interfered with, and, because of this, the end of the world is slowly coming. Through the lives that follow, Harry sets out to find out who is passing on technology before its time, and work out how to stop them. With its gently-moralising romp through the salient historical touchpoints of the twentieth century, I sometimes got a whiff of Forrest Gump. But it must be stressed that this book is far less certain of its 'right-on' liberal credentials than Robert Zemeckis' badly-aged film. And whilst we're on the topic of other media, if you liked the underlying conceit behind Stuart Turton's The Seven Deaths of Evelyn Hardcastle yet didn't enjoy the 'variations' of that particular tale, then I'd definitely give The First Fifteen Lives a try. At the very least, 15 is bigger than 7. More seriously, though, The First Fifteen Lives appears to reflect anxieties about technology, particularly around modern technological accelerationism. At no point does it seriously suggest that if we could somehow possess the technology from a decade in the future then our lives would be improved in any meaningful way. Indeed, precisely the opposite is invariably implied. To me, at least, homo sapiens often seems to be merely marking time until we can blow each other up and destroying the climate whilst sleepwalking into some crisis that might precipitate a thermonuclear genocide sometimes seems to be built into our DNA. In an era of cli-fi fiction and our non-fiction newspaper headlines, to label North's insight as 'prescience' might perhaps be overstating it, but perhaps that is the point: this destructive and negative streak is universal to all periods of our violent, insecure species.

The Goldfinch (2013) Donna Tartt After Breaking Bad, the second biggest runaway success of 2014 was probably Donna Tartt's doorstop of a novel, The Goldfinch. Yet upon its release and popular reception, it got a significant number of bad reviews in the literary press with, of course, an equal number of predictable think pieces claiming this was sour grapes on the part of the cognoscenti. Ah, to be in 2014 again, when our arguments were so much more trivial. For the uninitiated, The Goldfinch is a sprawling bildungsroman that centres on Theo Decker, a 13-year-old whose world is turned upside down when a terrorist bomb goes off whilst visiting the Metropolitan Museum of Art, killing his mother among other bystanders. Perhaps more importantly, he makes off with a painting in order to fulfil a promise to a dying old man: Carel Fabritius' 1654 masterpiece The Goldfinch. For the next 14 years (and almost 800 pages), the painting becomes the only connection to his lost mother as he's flung, almost entirely rudderless, around the Western world, encountering an array of eccentric characters. Whatever the critics claimed, Tartt's near-perfect evocation of scenes, from the everyday to the unimaginable, is difficult to summarise. I wouldn't label it 'cinematic' due to her evocation of the interiority of the characters. Take, for example: Even the suggestion that my father had close friends conveyed a misunderstanding of his personality that I didn't know how to respond it's precisely this kind of relatable inner subjectivity that cannot be easily conveyed by film, likely is one of the main reasons why the 2019 film adaptation was such a damp squib. Tartt's writing is definitely not 'impressionistic' either: there are many near-perfect evocations of scenes, even ones we hope we cannot recognise from real life. In particular, some of the drug-taking scenes feel so credibly authentic that I sometimes worried about the author herself. Almost eight months on from first reading this novel, what I remember most was what a joy this was to read. I do worry that it won't stand up to a more critical re-reading (the character named Xandra even sounds like the pharmaceuticals she is taking), but I think I'll always treasure the first days I spent with this often-beautiful novel.

Beyond Black (2005) Hilary Mantel Published about five years before the hyperfamous Wolf Hall (2004), Hilary Mantel's Beyond Black is a deeply disturbing book about spiritualism and the nature of Hell, somewhat incongruously set in modern-day England. Alison Harte is a middle-aged physic medium who works in the various towns of the London orbital motorway. She is accompanied by her stuffy assistant, Colette, and her spirit guide, Morris, who is invisible to everyone but Alison. However, this is no gentle and musk-smelling world of the clairvoyant and mystic, for Alison is plagued by spirits from her past who infiltrate her physical world, becoming stronger and nastier every day. Alison's smiling and rotund persona thus conceals a truly desperate woman: she knows beyond doubt the terrors of the next life, yet must studiously conceal them from her credulous clients. Beyond Black would be worth reading for its dark atmosphere alone, but it offers much more than a chilling and creepy tale. Indeed, it is extraordinarily observant as well as unsettlingly funny about a particular tranche of British middle-class life. Still, the book's unnerving nature that sticks in the mind, and reading it noticeably changed my mood for days afterwards, and not necessarily for the best.

The Wall (2019) John Lanchester The Wall tells the story of a young man called Kavanagh, one of the thousands of Defenders standing guard around a solid fortress that envelopes the British Isles. A national service of sorts, it is Kavanagh's job to stop the so-called Others getting in. Lanchester is frank about what his wall provides to those who stand guard: the Defenders of the Wall are conscripted for two years on the Wall, with no exceptions, giving everyone in society a life plan and a story. But whilst The Wall is ostensibly about a physical wall, it works even better as a story about the walls in our mind. In fact, the book blends together of some of the most important issues of our time: climate change, increasing isolation, Brexit and other widening societal divisions. If you liked P. D. James' The Children of Men you'll undoubtedly recognise much of the same intellectual atmosphere, although the sterility of John Lanchester's dystopia is definitely figurative and textual rather than literal. Despite the final chapters perhaps not living up to the world-building of the opening, The Wall features a taut and engrossing narrative, and it undoubtedly warrants even the most cursory glance at its symbolism. I've yet to read something by Lanchester I haven't enjoyed (even his short essay on cheating in sports, for example) and will be definitely reading more from him in 2022.

The Only Story (2018) Julian Barnes The Only Story is the story of Paul, a 19-year-old boy who falls in love with 42-year-old Susan, a married woman with two daughters who are about Paul's age. The book begins with how Paul meets Susan in happy (albeit complicated) circumstances, but as the story unfolds, the novel becomes significantly more tragic and moving. Whilst the story begins from the first-person perspective, midway through the book it shifts into the second person, and, later, into the third as well. Both of these narrative changes suggested to me an attempt on the part of Paul the narrator (if not Barnes himself), to distance himself emotionally from the events taking place. This effect is a lot more subtle than it sounds, however: far more prominent and devastating is the underlying and deeply moving story about the relationship ends up. Throughout this touching book, Barnes uses his mastery of language and observation to avoid the saccharine and the maudlin, and ends up with a heart-wrenching and emotive narrative. Without a doubt, this is the saddest book I read this year.

30 December 2021

Chris Lamb: Favourite books of 2021: Non-fiction

As a follow-up to yesterday's post listing my favourite memoirs and biographies I read in 2021, today I'll be outlining my favourite works of non-fiction. Books that just missed the cut include: The Unusual Suspect by Ben Machell for its thrilleresque narrative of a modern-day Robin Hood (and if you get to the end, a completely unexpected twist); Paul Fussell's Class: A Guide to the American Status System as an amusing chaser of sorts to Kate Fox's Watching the English; John Carey's Little History of Poetry for its exhilarating summation of almost four millennia of verse; David Graeber's Debt: The First 5000 Years for numerous historical insights, not least its rejoinder to our dangerously misleading view of ancient barter systems; and, although I didn't treasure everything about it, I won't hesitate to gift Pen Vogler's Scoff to a number of friends over the next year. The weakest book of non-fiction I read this year was undoubtedly Roger Scruton's How to Be a Conservative: I much preferred The Decadent Society for Ross Douthat for my yearly ration of the 'intellectual right'. I also very much enjoyed reading a number of classic texts from academic sociology, but they are difficult to recommend or even summarise. These included One-Dimensional Man by Herbert Marcuse, Postmodernism: Or, the Cultural Logic of Late Capitalism by Frederic Jameson and The Protestant Ethic and the Spirit of Capitalism by Max Weber. 'These are heavy books', remarks John Proctor in Arthur Miller's The Crucible... All round-up posts for 2021: Memoir/biography, Non-fiction (this post) & Fiction (coming soon).

Hidden Valley Road (2020) Robert Kolker A compelling and disturbing account of the Galvin family six of whom were diagnosed with schizophrenia which details a journey through the study and misunderstanding of the condition. The story of the Galvin family offers a parallel history of the science of schizophrenia itself, from the era of institutionalisation, lobotomies and the 'schizo mother', to the contemporary search for genetic markers for the disease... all amidst fundamental disagreements about the nature of schizophrenia and, indeed, of all illnesses of the mind. Samples of the Galvins' DNA informed decades of research which, curiously, continues to this day, potentially offering paths to treatment, prediction and even eradication of the disease, although on this last point I fancy that I detect a kind of neo-Victorian hubris that we alone will be the ones to find a cure. Either way, a gentle yet ultimately tragic view of a curiously 'American' family, where the inherent lack of narrative satisfaction brings a frustration and sadness of its own.

Islands of Abandonment: Life in the Post-Human Landscape (2021) Cat Flyn In this disarmingly lyrical book, Cat Flyn addresses the twin questions of what happens after humans are gone and how far can our damage to nature be undone. From the forbidden areas of post-war France to the mining regions of Scotland, Islands of Abandonment explores the extraordinary places where humans no longer live in an attempt to give us a glimpse into what happens when mankind's impact on nature is, for one reason or another, forced to stop. Needless to say, if anxieties in this area are not curdling away in your subconscious mind, you are probably in some kind of denial. Through a journey into desolate, eerie and ravaged areas in the world, this artfully-written study offers profound insights into human nature, eschewing the usual dry sawdust of Wikipedia trivia. Indeed, I summed it up to a close friend remarking that, through some kind of hilarious administrative error, the book's publisher accidentally dispatched a poet instead of a scientist to write this book. With glimmers of hope within the (mostly) tragic travelogue, Islands of Abandonment is not only a compelling read, but also a fascinating insight into the relationship between Nature and Man.

The Anatomy of Fascism (2004) Robert O. Paxton Everyone is absolutely sure they know what fascism is... or at least they feel confident choosing from a buffet of features to suit the political mood. To be sure, this is not a new phenomenon: even as 'early' as 1946, George Orwell complained in Politics and the English Language that the word Fascism has now no meaning except in so far as it signifies something not desirable . Still, it has proved uncommonly hard to define the core nature of fascism and what differentiates it from related political movements. This is still of great significance in the twenty-first century, for the definition ultimately determines where the powerful label of 'fascist' can be applied today. Part of the enjoyment of reading this book was having my own cosy definition thoroughly dismantled and replaced with a robust system of abstractions and common themes. This is achieved through a study of the intellectual origins of fascism and how it played out in the streets of Berlin, Rome and Paris. Moreover, unlike Strongmen (see above), fascisms that failed to gain meaningful power are analysed too, including Oswald Mosley's British Union of Fascists. Curiously enough, Paxton's own definition of fascism is left to the final chapter, and by the time you reach it, you get an anti-climatic feeling of it being redundant. Indeed, whatever it actually is, fascism is really not quite like any other 'isms' at all, so to try and classify it like one might be a mistake. In his introduction, Paxton warns that many of those infamous images associated with fascism (eg. Hitler in Triumph of the Will, Mussolini speaking from a balcony, etc.) have the ability to induce facile errors about the fascist leader and the apparent compliance of the crowd. (Contemporary accounts often record how sceptical the common man was of the leader's political message, even if they were transfixed by their oratorical bombast.) As it happens, I thus believe I had something of an advantage of reading this via an audiobook, and completely avoided re-absorbing these iconic images. To me, this was an implicit reminder that, however you choose to reduce it to a definition, fascism is undoubtedly the most visual of all political forms, presenting itself to us in vivid and iconic primary images: ranks of disciplined marching youths, coloured-shirted militants beating up members of demonised minorities; the post-war pictures from the concentration camps... Still, regardless of you choose to read it, The Anatomy of Fascism is a powerful book that can teach a great deal about fascism in particular and history in general.

What Good are the Arts? (2005) John Carey What Good are the Arts? takes a delightfully sceptical look at the nature of art, and cuts through the sanctimony and cant that inevitably surrounds them. It begins by revealing the flaws in lofty aesthetic theories and, along the way, debunks the claims that art makes us better people. They may certainly bring joy into your life, but by no means do the fine arts make you automatically virtuous. Carey also rejects the entire enterprise of separating things into things that are art and things that are not, making a thoroughly convincing case that there is no transcendental category containing so-called 'true' works of art. But what is perhaps equally important to what Carey is claiming is the way he does all this. As in, this is an extremely enjoyable book to read, with not only a fine sense of pace and language, but a devilish sense of humour as well. To be clear, What Good are the Arts? it is no crotchety monograph: Leo Tolstoy's *What Is Art? (1897) is hilarious to read in similar ways, but you can't avoid feeling its cantankerous tone holds Tolstoy's argument back. By contrast, Carey makes his argument in a playful sort of manner, in a way that made me slightly sad to read other polemics throughout the year. It's definitely not that modern genre of boomer jeremiad about the young, political correctness or, heaven forbid, 'cancel culture'... which, incidentally, made Carey's 2014 memoir, The Unexpected Professor something of a disappointing follow-up. Just for fun, Carey later undermines his own argument by arguing at length for the value of one art in particular. Literature, Carey asserts, is the only art capable of reasoning and the only art with the ability to criticise. Perhaps so, and Carey spends a chapter or so contending that fiction has the exclusive power to inspire the mind and move the heart towards practical ends... or at least far better than any work of conceptual art. Whilst reading this book I found myself taking down innumerable quotations and laughing at the jokes far more than I disagreed. And the sustained and intellectual style of polemic makes this a pretty strong candidate for my favourite overall book of the year.

29 December 2021

Chris Lamb: Favourite books of 2021: Memoir/biography

Just as I did for 2020, I won't publically disclose exactly how many books I read in 2021, but they evidently provoked enough thoughts that felt it worth splitting my yearly writeup into separate posts. I will reveal, however, that I got through more books than the previous year, and, like before, I enjoyed the books I read this year even more in comparison as well. How much of this is due to refining my own preferences over time, and how much can be ascribed to feeling less pressure to read particular books? It s impossible to say, and the question is complicated further by the fact I found many of the classics I read well worth of their entry into the dreaded canon. But enough of the throat-clearing. In today's post I'll be looking at my favourite books filed under memoir and biography, in no particular order. Books that just missed the cut here include: Bernard Crick's celebrated 1980 biography of George Orwell, if nothing else because it was a pleasure to read; Hilary Mantel's exhilaratingly bitter early memoir, Giving up the Ghost (2003); and Patricia Lockwood's hilarious Priestdaddy (2017). I also had a soft spot for Tim Kreider's We Learn Nothing (2012) as well, despite not knowing anything about the author in advance, likely a sign of good writing. The strangest book in this category I read was definitely Michelle Zauner's Crying in H Mart. Based on a highly-recommended 2018 essay in the New Yorker, its rich broth of genuine yearning for a departed mother made my eyebrows raise numerous times when I encountered inadvertent extra details about Zauner's relationships.

Beethoven: A Life in Nine Pieces (2020) Laura Tunbridge Whilst it might immediately present itself as a clickbait conceit, organising an overarching narrative around just nine compositions by Beethoven turns out to be an elegant way of saying something fresh about this grizzled old bear. Some of Beethoven's most famous compositions are naturally included in the nine (eg. the Eroica and the Hammerklavier piano sonata), but the book raises itself above conventional Beethoven fare when it highlights, for instance, his Septet, Op. 20, an early work that is virtually nobody's favourite Beethoven piece today. The insight here is that it was widely popular in its time, played again and again around Vienna for the rest of his life. No doubt many contemporary authors can relate to this inability to escape being artistically haunted by an earlier runaway success. The easiest way to say something interesting about Beethoven in the twenty-first century is to talk about the myth of Beethoven instead. Or, as Tunbridge implies, perhaps that should really be 'Beethoven' in leaden quotation marks, given so much about what we think we know about the man is a quasi-fictional construction. Take Anton Schindler, Beethoven's first biographer and occasional amanuensis, who destroyed and fabricated details about Beethoven's life, casting himself in a favourable light and exaggerating his influence with the composer. Only a few decades later, the idea of a 'heroic' German was to be politically useful as well; the Anglosphere often need reminding that Germany did not exist as a nation-state prior to 1871, so it should be unsurprising to us that the late nineteenth-century saw a determined attempt to create a uniquely 'German' culture ex nihilo. (And the less we say about Immortal Beloved the better, even though I treasure that film.) Nevertheless, Tunbridge cuts through Beethoven's substantial legacy using surgical precision that not only avoids feeling like it is settling a score, but it also does so in a way that is unlikely to completely alienate anyone emotionally dedicated to some already-established idea of the man to bring forth the tediously predictable sentiment that Beethoven has 'gone woke'. With Alex Ross on the cult of Wagner, it seems that books about the 'myth of X' are somewhat in vogue right now. And this pattern within classical music might fit into some broader trend of deconstruction in popular non-fiction too, especially when we consider the numerous contemporary books on the long hangover of the Civil Rights era (Robin DiAngelo's White Fragility, etc.), the multifarious ghosts of Empire (Akala's Natives, Sathnam Sanghera's Empireland, etc.) or even the 'transmogrification' of George Orwell into myth. But regardless of its place in some wider canon, A Life in Nine Pieces is beautifully printed in hardback form (worth acquiring for that very reason alone), and it is one of the rare good books about classical music that can be recommended to both the connoisseur and the layperson alike.

Sea State (2021) Tabitha Lasley In her mid-30s and jerking herself out of a terrible relationship, Tabitha Lasley left London and put all her savings into a six-month lease on a flat within a questionable neighbourhood in Aberdeen, Scotland. She left to make good on a lukewarm idea for a book about oil rigs and the kinds of men who work on them: I wanted to see what men were like with no women around, she claims. The result is Sea State, a forthright examination of the life of North Sea oil riggers, and an unsparing portrayal of loneliness, masculinity, female desire and the decline of industry in Britain. (It might almost be said that Sea State is an update of a sort to George Orwell's visit to the mines in the North of England.) As bracing as the North Sea air, Sea State spoke to me on multiple levels but I found it additionally interesting to compare and contrast with Julian Barnes' The Man with Red Coat (see below). Women writers are rarely thought to be using fiction for higher purposes: it is assumed that, unlike men, whatever women commit to paper is confessional without any hint of artfulness. Indeed, it seems to me that the reaction against the decades-old genre of autofiction only really took hold when it became the domain of millennial women. (By contrast, as a 75-year-old male writer with a firmly established reputation in the literary establishment, Julian Barnes is allowed wide latitude in what he does with his sources and his writing can be imbued with supremely confident airs as a result.) Furthermore, women are rarely allowed metaphor or exaggeration for dramatic effect, and they certainly aren t permitted to emphasise darker parts in order to explore them... hence some of the transgressive gratification of reading Sea State. Sea State is admittedly not a work of autofiction, but the sense that you are reading about an author writing a book is pleasantly unavoidable throughout. It frequently returns to the topic of oil workers who live multiple lives, and Lasley admits to living two lives herself: she may be in love but she's also on assignment, and a lot of the pleasure in this candid and remarkably accessible book lies in the way these states become slowly inseparable.

Twilight of Democracy (2020) Anne Applebaum For the uninitiated, Anne Applebaum is a staff writer for The Atlantic magazine who won a Pulitzer-prize for her 2004 book on the Soviet Gulag system. Her latest book, however, Twilight of Democracy is part memoir and part political analysis and discusses the democratic decline and the rise of right-wing populism. This, according to Applebaum, displays distinctly authoritarian tendencies, and who am I to disagree? Applebaum does this through three main case studies (Poland, the United Kingdom and the United States), but the book also touches on Hungary as well. The strongest feature of this engaging book is that Appelbaum's analysis focuses on the intellectual classes and how they provide significant justification for a descent into authoritarianism. This is always an important point to be remembered, especially as much of the folk understanding of the rise of authoritarian regimes tends to place exaggerated responsibility on the ordinary and everyday citizen: the blame placed on the working-class in the Weimar Republic or the scorn heaped upon 'white trash' of the contemporary Rust Belt, for example. Applebaum is uniquely poised to discuss these intellectuals because, well, she actually knows a lot of them personally. Or at least, she used to know them. Indeed, the narrative of the book revolves around two parties she hosted, both in the same house in northwest Poland. The first party, on 31 December 1999, was attended by friends from around the Western world, but most of the guests were Poles from the broad anti-communist alliance. They all agreed about democracy, the rule of law and the route to prosperity whilst toasting in the new millennium. (I found it amusing to realise that War and Peace also starts with a party.) But nearly two decades later, many of the attendees have ended up as supporters of the problematic 'Law and Justice' party which currently governs the country. Applebaum would now cross the road to avoid them, and they would do the same to her, let alone behave themselves at a cordial reception. The result of this autobiographical detail is that by personalising the argument, Applebaum avoids the trap of making too much of high-minded abstract argument for 'democracy', and additionally makes her book compellingly spicy too. Yet the strongest part of this book is also its weakest. By individualising the argument, it often feels that Applebaum is settling a number of personal scores. She might be very well justified in doing this, but at times it feels like the reader has walked in halfway through some personal argument and is being asked to judge who is in the right. Furthermore, Applebaum's account of contemporary British politics sometimes deviates into the cartoonish: nothing was egregiously incorrect in any of her summations, but her explanation of the Brexit referendum result didn't read as completely sound. Nevertheless, this lively and entertaining book that can be read with profit, even if you disagree with significant portions of it, and its highly-personal approach makes it a refreshing change from similar contemporary political analysis (eg. David Runciman's How Democracy Ends) which reaches for that more 'objective' line.

The Man in the Red Coat (2019) Julian Barnes As rich as the eponymous red coat that adorns his cover, Julian Barnes quasi-biography of French gynaecologist Samuel-Jean Pozzi (1846 1918) is at once illuminating, perplexing and downright hilarious. Yet even that short description is rather misleading, for this book evades classification all manner number of ways. For instance, it is unclear that, with the biographer's narrative voice so obviously manifest, it is even a biography in the useful sense of the word. After all, doesn't the implied pact between author and reader require the biographer to at least pretend that they are hiding from the reader? Perhaps this is just what happens when an author of very fine fiction turns his hand to non-fiction history, and, if so, it represents a deeper incursion into enemy territory after his 1984 metafictional Flaubert's Parrot. Indeed, upon encountering an intriguing mystery in Pozzi's life crying out for a solution, Barnes baldly turns to the reader, winks and states: These matters could, of course, be solved in a novel. Well, quite. Perhaps Barnes' broader point is that, given that's impossible for the author to completely melt into air, why not simply put down your cards and have a bit of fun whilst you're at it? If there's any biography that makes the case for a rambling and lightly polemical treatment, then it is this one. Speaking of having fun, however, two qualities you do not expect in a typical biography is simply how witty they can be, as well as it having something of the whiff of the thriller about it. A bullet might be mentioned in an early chapter, but given the name and history of Monsieur Pozzi is not widely known, one is unlikely to learn how he lived his final years until the closing chapters. (Or what happened to that turtle.) Humour is primarily incorporated into the book in two main ways: first, by explicitly citing the various wits of the day ( What is a vice? Merely a taste you don t share. etc.), but perhaps more powerful is the gentle ironies, bon mots and observations in Barnes' entirely unflappable prose style, along with the satire implicit in him writing this moreish pseudo-biography to begin with. The opening page, with its steadfast refusal to even choose where to begin, is somewhat characteristic of Barnes' method, so if you don't enjoy the first few pages then you are unlikely to like the rest. (Indeed, the whole enterprise may be something of an acquired taste. Like Campari.) For me, though, I was left wryly grinning and often couldn't wait to turn the page. Indeed, at times it reminded me of a being at a dinner party with an extremely charming guest at the very peak of his form as a wit and raconteur, delighting the party with his rambling yet well-informed discursive on his topic de jour. A significant book, and a book of significance.

28 December 2021

Russ Allbery: Review: Out of Office

Review: Out of Office, by Charlie Warzel & Anne Helen Petersen
Publisher: Alfred A. Knopf
Copyright: 2021
ISBN: 0-593-32010-7
Format: Kindle
Pages: 260
Out of Office opens with the provocative assertion that you were not working from home during the pandemic, even if you were among the 42% of Americans who were able to work remotely.
You were, quite literally, doing your job from home. But you weren't working from home. You were laboring in confinement and under duress. Others have described it as living at work. You were frantically tapping out an email while trying to make lunch and supervise distance learning. You were stuck alone in a cramped apartment for weeks, unable to see friends or family, exhausted, and managing a level of stress you didn't know was possible. Work became life, and life became work. You weren't thriving. You were surviving.
The stated goal of this book is to reclaim the concept of working from home, not only from the pandemic, but also from the boundary-destroying metastasis of work into non-work life. It does work towards that goal, but the description of what would be required for working from home to live up to its promise becomes a sweeping critique of the organization and conception of work, leaving it nearly as applicable to those who continue working from an office. Turns out that the main problem with working from home is the work part, not the "from home" part. This was a fascinating book to read in conjunction with A World Without Email. Warzel and Petersen do the the structural and political analysis that I sometimes wish Newport would do more of, but as a result offer less concrete advice. Both, however, have similar diagnoses of the core problems of the sort of modern office work that could be done from home: it's poorly organized, poorly managed, and desperately inefficient. Rather than attempting to fix those problems, which is difficult, structural, and requires thought and institutional cooperation, we're compensating by working more. This both doesn't work and isn't sustainable. Newport has a background in productivity books and a love of systems and protocols, so his focus in A World Without Email is on building better systems of communication and organization of work. Warzel and Petersen come from a background of reporting and cultural critique, so they put more focus on power imbalances and power-serving myths about the American dream. Where Newport sees an easy-to-deploy ad hoc work style that isn't fit for purpose, Warzel and Petersen are more willing to point out intentional exploitation of workers in the guise of flexibility. But they arrive at some similar conclusions. The way office work is organized is not leading to more productivity. Tools like Slack encourage the public performance of apparent productivity at the cost of the attention and focus required to do meaningful work. And the process is making us miserable. Out of Office is, in part, a discussion of what would be required to do better work with less stress, but it also shares a goal with Newport and some (but not most) corners of productivity writing: spend less time and energy on work. The goal of Out of Office is not to get more work done. It's to work more efficiently and sustainably and thus work less. To reclaim the promise of flexibility so that it benefits the employee and not the employer. To recognize, in the authors' words, that the office can be a bully, locking people in to commute schedules and unnatural work patterns, although it also provides valuable moments of spontaneous human connection. Out of Office tries to envision a style of work that includes the office sometimes, home sometimes, time during the day to attend to personal chores or simply to take a mental break from an unnatural eight hours (or more) of continuous focus, universal design, real worker-centric flexibility, and an end to the constant productivity ratchet where faster work simply means more work for the same pay. That's a lot of topics for a short book, and structurally this is a grab bag. Some sections will land and some won't. Loom's video messages sound like a nightmare to me, and I rolled my eyes heavily at the VR boosterism, reluctant as it may be. The section on DEI (diversity, equity, and inclusion) was a valiant effort that at least gestures towards the dismal track record of most such efforts, but still left me unconvinced that anyone knows how to improve diversity in an existing organization without far more brute-force approaches than anyone with power is usually willing to consider. But there's enough here, and the authors move through topics quickly enough, that a section that isn't working for you will soon be over. And some of the sections that do work are great. For example, the whole discussion of management.
Many of these companies view middle management as bloat, waste, what David Graeber would call a "bullshit job." But that's because bad management is a waste; you're paying someone more money to essentially annoy everyone around them. And the more people experience that sort of bad management, and think of it as "just the way it is," the less they're going to value management in general.
I admit to a lot of confirmation bias here, since I've been ranting about this for years, but management must be the most wide-spread professional job for which we ignore both training and capability and assume that anyone who can do any type of useful work can also manage people doing that work. It's simply not true, it creates workplaces full of horrible management, and that in turn creates a deep and unhelpful cynicism about all management. There is still a tendency on the left to frame this problem in terms of class struggle, on the reasonable grounds that for decades under "scientific management" of manufacturing that's what it was. Managers were there to overwork workers and extract more profits for the owners, and labor unions were there to fight back against managers. But while some of this does happen in the sort of office work this book is focused on, I think Warzel and Petersen correctly point to a different cause.
"The reason she was underpaid on the team was not because her boss was cackling in the corner. It was because nobody told the boss it was their responsibility to look at the fucking spreadsheet."
We don't train managers, we have no clear expectations for what managers should do, we don't meaningfully measure their performance, we accept a high-overhead and high-chaos workstyle based on ad hoc one-to-one communication that de-emphasizes management, and many managers have never seen good management and therefore have no idea what they're supposed to be doing. The management problem for many office workers is less malicious management than incompetent management, or simply no effective management at all apart from an occasional reorg and a complicated and mind-numbing annual review form. The last section of this book (apart from concluding letters to bosses and workers) is on community, and more specifically on extracting time and energy from work (via the roadmap in previous chapters) and instead investing it in the people around you. Much ink has been spilled about the collapse of American civic life, about how we went from a nation of joiners to a nation of isolated individual workers with weak and failing community institutions. Warzel and Petersen correctly lay some blame for this at the foot of work, and see the reorganization of work and an increase in work from home (and thus a decrease in commutes) as an opportunity to reverse that trend. David Brooks recently filled in for Ezra Klein on his podcast and talked with University of Chicago professor Leon Kass, which I listened to shortly after reading this book. In one segment, they talked about marriage and complained about the decline in marriage rates. They were looking for causes in people's moral upbringing, in their life priorities, in the lack of aspiration for permanence in kids these days, and in any other personal or moral failing that would allow them to be smugly judgmental. It was a truly remarkable thing to witness. Neither man at any point in the conversation mentioned either money or time. Back in the world most Americans live in, real wages have been stagnant for decades, student loan debt is skyrocketing as people desperately try to keep up with the ever-shifting requirements for a halfway-decent job, and work has expanded to fill all hours of the day, even for people who don't have to work multiple jobs to make ends meet. Employers have fully embraced a "flexible" workforce via layoffs, micro-optimizing work scheduling, eliminating benefits, relying on contract and gig labor, and embracing exceptional levels of employee turnover. The American worker has far less of money, time, and stability, three important foundations for marriage and family as well as participation in most other civic institutions. People like Brooks and Kass stubbornly cling to their feelings of moral superiority instead of seeing a resource crisis. Work has stolen the resources that people previously put into those other areas of their life. And it's not even using those resources effectively. That's, in a way, a restatement of the topic of this book. Our current way of organizing work is not sustainable, healthy, or wise. Working from home may be part of a strategy for changing it. The pandemic has already heavily disrupted work, and some of those changes, including increased working from home, seem likely to stick. That provides a narrow opportunity to renegotiate our arrangement with work and try to make those changes stick. I largely agree with the analysis, but I'm pessimistic. I think the authors are as well. We're very bad at social change, and there will be immense pressure for everything to go "back to normal." Those in the best bargaining position to renegotiate work for themselves are not in the habit of sharing that renegotiation with anyone else. But I'm somewhat heartened by how much public discussion there currently is about a more fundamental renegotiation of the rules of office work. I'm also reminded of a deceptively profound aphorism from economist Herbert Stein: "If something cannot go on forever, it will stop." This book is a bit uneven and is more of a collection of related thoughts than a cohesive argument, but if you are hungry for more worker-centric analyses of the dynamics of office work (inside or outside the office), I think it's worth reading. Rating: 7 out of 10

26 December 2021

Russ Allbery: rra-c-util 10.1

This is my collection of utility functions, Autoconf macros, test programs, and other support infrastructure for my other packages. Changes in this release: I go a few years between battles with perltidy, then dive in again, try to see if anything has gotten better, make a few tweaks, and leave vaguely frustrated. I've been so spoiled by Python's black that, despite knowing how much harder Perl is to parse, I keep hoping for the same thing. If anyone knows how to get perltidy to format long method calls the way black does, namely:
$spin = App::DocKnot::Spin->new(
      delete => 1, 'style-url' => '/~eagle/styles/'  
);
and not (as perltidy insists) either:
$spin
  = App::DocKnot::Spin->new(  delete => 1, 'style-url' => '/~eagle/styles/'  );
or:
$spin = App::DocKnot::Spin->new(
    delete => 1, 'style-url' => '/~eagle/styles/'  );
or some other weirdness, I'd love to hear it. Also this time around I discovered that changing the indent for continuation lines also changes the outdent for labels, because those are totally the same thing. A very perltidy sort of problem. Disabling vertical alignment (which makes code formatting very unstable and fiddly because perltidy is AMAZINGLY aggressive about what it tries to vertically align) means that perltidy instead aggressively breaks all of your vertical alignment. I just want => to line up, but not the arguments to any random function call or the equal signs for every random assignment. Is this too much to ask? (To be fair, black doesn't do this either, but for some reason it feels more important to line up fat commas in Perl than hash constructors in Python.) I should probably go file bugs against perltidy, and I admit I have not done that and thus there's no reason to expect anyone to know about my wishes. What I want seems different enough from how the program works that I'm not sure that would be appreciated, but I'm probably wrong. Anyway, you can get the latest version of rra-c-util from its distribution page.

19 December 2021

Russ Allbery: Review: Raybearer

Review: Raybearer, by Jordan Ifueko
Series: Raybearer #1
Publisher: Amulet Books
Copyright: 2020
ISBN: 1-68335-719-1
Format: Kindle
Pages: 308
Tarisai was raised alone in Bhekina House by an array of servants and tutors who were not allowed to touch her. Glimpses of the world were fleeting and sometimes ended by nailed-shut windows. Her life revolved around her rarely-seen mother, The Lady, who treats her with deep affection but rarely offers a word of praise, instead only pushing her to study harder. The servants whispered behind her back (but still in her hearing) that she was not human. At the age of seven, in a child's attempt to locate her absent mother by sneaking out of the house, she finds her father and is told a piece of the truth: she is the daughter of the Lady and a captive ehru, a djinn. At the age of eleven she's sent with two guardians to Oluwan City, the capital of Aritsar, to enter a competition she knows nothing about, for reasons no one has ever explained. Raybearer is a young adult fantasy novel, the first of a duology. Like a lot of young adult novels, it is a coming of age story that follows Tarisai from the end of her highly manipulated childhood through her introduction to a world she was carefully never taught about. Like a lot of young adult fantasy novels, Tarisai has some unusual abilities. What those are, and why she has them, is perhaps less obvious than it may appear at first. Unlike a lot of young adult fantasy novels, Raybearer is not set in a facsimile of Western Europe, the structure of gods and religion is not obviously derived from Christianity or Greek or Norse mythology, and neither Tarisai nor most of the characters of this story are white. Some of the characters are; Ifueko draws from a grab bag of cultures that does include European as well as African, Middle Eastern, and Asian. But the food, the physical descriptions, the landscape, and the hair and hair styles feel primarily African not in the sense of specific identifiable regions, but in the same way that most fantasy feels European even if the map isn't recognizable. That gives this story a freshness that I found delightful. The mythology of this world shares some similarities to standard fantasy tropes, including a bargain with the underworld that plays a similar role to fae bargains in some European fantasy, but it also goes in different directions and finds atypical balances, which gave the story room to catch me by surprise. The magical center of this book (and series), which Tarisai is carefully not told about until the story starts, is a system for anointing and protecting the emperor: selection of people who swear loyalty to him and each other and become his innermost circle, and thereby grant him magical protection. The emperor himself is the Raybearer, possessing an artifact that makes him invulnerable to one form of death for each member of his council he anoints. At eleven council members, he becomes invulnerable to anything but old age, or an attack from one of the council themselves. As the reader learns early in the book, that last part is important. Tarisai is an assassin; her mother's goal is for her to be selected as a member of the council for the prince, who will become the next emperor. But there is rather more to this system of magic than it may first appear, in a way that adds good depth to the mythology. And there is quite a bit more to Tarisai herself than anyone expects. Tarisai as a protagonist follows a more typical young adult pattern, but it's a formula that works for me. Her upbringing isolated from any other children has left her craving connection, but it also made her self-reliant, stubborn, and good at keeping her own counsel. One of the things that I loved about this book is that she's not thrown into a nest of vipers and cynical politics. Some of that is happening in the background, but the first step of her mother's plan is for her to earn the trust of the prince in a competition with other potential council members, all of whom are, well, kids. They fight (some), but they also make friends, helped along by the goal and requirement that they join a cooperative council or be sent home. That gives the plot a more collaborative and social feel than one would otherwise expect from the setup. Ifueko does a great job juggling a challenging cast size by focusing on a few kids with whom Tarisai strikes up a friendship but giving the others distinct-enough personalities that their presence is still felt in the story. There are two character dynamics that stand out: Tarisai's relationship with Prince Ekundayo, and her friendship with Sanjeet. The first carries much of the weight of the plot, of course; Tarisai is supposed to gain his trust and then kill him, and the reader will be unsurprised that this takes twists and turns no one expected. But Ifueko, refreshingly, does not reach for the stock plot development of a romance to complicate matters, even though many of the characters expect that. To the contrary, this is a rare story that at least hints at an acknowledgment that some people are not interested in romance at all, and there are other forms that mutual respect can take. Tarisai's relationship with Sanjeet is a different type of depth: two kids with very different histories finding a common understanding in the ways that they were both abused, and create space for each other. It's a great friendship that includes some deeply touching moments. It took me a bit to get into this book, but once Tarisai starts finding her feet and navigating her new relationships, I was engrossed. The story takes a sharp and nasty turn that was hard to read, but Ifueko chooses to turn it into a story of resiliency rather than survival, which makes it much easier to read than it could have been. She also pulls off the kind of plot that complicates and deepens the motives of the obvious villains in a way that gives the story much greater heft, but without disregarding the damage that they have done. I think the plot did fall apart a bit at the end of the book, with too much quick travel and world-building revelations at the cost of development of the relationships that were otherwise at the center of the book, but I'm hoping the sequel will pull those threads back together. And it's so refreshing to read a fantasy novel of this type with a different setting. It's not perfect: Ifueko falls back on Planet of the Hats regional characterization in a few places, and Songland is so obviously Korea that it felt jarring and out of place. Christianity also snuck its nose into the world-building tent near the end in ways that bugged me a bit, although it was subtle enough that I think most readers won't notice. But compared to most fantasy settings, it feels original and fresh. More of this! Ifueko starts this book with a wonderfully memorable dedication:
For the kid scanning fairy tales for a hero with a face like theirs. And for the girls whose stories we compressed into pities and wonders, triumphs and cautions, without asking, even once, for their names.
I think she was successful on both parts of that promise, and it makes for some great reading. Recommended. Followed by Redemptor. Rating: 8 out of 10

4 December 2021

Jonathan Dowland: Haskell mortgage calculator

A few months ago I was trying to compare two mortgage offers, and ended up writing a small mortgage calculator to help me. Both mortgages were fixed-term for the same time period (5 years). One of the mortgages had a lower rate than the other, but much higher arrangement fees. A broker recommended the mortgage with the higher rate but lower fee, on an affordability basis for the fixed term: over all, we would spend less money within the fixed term on that deal than the other. (I thought) this left one bit of information missing: what remaining balance would there be at the end of the term? The mortgages I want to model are defined in terms of a monthly repayment figure and an annual interest rate for the fixed period. I think interest is usually recalculated on a daily basis, so I convert the annual rate down to a daily rate. Repayments only happen once a month. Months are not all the same size. Using mod 30 on the 'day' approximates a monthly payment. Over 5 years, there would be 60 months, meaning 60 repayments. (I'm ignoring leap years)
 > length . filter id .take (5*365) $ [ x mod 30==0   x <- [1..]]
60
Here's what I came up with. I was a little concerned the repayment approximation was too far out so I compared the output with a more precise (but boring) spreadsheet and they agreed to within an acceptable tolerance. The numbers that follow are all made up to illustrate the function and don't reflect my actual mortgage. :)
borrowed = 1000000 -- day 0 amount outstanding
aer   = 0.89
repay = 1000
der   = aer / 36
owed n   n == 0          = borrowed
         n  mod  30 == 0 = last + interest - repay
         otherwise       = last + interest
    where
        last     = owed (n - 1)
        interest = last * der

Paul Tagliamonte: Receiving BPSK symbols (Part 3/5)

This post is part of a series called "PACKRAT". If this is the first post you've found, it'd be worth reading the intro post first and then looking over all posts in the series.
In the last post, we worked through how to generate a BPSK signal, and hopefully transmit it using one of our SDRs. Let s take that and move on to Receiving BPSK and turning that back into symbols! Demodulating BPSK data is a bit more tricky than transmitting BPSK data, mostly due to tedious facts of life such as space, time, and hardware built with compromises because not doing that makes the problem impossible. Unfortunately, it s now our job to work within our imperfect world to recover perfect data. We need to handle the addition of noise, differences in frequency, clock synchronization and interference in order to recover our information. This makes life a lot harder than when we transmit information, and as a result, a lot more complex.

Coarse Sync Our starting point for this section will be working from a capture of a number of generated PACKRAT packets as heard by a PlutoSDR at (xz compressed interleaved int16, 2,621,440 samples per second) Every SDR has its own oscillator, which eventually controls a number of different components of an SDR, such as the IF (if it s a superheterodyne architecture) and the sampling rate. Drift in oscillators lead to drifts in frequency such that what one SDR may think is 100MHz may be 100.01MHz for another radio. Even if the radios were perfectly in sync, other artifacts such as doppler time dilation due to motion can cause the frequency to appear higher or lower in frequency than it was transmitted. All this is a long way of saying, we need to determine when we see a strong signal that s close-ish to our tuned frequency, and take steps to roughly correct it to our center frequency (in the order of 100s of Hz to kHz) in order to acquire a phase lock on the signal to attempt to decode information contained within. The easiest way of detecting the loudest signal of interest is to use an FFT. Getting into how FFTs work is out of scope of this post, so if this is the first time you re seeing mention of an FFT, it may be a good place to take a quick break to learn a bit about the time domain (which is what the IQ data we ve been working with so far is), frequency domain, and how the FFT and iFFT operations can convert between them. Lastly, because FFTs average power over the window, swapping phases such that the transmitted wave has the same number of in-phase and inverted-phase symbols the power would wind up averaging to zero. This is not helpful, so I took a tip from Dr. Marc Lichtman s PySDR project and used complex squaring to drive our BPSK signal into a single detectable carrier by squaring the IQ data. Because points are on the unit circle and at tau/2 (specifically, tau/(2^1) for BPSK, 2^2 for QPSK) angles, and given that squaring has the effect of doubling the angle, and angles are all mod tau, this will drive our wave comprised of two opposite phases back into a continuous wave effectively removing our BPSK modulation, making it much easier to detect in the frequency domain. Thanks to Tom Bereknyei for helping me with that!
...
var iq []complex 
var freq []complex 
for i := range iq  
iq[i] = iq[i] * iq[i]
 
// perform an fft, computing the frequency
 // domain vector in  freq  given the iq data
 // contained in  iq .
 fft(iq, freq)
// get the array index of the max value in the
 // freq array given the magnitude value of the
 // complex numbers.
 var binIdx = max(abs(freq))
...
Now, most FFT operations will lay the frequency domain data out a bit differently than you may expect (as a human), which is that the 0th element of the FFT is 0Hz, not the most negative number (like in a waterfall). Generally speaking, zero first is the most common frequency domain layout (and generally speaking the most safe assumption if there s no other documentation on fft layout). Negative first is usually used when the FFT is being rendered for human consumption such as a waterfall plot. Given that we now know which FFT bin (which is to say, which index into the FFT array) contains the strongest signal, we ll go ahead and figure out what frequency that bin relates to. In the time domain, each complex number is the next time instant. In the frequency domain, each bin is a discrete frequency or more specifically a frequency range. The bandwidth of the bin is a function of the sampling rate and number of time domain samples used to do the FFT operation. As you increase the amount of time used to preform the FFT, the more precise the FFT measurement of frequency can be, but it will cover the same bandwidth, as defined by the sampling rate.
...
var sampleRate = 2,621,440
// bandwidth is the range of frequencies
 // contained inside a single FFT bin,
 // measured in Hz.
 var bandwidth = sampleRate/len(freq)
...
Now that we know we have a zero-first layout and the bin bandwidth, we can compute what our frequency offset is in Hz.
...
// binIdx is the index into the freq slice
 // containing the frequency domain data.
 var binIdx = 0
// binFreq is the frequency of the bin
 // denoted by binIdx
 var binFreq = 0
if binIdx > len(freq)/2  
// This branch covers the case where the bin
 // is past the middle point - which is to say,
 // if this is a negative frequency.
 binFreq = bandwidth * (binIdx - len(freq))
  else  
// This branch covers the case where the bin
 // is in the first half of the frequency array,
 // which is to say - if this frequency is
 // a positive frequency.
 binFreq = bandwidth * binIdx
 
...
However, sice we squared the IQ data, we re off in frequency by twice the actual frequency if we are reading 12kHz, the bin is actually 6kHz. We need to adjust for that before continuing with processing.
...
var binFreq = 0
...
// [compute the binFreq as above]
 ...
// Adjust for the squaring of our IQ data
 binFreq = binFreq / 2
...
Finally, we need to shift the frequency by the inverse of the binFreq by generating a carrier wave at a specific frequency and rotating every sample by our carrier wave so that a wave at the same frequency will slow down (or stand still!) as it approaches 0Hz relative to the carrier wave.
 var tau = pi * 2
// ts tracks where in time we are (basically: phase)
 var ts float
// inc is the amount we step forward in time (seconds)
 // each sample.
 var inc float = (1 / sampleRate)
// amount to shift frequencies, in Hz,
 // in this case, shift +12 kHz to 0Hz
 var shift = -12,000
for i := range iq  
ts += inc
if ts > tau  
// not actually needed, but keeps ts within
 // 0 to 2*pi (since it is modulus 2*pi anyway)
 ts -= tau
 
// Here, we're going to create a carrier wave
 // at the provided frequency (in this case,
 // -12kHz)
 cwIq = complex(cos(tau*shift*ts), sin(tau*shift*ts))
iq[i] = iq[i] * cwIq
 
Now we ve got the strong signal we ve observed (which may or may not be our BPSK modulated signal!) close enough to 0Hz that we ought to be able to Phase Lock the signal in order to begin demodulating the signal.

Filter After we re roughly in the neighborhood of a few kHz, we can now take some steps to cut out any high frequency components (both positive high frequencies and negative high frequencies). The normal way to do this would be to do an FFT, apply the filter in the frequency domain, and then do an iFFT to turn it back into time series data. This will work in loads of cases, but I ve found it to be incredibly tricky to get right when doing PSK. As such, I ve opted to do this the old fashioned way in the time domain. I ve again opted to go simple rather than correct, and haven t used nearly any of the advanced level trickery I ve come across for fear of using it wrong. As a result, our process here is going to be generating a sinc filter by computing a number of taps, and applying that in the time domain directly on the IQ stream.
// Generate sinc taps

func sinc(x float) float  
if x == 0  
return 1
 
var v = pi * x
return sin(v) / v
 
...
var dst []float
var length = float(len(dst))
if int(length)%2 == 0  
length++
 
for j := range dst  
i := float(j)
dst[j] = sinc(2 * cutoff * (i - (length-1)/2))
 
...
then we apply it in the time domain
...
// Apply sinc taps to an IQ stream

var iq []complex
// taps as created in  dst  above
 var taps []float
var delay = make([]complex, len(taps))
for i := range iq  
// let's shift the next sample into
 // the delay buffer
 copy(delay[1:], delay)
delay[0] = iq[i]
var phasor complex
for j := range delay  
// for each sample in the buffer, let's
 // weight them by the tap values, and
 // create a new complex number based on
 // filtering the real and imag values.
 phasor += complex(
taps[j] * real(delay[j]),
taps[j] * imag(delay[j]),
)
 
// now that we've run this sample
 // through the filter, we can go ahead
 // and scale it back (since we multiply
 // above) and drop it back into the iq
 // buffer.
 iq[i] = complex(
real(phasor) / len(taps),
imag(phasor) / len(taps),
)
 
...
After running IQ samples through the taps and back out, we ll have a signal that s been filtered to the shape of our designed Sinc filter which will cut out captured high frequency components (both positive and negative). Astute observers will note that we re using the real (float) valued taps on both the real and imaginary values independently. I m sure there s a way to apply taps using complex numbers, but it was a bit confusing to work through without being positive of the outcome. I may revisit this in the future!

Downsample Now, post-filter, we ve got a lot of extra RF bandwidth being represented in our IQ stream at our high sample rate All the high frequency values are now filtered out, which means we can reduce our sampling rate without losing much information at all. We can either do nothing about it and process at the fairly high sample rate we re capturing at, or we can drop the sample rate down and help reduce the volume of numbers coming our way. There s two big ways of doing this; either you can take every Nth sample (e.g., take every other sample to half the sample rate, or take every 10th to decimate the sample stream to a 10th of what it originally was) which is the easiest to implement (and easy on the CPU too), or to average a number of samples to create a new sample. A nice bonus to averaging samples is that you can trade-off some CPU time for a higher effective number of bits (ENOB) in your IQ stream, which helps reduce noise, among other things. Some hardware does exactly this (called Oversampling ), and like many things, it has some pros and some cons. I ve opted to treat our IQ stream like an oversampled IQ stream and average samples to get a marginal bump in ENOB. Taking a group of 4 samples and averaging them results in a bit of added precision. That means that a stream of IQ data at 8 ENOB can be bumped to 9 ENOB of precision after the process of oversampling and averaging. That resulting stream will be at 1/4 of the sample rate, and this process can be repeated 4 samples can again be taken for a bit of added precision; which is going to be 1/4 of the sample rate (again), or 1/16 of the original sample rate. If we again take a group of 4 samples, we ll wind up with another bit and a sample rate that s 1/64 of the original sample rate.

Phase Lock Our starting point for this section is the same capture as above, but post-coarse sync, filtering downsampling (xz compressed interleaved float32, 163,840 samples per second) The PLL in PACKRAT was one of the parts I spent the most time stuck on. There s no shortage of discussions of how hardware PLLs work, or even a few software PLLs, but very little by way of how to apply them and/or troubleshoot them. After getting frustrated trying to follow the well worn path, I decided to cut my own way through the bush using what I had learned about the concept, and hope that it works well enough to continue on. PLLs, in concept are fairly simple you generate a carrier wave at a frequency, compare the real-world SDR IQ sample to where your carrier wave is in phase, and use the difference between the local wave and the observed wave to adjust the frequency and phase of your carrier wave. Eventually, if all goes well, that delta is driven as small as possible, and your carrier wave can be used as a reference clock to determine if the observed signal changes in frequency or phase. In reality, tuning PLLs is a total pain, and basically no one outlines how to apply them to BPSK signals in a descriptive way. I ve had to steal an approach I ve seen in hardware to implement my software PLL, with any hope it s close enough that this isn t a hazard to learners. The concept is to generate the carrier wave (as above) and store some rolling averages to tune the carrier wave over time. I use two constants, alpha and beta (which appear to be traditional PLL variable names for this function) which control how quickly the frequency and phase is changed according to observed mismatches. Alpha is set fairly high, which means discrepancies between our carrier and observed data are quickly applied to the phase, and a lower constant for Beta, which will take long-term errors and attempt to use that to match frequency. This is all well and good. Getting to this point isn t all that obscure, but the trouble comes when processing a BPSK signal. Phase changes kick the PLL out of alignment and it tends to require some time to get back into phase lock, when we really shouldn t even be loosing it in the first place. My attempt is to generate two predicted samples, one for each phase of our BPSK signal. The delta is compared, and the lower error of the two is used to adjust the PLL, but the carrier wave itself is used to rotate the sample.
 var alpha = 0.1
var beta = (alpha * alpha) / 2
var phase = 0.0
var frequency = 0.0
...
for i := range iq  
predicted = complex(cos(phase), sin(phase))
sample = iq[i] * conj(predicted)
delta = phase(sample)
predicted2 = complex(cos(phase+pi), sin(phase+pi))
sample2 = iq[i] * conj(predicted2)
delta2 = phase(sample2)
if abs(delta2) < abs(delta)  
// note that we do not update 'sample'.
 delta = delta2
 
phase += alpha * delta
frequency += beta * delta
// adjust the iq sample to the PLL rotated
 // sample.
 iq[i] = sample
 
...
If all goes well, this loop has the effect of driving a BPSK signal s imaginary values to 0, and the real value between +1 and -1.

Average Idle / Carrier Detect Our starting point for this section is the same capture as above, but post-PLL (xz compressed interleaved float32, 163,840 samples per second) When we start out, we have IQ samples that have been mostly driven to an imaginary component of 0 and real value range between +1 and -1 for each symbol period. Our goal now is to determine if we re receiving a signal, and if so, determine if it s +1 or -1. This is a deceptively hard problem given it spans a lot of other similarly entertaining hard problems. I ve opted to not solve the hard problems involved and hope that in practice my very haphazard implementation works well enough. This turns out to be both good (not solving a problem is a great way to not spend time on it) and bad (turns out it does materially impact performance). This segment is the one I plan on revisiting, first. Expect more here at some point! Given that I want to be able to encapsulate three states in the output from this section (our Symbols are no carrier detected ( 0 ), real value 1 ( 1 ) or real value -1 ("-1")), which means spending cycles to determine what the baseline noise is to try and identify when a signal breaks through the noise becomes incredibly important.
var idleThreshold
var thresholdFactor = 10
...
// sigThreshold is used to determine if the symbol
 // is -1, +1 or 0. It's 1.3 times the idle signal
 // threshold.
 var sigThreshold = (idleThreshold * 0.3) + idleThreshold
// iq contains a single symbol's worth of IQ samples.
 // clock alignment isn't really considered; so we'll
 // get a bad packet if we have a symbol transition
 // in the middle of this buffer. No attempt is made
 // to correct for this yet.
 var iq []complex
// avg is used to average a chunk of samples in the
 // symbol buffer.
 var avg float
var mid = len(iq) / 2
// midNum is used to determine how many symbols to
 // average at the middle of the symbol.
 var midNum = len(iq) / 50
for j := mid; j < mid+midNum; j++  
avg += real(iq[j])
 
avg /= midNum
var symbol float
switch  
case avg > sigThreshold:
symbol = 1
case avg < -sigThreshold:
symbol = -1
default:
symbol = 0
// update the idleThreshold using the thresholdFactor
 // to average the idleThreshold over more samples to
 // get a better idea of average noise.
 idleThreshold = (
(idleThreshold*(thresholdFactor-1) + symbol) \
/ thresholdFactor
)
 
// write symbol to output somewhere
...

Next Steps Now that we have a stream of values that are either +1, -1 or 0, we can frame / unframe the data contained in the stream, and decode Packets contained inside, coming next in Part 4!

1 December 2021

Russ Allbery: Review: A World Without Email

Review: A World Without Email, by Cal Newport
Publisher: Portfolio/Penguin
Copyright: 2021
ISBN: 0-525-53657-4
Format: Kindle
Pages: 264
A World Without Email is the latest book by computer science professor and productivity writer Cal Newport. After a detour to comment on the drawbacks of social media in Digital Minimalism, Newport is back to writing about focus and concentration in the vein of Deep Work. This time, though, the topic is workplace structure and collaborative process rather than personal decisions. This book is a bit hard for me to review because I spoiled myself for the contents by listening to a lot of Newport's podcast, where he covers the same material. I therefore didn't enjoy it as much as I otherwise would have because the ideas were familiar. I recommend the book over the podcast, though; it's tighter, more coherent, and more comprehensive. The core contention of this book is that knowledge work (roughly, jobs where one spends significant time working on a computer processing information) has stumbled into a superficially tempting but inefficient and psychologically harmful structure that Newport calls the hyperactive hive mind. This way of organizing work is a local maxima: it feels productive, it's flexible and very easy to deploy, and most minor changes away from it make overall productivity worse. However, the incentive structure is all wrong. It prioritizes quick responses and coordination overhead over deep thinking and difficult accomplishments. The characteristic property of the hyperactive hive mind is free-flowing, unstructured communication between co-workers. If you need something from someone else, you ask them for it and they send it to you. The "email" in the title is not intended literally; Slack and related instant messaging apps are even more deeply entrenched in the hyperactive hive mind than email is. The key property of this workflow is that most collaborative work is done by contacting other people directly via ad hoc, unstructured messages. Newport's argument is that this workflow has multiple serious problems, not the least of which is that it makes us miserable. If you have read his previous work, you will correctly expect this to tie into his concept of deep work. Ad hoc, unstructured communication creates a constant barrage of unimportant small tasks and interrupts, most of which require several asynchronous exchanges before your brain can stop tracking the task. This creates constant context-shifting, loss of focus and competence, and background stress from ever-growing email inboxes, unread message notifications, and the semi-frantic feeling that you're forgetting something you need to do. This is not an original observation, of course. Many authors have suggested individual ways to improve this workflow: rules about how often to check one's email, filtering approaches, task managers, and other personal systems. Newport's argument is that none of these individual approaches can address the problem due to social effects. It's all well and good to say that you should unplug from distractions and ignore requests while you concentrate, but everyone else's workflow assumes that their co-workers are responsive to ad hoc requests. Ignoring this social contract makes the job of everyone still stuck in the hyperactive hive mind harder. They won't appreciate that, and your brain will not be able to relax knowing that you're not meeting your colleagues' expectations. In Newport's analysis, the necessary solution is a comprehensive redesign of how we do knowledge work, akin to the redesign of factory work that came with the assembly line. It's a collective problem that requires a collective solution. In other industries, organizing work for efficiency and quality is central to the job of management, but in knowledge work (for good historical reasons) employees are mostly left to organize their work on their own. That self-organization has produced a system that doesn't require centralized coordination or decisions and provides a lot of superficial flexibility, but which may be significantly inferior to a system designed for how people think and work. Even if you find this convincing (and I think Newport makes a good case), there are reasons to be suspicious of corporations trying to make people more productive. The assembly line made manufacturing much more efficient, but it also increased the misery of workers so much that Henry Ford had to offer substantial raises to retain workers. As one of Newport's knowledge workers, I'm not enthused about that happening to my job. Newport recognizes this and tries to address it by drawing a distinction between the workflow (how information moves between workers) and the work itself (how individual workers solve problems in their area of expertise). He argues that companies need to redesign the former, but should leave the latter to each worker. It's a nice idea, and it will probably work in industries like tech with substantial labor bargaining power. I'm more cynical about other industries. The second half of the book is Newport's specific principles and recommendations for designing better workflows that don't rely on unstructured email. Some of this will be familiar (and underwhelming) to anyone who works in tech; Newport recommends ticket systems and thinks agile, scrum, and kanban are pointed in the right direction. But there are some other good ideas in here, such as embracing specialization. Newport argues (with some evidence) that the drastic reduction in secretarial jobs, on the grounds that workers with computers can do the same work themselves, was a mistake. Even with new automation, this approach increased the range of tasks required in every other job. Not only was this a drain on the time of other workers, it caused more context switching, which made everyone less efficient and undermined work quality. He argues for reversing that trend: where the work cannot be automated, hire more support workers and more specialized workers in general, stop expecting everyone to be their own generalist admin, and empower support workers to create better systems rather than using the hyperactive hive mind model to answer requests. There's more here, ranging from specifics of how to develop a structured process for a type of work to the importance of enabling sustained concentration on a task. It's a less immediately actionable book than Newport's previous writing, but I welcome the partial shift in focus to more systemic issues. Newport continues to be relentlessly apolitical, but here it feels less like he's eliding important analysis and more like he thinks the interests of workers and good employers are both served by the approach he's advocating. I will warn that Newport leans heavily on evolutionary psychology in his argument that the hyperactive hive mind is bad for us. I think he has some good arguments about the anxiety that comes with not responding to requests from others, but I'm not sure intrusive experiments on spectacularly-unusual remnant hunter-gatherer groups, who are treated like experimental animals, are the best way of making that case. I realize this isn't Newport's research, but I think he could have made his point with more directly relevant experiments. He also continues his obsession with the superiority of in-person conversation over written communication, and while he has a few good arguments, he has a tendency to turn them into sweeping generalizations that are directly contradicted by, well, my entire life. It would be nice if he were more willing to acknowledge that it's possible to express deep emotional nuance and complex social signaling in writing; it simply requires a level of practice and familiarity (and shared vocabulary) that's often missing from the workplace. I was muttering a lot near the start of this book, but thankfully those sections are short, and I think the rest of his argument sits on a stronger foundation. I hope Newport continues moving in the direction of more systemic analysis. If you enjoyed Deep Work, you will probably find A World Without Email interesting. If you're new to Newport, this is not a bad place to start, particularly if you have influence on how communication is organized in your workplace. Those who work in tech will find some bits of this less interesting, but Newport approaches the topic from a different angle than most agile books and covers a broader range if ideas. Recommended if you like reading this sort of thing. Rating: 7 out of 10

30 November 2021

Russell Coker: Your Device Has Been Improved

I ve just started a Samsung tablet downloading a 770MB update, the description says:
Technically I have no doubt that both those claims are true and accurate. But according to common understanding of the English language I think they are both misleading. By stability improved they mean fixed some bugs that made it unstable and no technical person would imagine that after a certain number of such updates the number of bugs will ever reach zero and the tablet will be perfectly reliable. In fact if you should consider yourself lucky if they fix more bugs than they add. It s not THAT uncommon for phones and tablets to be bricked (rendered unusable by software) by an update. In the past I got a Huawei Mate9 as a warranty replacement for a Nexus 6P because an update caused so many Nexus 6P phones to fail that they couldn t be replaced with an identical phone [1]. By security improved they usually mean fixed some security flaws that were recently discovered to make it almost as secure as it was designed to be . Note that I deliberately say almost as secure because it s sometimes impossible to fix a security flaw without making significant changes to interfaces which requires more work than desired for an old product and also gives a higher probability of things going wrong. So it s sometimes better to aim for almost as secure or alternatively just as secure but with some features disabled. Device manufacturers (and most companies in the Android space make the same claims while having the exact same bugs to deal with, Samsung is no different from the others in this regards) are not making devices more secure or more reliable than when they were initially released. They are aiming to make them almost as secure and reliable as when they were released. They don t have much incentive to try too hard in this regard, Samsung won t suffer if I decide my old tablet isn t reliable enough and buy a new one, which will almost certainly be from Samsung because they make nice tablets. As a thought experiment, consider if car repairers did the same thing. Getting us to service your car will improve fuel efficiency , great how much more efficient will it be than when I purchased it? As another thought experiment, consider if car companies stopped providing parts for car repair a few years after releasing a new model. This is effectively what phone and tablet manufacturers have been doing all along, software updates for stability and security are to devices what changing oil etc is for cars.

21 November 2021

Antoine Beaupr : mbsync vs OfflineIMAP

After recovering from my latest email crash (previously, previously), I had to figure out which tool I should be using. I had many options but I figured I would start with a popular one (mbsync). But I also evaluated OfflineIMAP which was resurrected from the Python 2 apocalypse, and because I had used it before, for a long time. Read on for the details.

Benchmark setup All programs were tested against a Dovecot 1:2.3.13+dfsg1-2 server, running Debian bullseye. The client is a Purism 13v4 laptop with a Samsung SSD 970 EVO 1TB NVMe drive. The server is a custom build with a AMD Ryzen 5 2600 CPU, and a RAID-1 array made of two NVMe drives (Intel SSDPEKNW010T8 and WDC WDS100T2B0C). The mail spool I am testing against has almost 400k messages and takes 13GB of disk space:
$ notmuch count --exclude=false
372758
$ du -sh --exclude xapian Maildir
13G Maildir
The baseline we are comparing against is SMD (syncmaildir) which performs the sync in about 7-8 seconds locally (3.5 seconds for each push/pull command) and about 10-12 seconds remotely. Anything close to that or better is good enough. I do not have recent numbers for a SMD full sync baseline, but the setup documentation mentions 20 minutes for a full sync. That was a few years ago, and the spool has obviously grown since then, so that is not a reliable baseline. A baseline for a full sync might be also set with rsync, which copies files at nearly 40MB/s, or 317Mb/s!
anarcat@angela:tmp(main)$ time rsync -a --info=progress2 --exclude xapian  shell.anarc.at:Maildir/ Maildir/
 12,647,814,731 100%   37.85MB/s    0:05:18 (xfr#394981, to-chk=0/395815)    
72.38user 106.10system 5:19.59elapsed 55%CPU (0avgtext+0avgdata 15988maxresident)k
8816inputs+26305112outputs (0major+50953minor)pagefaults 0swaps
That is 5 minutes to transfer the entire spool. Incremental syncs are obviously pretty fast too:
anarcat@angela:tmp(main)$ time rsync -a --info=progress2 --exclude xapian  shell.anarc.at:Maildir/ Maildir/
              0   0%    0.00kB/s    0:00:00 (xfr#0, to-chk=0/395815)    
1.42user 0.81system 0:03.31elapsed 67%CPU (0avgtext+0avgdata 14100maxresident)k
120inputs+0outputs (3major+12709minor)pagefaults 0swaps
As an extra curiosity, here's the performance with tar, pretty similar with rsync, minus incremental which I cannot be bothered to figure out right now:
anarcat@angela:tmp(main)$ time ssh shell.anarc.at tar --exclude xapian -cf - Maildir/   pv -s 13G   tar xf - 
56.68user 58.86system 5:17.08elapsed 36%CPU (0avgtext+0avgdata 8764maxresident)k
0inputs+0outputs (0major+7266minor)pagefaults 0swaps
12,1GiO 0:05:17 [39,0MiB/s] [===================================================================> ] 92%
Interesting that rsync manages to almost beat a plain tar on file transfer, I'm actually surprised by how well it performs here, considering there are many little files to transfer. (But then again, this maybe is exactly where rsync shines: while tar needs to glue all those little files together, rsync can just directly talk to the other side and tell it to do live changes. Something to look at in another article maybe?) Since both ends are NVMe drives, those should easily saturate a gigabit link. And in fact, a backup of the server mail spool achieves much faster transfer rate on disks:
anarcat@marcos:~$ tar fc - Maildir   pv -s 13G > Maildir.tar
15,0GiO 0:01:57 [ 131MiB/s] [===================================] 115%
That's 131Mibyyte per second, vastly faster than the gigabit link. The client has similar performance:
anarcat@angela:~(main)$ tar fc - Maildir   pv -s 17G > Maildir.tar
16,2GiO 0:02:22 [ 116MiB/s] [==================================] 95%
So those disks should be able to saturate a gigabit link, and they are not the bottleneck on fast links. Which begs the question of what is blocking performance of a similar transfer over the gigabit link, but that's another question altogether, because no sync program ever reaches the above performance anyways. Finally, note that when I migrated to SMD, I wrote a small performance comparison that could be interesting here. It show SMD to be faster than OfflineIMAP, but not as much as we see here. In fact, it looks like OfflineIMAP slowed down significantly since then (May 2018), but this could be due to my larger mail spool as well.

mbsync The isync (AKA mbsync) project is written in C and supports syncing Maildir and IMAP folders, with possibly multiple replicas. I haven't tested this but I suspect it might be possible to sync between two IMAP servers as well. It supports partial mirorrs, message flags, full folder support, and "trash" functionality.

Complex configuration file I started with this .mbsyncrc configuration file:
SyncState *
Sync New ReNew Flags
IMAPAccount anarcat
Host imap.anarc.at
User anarcat
PassCmd "pass imap.anarc.at"
SSLType IMAPS
CertificateFile /etc/ssl/certs/ca-certificates.crt
IMAPStore anarcat-remote
Account anarcat
MaildirStore anarcat-local
# Maildir/top/sub/sub
#SubFolders Verbatim
# Maildir/.top.sub.sub
SubFolders Maildir++
# Maildir/top/.sub/.sub
# SubFolders legacy
# The trailing "/" is important
#Path ~/Maildir-mbsync/
Inbox ~/Maildir-mbsync/
Channel anarcat
# AKA Far, convert when all clients are 1.4+
Master :anarcat-remote:
# AKA Near
Slave :anarcat-local:
# Exclude everything under the internal [Gmail] folder, except the interesting folders
#Patterns * ![Gmail]* "[Gmail]/Sent Mail" "[Gmail]/Starred" "[Gmail]/All Mail"
# Or include everything
Patterns *
# Automatically create missing mailboxes, both locally and on the server
#Create Both
Create slave
# Sync the movement of messages between folders and deletions, add after making sure the sync works
#Expunge Both
Long gone are the days where I would spend a long time reading a manual page to figure out the meaning of every option. If that's your thing, you might like this one. But I'm more of a "EXAMPLES section" kind of person now, and I somehow couldn't find a sample file on the website. I started from the Arch wiki one but it's actually not great because it's made for Gmail (which is not a usual Dovecot server). So a sample config file in the manpage would be a great addition. Thankfully, the Debian packages ships one in /usr/share/doc/isync/examples/mbsyncrc.sample but I only found that after I wrote my configuration. It was still useful and I recommend people take a look if they want to understand the syntax. Also, that syntax is a little overly complicated. For example, Far needs colons, like:
Far :anarcat-remote:
Why? That seems just too complicated. I also found that sections are not clearly identified: IMAPAccount and Channel mark section beginnings, for example, which is not at all obvious until you learn about mbsync's internals. There are also weird ordering issues: the SyncState option needs to be before IMAPAccount, presumably because it's global. Using a more standard format like .INI or TOML could improve that situation.

Stellar performance A transfer of the entire mail spool takes 56 minutes and 6 seconds, which is impressive. It's not quite "line rate": the resulting mail spool was 12GB (which is a problem, see below), which turns out to be about 29Mbit/s and therefore not maxing the gigabit link, and an order of magnitude slower than rsync. The incremental runs are roughly 2 seconds, which is even more impressive, as that's actually faster than rsync:
===> multitime results
1: mbsync -a
            Mean        Std.Dev.    Min         Median      Max
real        2.015       0.052       1.930       2.029       2.105       
user        0.660       0.040       0.592       0.661       0.722       
sys         0.338       0.033       0.268       0.341       0.387    
Those tests were performed with isync 1.3.0-2.2 on Debian bullseye. Tests with a newer isync release originally failed because of a corrupted message that triggered bug 999804 (see below). Running 1.4.3 under valgrind works around the bug, but adds a 50% performance cost, the full sync running in 1h35m. Once the upstream patch is applied, performance with 1.4.3 is fairly similar, considering that the new sync included the register folder with 4000 messages:
120.74user 213.19system 59:47.69elapsed 9%CPU (0avgtext+0avgdata 105420maxresident)k
29128inputs+28284376outputs (0major+45711minor)pagefaults 0swaps
That is ~13GB in ~60 minutes, which gives us 28.3Mbps. Incrementals are also pretty similar to 1.3.x, again considering the double-connect cost:
===> multitime results
1: mbsync -a
            Mean        Std.Dev.    Min         Median      Max
real        2.500       0.087       2.340       2.491       2.629       
user        0.718       0.037       0.679       0.711       0.793       
sys         0.322       0.024       0.284       0.320       0.365
Those tests were all done on a Gigabit link, but what happens on a slower link? My server uplink is slow: 25 Mbps down, 6 Mbps up. There mbsync is worse than the SMD baseline:
===> multitime results
1: mbsync -a
Mean        Std.Dev.    Min         Median      Max
real        31.531      0.724       30.764      31.271      33.100      
user        1.858       0.125       1.721       1.818       2.131       
sys         0.610       0.063       0.506       0.600       0.695       
That's 30 seconds for a sync, which is an order of magnitude slower than SMD.

Great user interface Compared to OfflineIMAP and (ahem) SMD, the mbsync UI is kind of neat:
anarcat@angela:~(main)$ mbsync -a
Notice: Master/Slave are deprecated; use Far/Near instead.
C: 1/2  B: 204/205  F: +0/0 *0/0 #0/0  N: +1/200 *0/0 #0/0
(Note that nice switch away from slavery-related terms too.) The display is minimal, and yet informative. It's not obvious what does mean at first glance, but the manpage is useful at least for clarifying that:
This represents the cumulative progress over channels, boxes, and messages affected on the far and near side, respectively. The message counts represent added messages, messages with updated flags, and trashed messages, respectively. No attempt is made to calculate the totals in advance, so they grow over time as more information is gathered. (Emphasis mine).
In other words:
  • C 2/2: channels done/total (2 done out of 2)
  • B 204/205: mailboxes done/total (204 out of 205)
  • F: changes on the far side
  • N: +10/200 *0/0 #0/0: changes on the "near" side:
    • +10/200: 10 out of 200 messages downloaded
    • *0/0: no flag changed
    • #0/0: no message deleted
You get used to it, in a good way. It does not, unfortunately, show up when you run it in systemd, which is a bit annoying as I like to see a summary mail traffic in the logs.

Interoperability issue In my notmuch setup, I have bound key S to "mark spam", which basically assigns the tag spam to the message and removes a bunch of others. Then I have a notmuch-purge script which moves that message to the spam folder, for training purposes. It basically does this:
notmuch search --output=files --format=text0 "$search_spam" \
      xargs -r -0 mv -t "$HOME/Maildir/$ PREFIX junk/cur/"
This method, which worked fine in SMD (and also OfflineIMAP) created this error on sync:
Maildir error: duplicate UID 37578.
And indeed, there are now two messages with that UID in the mailbox:
anarcat@angela:~(main)$ find Maildir/.junk/ -name '*U=37578*'
Maildir/.junk/cur/1637427889.134334_2.angela,U=37578:2,S
Maildir/.junk/cur/1637348602.2492889_221804.angela,U=37578:2,S
This is actually a known limitation or, as mbsync(1) calls it, a "RECOMMENDATION":
When using the more efficient default UID mapping scheme, it is important that the MUA renames files when moving them between Maildir fold ers. Mutt always does that, while mu4e needs to be configured to do it:
(setq mu4e-change-filenames-when-moving t)
So it seems I would need to fix my script. It's unclear how the paths should be renamed, which is unfortunate, because I would need to change my script to adapt to mbsync, but I can't tell how just from reading the above. (A manual fix is actually to rename the file to remove the U= field: mbsync will generate a new one and then sync correctly.) Fortunately, someone else already fixed that issue: afew, a notmuch tagging script (much puns, such hurt), has a move mode that can rename files correctly, specifically designed to deal with mbsync. I had already been told about afew, but it's one more reason to standardize my notmuch hooks on that project, it looks like. Update: I have tried to use afew and found it has significant performance issues. It also has a completely different paradigm to what I am used to: it assumes all incoming mail has a new and lays its own tags on top of that (inbox, sent, etc). It can only move files from one folder at a time (see this bug) which breaks my spam training workflow. In general, I sync my tags into folders (e.g. ham, spam, sent) and message flags (e.g. inbox is F, unread is "not S", etc), and afew is not well suited for this (although there are hacks that try to fix this). I have worked hard to make my tagging scripts idempotent, and it's something afew doesn't currently have. Still, it would be better to have that code in Python than bash, so maybe I should consider my options here.

Stability issues The newer release in Debian bookworm (currently at 1.4.3) has stability issues on full sync. I filed bug 999804 in Debian about this, which lead to a thread on the upstream mailing list. I have found at least three distinct crashes that could be double-free bugs "which might be exploitable in the worst case", not a reassuring prospect. The thing is: mbsync is really fast, but the downside of that is that it's written in C, and with that comes a whole set of security issues. The Debian security tracker has only three CVEs on isync, but the above issues show there could be many more. Reading the source code certainly did not make me very comfortable with trusting it with untrusted data. I considered sandboxing it with systemd (below) but having systemd run as a --user process makes that difficult. I also considered using an apparmor profile but that is not trivial because we need to allow SSH and only some parts of it... Thankfully, upstream has been diligent at addressing the issues I have found. They provided a patch within a few days which did fix the sync issues. Update: upstream actually took the issue very seriously. They not only got CVE-2021-44143 assigned for my bug report, they also audited the code and found several more issues collectively identified as CVE-2021-3657, which actually also affect 1.3 (ie. Debian 11/bullseye/stable). Somehow my corpus doesn't trigger that issue, but it was still considered serious enough to warrant a CVE. So one the one hand: excellent response from upstream; but on the other hand: how many more of those could there be in there?

Automation with systemd The Arch wiki has instructions on how to setup mbsync as a systemd service. It suggests using the --verbose (-V) flag which is a little intense here, as it outputs 1444 lines of messages. I have used the following .service file:
[Unit]
Description=Mailbox synchronization service
ConditionHost=!marcos
Wants=network-online.target
After=network-online.target
Before=notmuch-new.service
[Service]
Type=oneshot
ExecStart=/usr/bin/mbsync -a
Nice=10
IOSchedulingClass=idle
NoNewPrivileges=true
[Install]
WantedBy=default.target
And the following .timer:
[Unit]
Description=Mailbox synchronization timer
ConditionHost=!marcos
[Timer]
OnBootSec=2m
OnUnitActiveSec=5m
Unit=mbsync.service
[Install]
WantedBy=timers.target
Note that we trigger notmuch through systemd, with the Before and also by adding mbsync.service to the notmuch-new.service file:
[Unit]
Description=notmuch new
After=mbsync.service
[Service]
Type=oneshot
Nice=10
ExecStart=/usr/bin/notmuch new
[Install]
WantedBy=mbsync.service
An improvement over polling repeatedly with a .timer would be to wake up only on IMAP notify, but neither imapnotify nor goimapnotify seem to be packaged in Debian. It would also not cover for the "sent folder" use case, where we need to wake up on local changes.

Password-less setup The sample file suggests this should work:
IMAPStore remote
Tunnel "ssh -q host.remote.com /usr/sbin/imapd"
Add BatchMode, restrict to IdentitiesOnly, provide a password-less key just for this, add compression (-C), find the Dovecot imap binary, and you get this:
IMAPAccount anarcat-tunnel
Tunnel "ssh -o BatchMode=yes -o IdentitiesOnly=yes -i ~/.ssh/id_ed25519_mbsync -o HostKeyAlias=shell.anarc.at -C anarcat@imap.anarc.at /usr/lib/dovecot/imap"
And it actually seems to work:
$ mbsync -a
Notice: Master/Slave are deprecated; use Far/Near instead.
C: 0/2  B: 0/1  F: +0/0 *0/0 #0/0  N: +0/0 *0/0 #0/0imap(anarcat): Error: net_connect_unix(/run/dovecot/stats-writer) failed: Permission denied
C: 2/2  B: 205/205  F: +0/0 *0/0 #0/0  N: +1/1 *3/3 #0/0imap(anarcat)<1611280><90uUOuyElmEQlhgAFjQyWQ>: Info: Logged out in=10808 out=15396642 deleted=0 expunged=0 trashed=0 hdr_count=0 hdr_bytes=0 body_count=1 body_bytes=8087
It's a bit noisy, however. dovecot/imap doesn't have a "usage" to speak of, but even the source code doesn't hint at a way to disable that Error message, so that's unfortunate. That socket is owned by root:dovecot so presumably Dovecot runs the imap process as $user:dovecot, which we can't do here. Oh well? Interestingly, the SSH setup is not faster than IMAP. With IMAP:
===> multitime results
1: mbsync -a
            Mean        Std.Dev.    Min         Median      Max
real        2.367       0.065       2.220       2.376       2.458       
user        0.793       0.047       0.731       0.776       0.871       
sys         0.426       0.040       0.364       0.434       0.476
With SSH:
===> multitime results
1: mbsync -a
            Mean        Std.Dev.    Min         Median      Max
real        2.515       0.088       2.274       2.532       2.594       
user        0.753       0.043       0.645       0.766       0.804       
sys         0.328       0.045       0.212       0.340       0.393
Basically: 200ms slower. Tolerable.

Migrating from SMD The above was how I migrated to mbsync on my first workstation. The work on the second one was more streamlined, especially since the corruption on mailboxes was fixed:
  1. install isync, with the patch:
    dpkg -i isync_1.4.3-1.1~_amd64.deb
    
  2. copy all files over from previous workstation to avoid a full resync (optional):
    rsync -a --info=progress2 angela:Maildir/ Maildir-mbsync/
    
  3. rename all files to match new hostname (optional):
    find Maildir-mbsync/ -type f -name '*.angela,*' -print0    rename -0 's/\.angela,/\.curie,/'
    
  4. trash the notmuch database (optional):
    rm -rf Maildir-mbsync/.notmuch/xapian/
    
  5. disable all smd and notmuch services:
    systemctl --user --now disable smd-pull.service smd-pull.timer smd-push.service smd-push.timer notmuch-new.service notmuch-new.timer
    
  6. do one last sync with smd:
    smd-pull --show-tags ; smd-push --show-tags ; notmuch new ; notmuch-sync-flagged -v
    
  7. backup notmuch on the client and server:
    notmuch dump   pv > notmuch.dump
    
  8. backup the maildir on the client and server:
    cp -al Maildir Maildir-bak
    
  9. create the SSH key:
    ssh-keygen -t ed25519 -f .ssh/id_ed25519_mbsync
    cat .ssh/id_ed25519_mbsync.pub
    
  10. add to .ssh/authorized_keys on the server, like this: command="/usr/lib/dovecot/imap",restrict ssh-ed25519 AAAAC...
  11. move old files aside, if present:
    mv Maildir Maildir-smd
    
  12. move new files in place (CRITICAL SECTION BEGINS!):
    mv Maildir-mbsync Maildir
    
  13. run a test sync, only pulling changes: mbsync --create-near --remove-none --expunge-none --noop anarcat-register
  14. if that works well, try with all mailboxes: mbsync --create-near --remove-none --expunge-none --noop -a
  15. if that works well, try again with a full sync: mbsync register mbsync -a
  16. reindex and restore the notmuch database, this should take ~25 minutes:
    notmuch new
    pv notmuch.dump   notmuch restore
    
  17. enable the systemd services and retire the smd-* services: systemctl --user enable mbsync.timer notmuch-new.service systemctl --user start mbsync.timer rm ~/.config/systemd/user/smd* systemctl daemon-reload
During the migration, notmuch helpfully told me the full list of those lost messages:
[...]
Warning: cannot apply tags to missing message: CAN6gO7_QgCaiDFvpG3AXHi6fW12qaN286+2a7ERQ2CQtzjSEPw@mail.gmail.com
Warning: cannot apply tags to missing message: CAPTU9Wmp0yAmaxO+qo8CegzRQZhCP853TWQ_Ne-YF94MDUZ+Dw@mail.gmail.com
Warning: cannot apply tags to missing message: F5086003-2917-4659-B7D2-66C62FCD4128@gmail.com
[...]
Warning: cannot apply tags to missing message: mailman.2.1316793601.53477.sage-members@mailman.sage.org
Warning: cannot apply tags to missing message: mailman.7.1317646801.26891.outages-discussion@outages.org
Warning: cannot apply tags to missing message: notmuch-sha1-000458df6e48d4857187a000d643ac971deeef47
Warning: cannot apply tags to missing message: notmuch-sha1-0079d8e0c3340e6f88c66f4c49fca758ea71d06d
Warning: cannot apply tags to missing message: notmuch-sha1-0194baa4cfb6d39bc9e4d8c049adaccaa777467d
Warning: cannot apply tags to missing message: notmuch-sha1-02aede494fc3f9e9f060cfd7c044d6d724ad287c
Warning: cannot apply tags to missing message: notmuch-sha1-06606c625d3b3445420e737afd9a245ae66e5562
Warning: cannot apply tags to missing message: notmuch-sha1-0747b020f7551415b9bf5059c58e0a637ba53b13
[...]
As detailed in the crash report, all of those were actually innocuous and could be ignored. Also note that we completely trash the notmuch database because it's actually faster to reindex from scratch than let notmuch slowly figure out that all mails are new and all the old mails are gone. The fresh indexing took:
nov 19 15:08:54 angela notmuch[2521117]: Processed 384679 total files in 23m 41s (270 files/sec.).
nov 19 15:08:54 angela notmuch[2521117]: Added 372610 new messages to the database.
While a reindexing on top of an existing database was going twice as slow, at about 120 files/sec.

Current config file Putting it all together, I ended up with the following configuration file:
SyncState *
Sync All
# IMAP side, AKA "Far"
IMAPAccount anarcat-imap
Host imap.anarc.at
User anarcat
PassCmd "pass imap.anarc.at"
SSLType IMAPS
CertificateFile /etc/ssl/certs/ca-certificates.crt
IMAPAccount anarcat-tunnel
Tunnel "ssh -o BatchMode=yes -o IdentitiesOnly=yes -i ~/.ssh/id_ed25519_mbsync -o HostKeyAlias=shell.anarc.at -C anarcat@imap.anarc.at /usr/lib/dovecot/imap"
IMAPStore anarcat-remote
Account anarcat-tunnel
# Maildir side, AKA "Near"
MaildirStore anarcat-local
# Maildir/top/sub/sub
#SubFolders Verbatim
# Maildir/.top.sub.sub
SubFolders Maildir++
# Maildir/top/.sub/.sub
# SubFolders legacy
# The trailing "/" is important
#Path ~/Maildir-mbsync/
Inbox ~/Maildir/
# what binds Maildir and IMAP
Channel anarcat
Far :anarcat-remote:
Near :anarcat-local:
# Exclude everything under the internal [Gmail] folder, except the interesting folders
#Patterns * ![Gmail]* "[Gmail]/Sent Mail" "[Gmail]/Starred" "[Gmail]/All Mail"
# Or include everything
#Patterns *
Patterns * !register  !.register
# Automatically create missing mailboxes, both locally and on the server
Create Both
#Create Near
# Sync the movement of messages between folders and deletions, add after making sure the sync works
Expunge Both
# Propagate mailbox deletion
Remove both
IMAPAccount anarcat-register-imap
Host imap.anarc.at
User register
PassCmd "pass imap.anarc.at-register"
SSLType IMAPS
CertificateFile /etc/ssl/certs/ca-certificates.crt
IMAPAccount anarcat-register-tunnel
Tunnel "ssh -o BatchMode=yes -o IdentitiesOnly=yes -i ~/.ssh/id_ed25519_mbsync -o HostKeyAlias=shell.anarc.at -C register@imap.anarc.at /usr/lib/dovecot/imap"
IMAPStore anarcat-register-remote
Account anarcat-register-tunnel
MaildirStore anarcat-register-local
SubFolders Maildir++
Inbox ~/Maildir/.register/
Channel anarcat-register
Far :anarcat-register-remote:
Near :anarcat-register-local:
Create Both
Expunge Both
Remove both
Note that it may be out of sync with my live (and private) configuration file, as I do not publish my "dotfiles" repository publicly for security reasons.

OfflineIMAP I've used OfflineIMAP for a long time before switching to SMD. I don't exactly remember why or when I started using it, but I do remember it became painfully slow as I started using notmuch, and would sometimes crash mysteriously. It's been a while, so my memory is hazy on that. It also kind of died in a fire when Python 2 stop being maintained. The main author moved on to a different project, imapfw which could serve as a framework to build IMAP clients, but never seemed to implement all of the OfflineIMAP features and certainly not configuration file compatibility. Thankfully, a new team of volunteers ported OfflineIMAP to Python 3 and we can now test that new version to see if it is an improvement over mbsync.

Crash on full sync The first thing that happened on a full sync is this crash:
Copy message from RemoteAnarcat:junk:
 ERROR: Copying message 30624 [acc: Anarcat]
  decoding with 'X-EUC-TW' codec failed (AttributeError: 'memoryview' object has no attribute 'decode')
Thread 'Copy message from RemoteAnarcat:junk' terminated with exception:
Traceback (most recent call last):
  File "/usr/share/offlineimap3/offlineimap/imaputil.py", line 406, in utf7m_decode
    for c in binary.decode():
AttributeError: 'memoryview' object has no attribute 'decode'
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
  File "/usr/share/offlineimap3/offlineimap/threadutil.py", line 146, in run
    Thread.run(self)
  File "/usr/lib/python3.9/threading.py", line 892, in run
    self._target(*self._args, **self._kwargs)
  File "/usr/share/offlineimap3/offlineimap/folder/Base.py", line 802, in copymessageto
    message = self.getmessage(uid)
  File "/usr/share/offlineimap3/offlineimap/folder/IMAP.py", line 342, in getmessage
    data = self._fetch_from_imap(str(uid), self.retrycount)
  File "/usr/share/offlineimap3/offlineimap/folder/IMAP.py", line 908, in _fetch_from_imap
    ndata1 = self.parser['8bit-RFC'].parsebytes(data[0][1])
  File "/usr/lib/python3.9/email/parser.py", line 123, in parsebytes
    return self.parser.parsestr(text, headersonly)
  File "/usr/lib/python3.9/email/parser.py", line 67, in parsestr
    return self.parse(StringIO(text), headersonly=headersonly)
  File "/usr/lib/python3.9/email/parser.py", line 56, in parse
    feedparser.feed(data)
  File "/usr/lib/python3.9/email/feedparser.py", line 176, in feed
    self._call_parse()
  File "/usr/lib/python3.9/email/feedparser.py", line 180, in _call_parse
    self._parse()
  File "/usr/lib/python3.9/email/feedparser.py", line 385, in _parsegen
    for retval in self._parsegen():
  File "/usr/lib/python3.9/email/feedparser.py", line 298, in _parsegen
    for retval in self._parsegen():
  File "/usr/lib/python3.9/email/feedparser.py", line 385, in _parsegen
    for retval in self._parsegen():
  File "/usr/lib/python3.9/email/feedparser.py", line 256, in _parsegen
    if self._cur.get_content_type() == 'message/delivery-status':
  File "/usr/lib/python3.9/email/message.py", line 578, in get_content_type
    value = self.get('content-type', missing)
  File "/usr/lib/python3.9/email/message.py", line 471, in get
    return self.policy.header_fetch_parse(k, v)
  File "/usr/lib/python3.9/email/policy.py", line 163, in header_fetch_parse
    return self.header_factory(name, value)
  File "/usr/lib/python3.9/email/headerregistry.py", line 601, in __call__
    return self[name](name, value)
  File "/usr/lib/python3.9/email/headerregistry.py", line 196, in __new__
    cls.parse(value, kwds)
  File "/usr/lib/python3.9/email/headerregistry.py", line 445, in parse
    kwds['parse_tree'] = parse_tree = cls.value_parser(value)
  File "/usr/lib/python3.9/email/_header_value_parser.py", line 2675, in parse_content_type_header
    ctype.append(parse_mime_parameters(value[1:]))
  File "/usr/lib/python3.9/email/_header_value_parser.py", line 2569, in parse_mime_parameters
    token, value = get_parameter(value)
  File "/usr/lib/python3.9/email/_header_value_parser.py", line 2492, in get_parameter
    token, value = get_value(value)
  File "/usr/lib/python3.9/email/_header_value_parser.py", line 2403, in get_value
    token, value = get_quoted_string(value)
  File "/usr/lib/python3.9/email/_header_value_parser.py", line 1294, in get_quoted_string
    token, value = get_bare_quoted_string(value)
  File "/usr/lib/python3.9/email/_header_value_parser.py", line 1223, in get_bare_quoted_string
    token, value = get_encoded_word(value)
  File "/usr/lib/python3.9/email/_header_value_parser.py", line 1064, in get_encoded_word
    text, charset, lang, defects = _ew.decode('=?' + tok + '?=')
  File "/usr/lib/python3.9/email/_encoded_words.py", line 181, in decode
    string = bstring.decode(charset)
AttributeError: decoding with 'X-EUC-TW' codec failed (AttributeError: 'memoryview' object has no attribute 'decode')
Last 1 debug messages logged for Copy message from RemoteAnarcat:junk prior to exception:
thread: Register new thread 'Copy message from RemoteAnarcat:junk' (account 'Anarcat')
ERROR: Exceptions occurred during the run!
ERROR: Copying message 30624 [acc: Anarcat]
  decoding with 'X-EUC-TW' codec failed (AttributeError: 'memoryview' object has no attribute 'decode')
Traceback:
  File "/usr/share/offlineimap3/offlineimap/folder/Base.py", line 802, in copymessageto
    message = self.getmessage(uid)
  File "/usr/share/offlineimap3/offlineimap/folder/IMAP.py", line 342, in getmessage
    data = self._fetch_from_imap(str(uid), self.retrycount)
  File "/usr/share/offlineimap3/offlineimap/folder/IMAP.py", line 908, in _fetch_from_imap
    ndata1 = self.parser['8bit-RFC'].parsebytes(data[0][1])
  File "/usr/lib/python3.9/email/parser.py", line 123, in parsebytes
    return self.parser.parsestr(text, headersonly)
  File "/usr/lib/python3.9/email/parser.py", line 67, in parsestr
    return self.parse(StringIO(text), headersonly=headersonly)
  File "/usr/lib/python3.9/email/parser.py", line 56, in parse
    feedparser.feed(data)
  File "/usr/lib/python3.9/email/feedparser.py", line 176, in feed
    self._call_parse()
  File "/usr/lib/python3.9/email/feedparser.py", line 180, in _call_parse
    self._parse()
  File "/usr/lib/python3.9/email/feedparser.py", line 385, in _parsegen
    for retval in self._parsegen():
  File "/usr/lib/python3.9/email/feedparser.py", line 298, in _parsegen
    for retval in self._parsegen():
  File "/usr/lib/python3.9/email/feedparser.py", line 385, in _parsegen
    for retval in self._parsegen():
  File "/usr/lib/python3.9/email/feedparser.py", line 256, in _parsegen
    if self._cur.get_content_type() == 'message/delivery-status':
  File "/usr/lib/python3.9/email/message.py", line 578, in get_content_type
    value = self.get('content-type', missing)
  File "/usr/lib/python3.9/email/message.py", line 471, in get
    return self.policy.header_fetch_parse(k, v)
  File "/usr/lib/python3.9/email/policy.py", line 163, in header_fetch_parse
    return self.header_factory(name, value)
  File "/usr/lib/python3.9/email/headerregistry.py", line 601, in __call__
    return self[name](name, value)
  File "/usr/lib/python3.9/email/headerregistry.py", line 196, in __new__
    cls.parse(value, kwds)
  File "/usr/lib/python3.9/email/headerregistry.py", line 445, in parse
    kwds['parse_tree'] = parse_tree = cls.value_parser(value)
  File "/usr/lib/python3.9/email/_header_value_parser.py", line 2675, in parse_content_type_header
    ctype.append(parse_mime_parameters(value[1:]))
  File "/usr/lib/python3.9/email/_header_value_parser.py", line 2569, in parse_mime_parameters
    token, value = get_parameter(value)
  File "/usr/lib/python3.9/email/_header_value_parser.py", line 2492, in get_parameter
    token, value = get_value(value)
  File "/usr/lib/python3.9/email/_header_value_parser.py", line 2403, in get_value
    token, value = get_quoted_string(value)
  File "/usr/lib/python3.9/email/_header_value_parser.py", line 1294, in get_quoted_string
    token, value = get_bare_quoted_string(value)
  File "/usr/lib/python3.9/email/_header_value_parser.py", line 1223, in get_bare_quoted_string
    token, value = get_encoded_word(value)
  File "/usr/lib/python3.9/email/_header_value_parser.py", line 1064, in get_encoded_word
    text, charset, lang, defects = _ew.decode('=?' + tok + '?=')
  File "/usr/lib/python3.9/email/_encoded_words.py", line 181, in decode
    string = bstring.decode(charset)
Folder junk [acc: Anarcat]:
 Copy message UID 30626 (29008/49310) RemoteAnarcat:junk -> LocalAnarcat:junk
Command exited with non-zero status 100
5252.91user 535.86system 3:21:00elapsed 47%CPU (0avgtext+0avgdata 846304maxresident)k
96344inputs+26563792outputs (1189major+2155815minor)pagefaults 0swaps
That only transferred about 8GB of mail, which gives us a transfer rate of 5.3Mbit/s, more than 5 times slower than mbsync. This bug is possibly limited to the bullseye version of offlineimap3 (the lovely 0.0~git20210225.1e7ef9e+dfsg-4), while the current sid version (the equally gorgeous 0.0~git20211018.e64c254+dfsg-1) seems unaffected.

Tolerable performance The new release still crashes, except it does so at the very end, which is an improvement, since the mails do get transferred:
 *** Finished account 'Anarcat' in 511:12
ERROR: Exceptions occurred during the run!
ERROR: Exception parsing message with ID (<20190619152034.BFB8810E07A@marcos.anarc.at>) from imaplib (response type: bytes).
 AttributeError: decoding with 'X-EUC-TW' codec failed (AttributeError: 'memoryview' object has no attribute 'decode')
Traceback:
  File "/usr/share/offlineimap3/offlineimap/folder/Base.py", line 810, in copymessageto
    message = self.getmessage(uid)
  File "/usr/share/offlineimap3/offlineimap/folder/IMAP.py", line 343, in getmessage
    data = self._fetch_from_imap(str(uid), self.retrycount)
  File "/usr/share/offlineimap3/offlineimap/folder/IMAP.py", line 910, in _fetch_from_imap
    raise OfflineImapError(
ERROR: Exception parsing message with ID (<40A270DB.9090609@alternatives.ca>) from imaplib (response type: bytes).
 AttributeError: decoding with 'x-mac-roman' codec failed (AttributeError: 'memoryview' object has no attribute 'decode')
Traceback:
  File "/usr/share/offlineimap3/offlineimap/folder/Base.py", line 810, in copymessageto
    message = self.getmessage(uid)
  File "/usr/share/offlineimap3/offlineimap/folder/IMAP.py", line 343, in getmessage
    data = self._fetch_from_imap(str(uid), self.retrycount)
  File "/usr/share/offlineimap3/offlineimap/folder/IMAP.py", line 910, in _fetch_from_imap
    raise OfflineImapError(
ERROR: IMAP server 'RemoteAnarcat' does not have a message with UID '32686'
Traceback:
  File "/usr/share/offlineimap3/offlineimap/folder/Base.py", line 810, in copymessageto
    message = self.getmessage(uid)
  File "/usr/share/offlineimap3/offlineimap/folder/IMAP.py", line 343, in getmessage
    data = self._fetch_from_imap(str(uid), self.retrycount)
  File "/usr/share/offlineimap3/offlineimap/folder/IMAP.py", line 889, in _fetch_from_imap
    raise OfflineImapError(reason, severity)
Command exited with non-zero status 1
8273.52user 983.80system 8:31:12elapsed 30%CPU (0avgtext+0avgdata 841936maxresident)k
56376inputs+43247608outputs (811major+4972914minor)pagefaults 0swaps
"offlineimap  -o " took 8 hours 31 mins 15 secs
This is 8h31m for transferring 12G, which is around 3.1Mbit/s. That is nine times slower than mbsync, almost an order of magnitude! Now that we have a full sync, we can test incremental synchronization. That is also much slower:
===> multitime results
1: sh -c "offlineimap -o   true"
            Mean        Std.Dev.    Min         Median      Max
real        24.639      0.513       23.946      24.526      25.708      
user        23.912      0.473       23.404      23.795      24.947      
sys         1.743       0.105       1.607       1.729       2.002
That is also an order of magnitude slower than mbsync, and significantly slower than what you'd expect from a sync process. ~30 seconds is long enough to make me impatient and distracted; 3 seconds, less so: I can wait and see the results almost immediately.

Integrity check That said: this is still on a gigabit link. It's technically possible that OfflineIMAP performs better than mbsync over a slow link, but I Haven't tested that theory. The OfflineIMAP mail spool is missing quite a few messages as well:
anarcat@angela:~(main)$ find Maildir-offlineimap -type f -type f -a \! -name '.*'   wc -l 
381463
anarcat@angela:~(main)$ find Maildir -type f -type f -a \! -name '.*'   wc -l 
385247
... although that's probably all either new messages or the register folder, so OfflineIMAP might actually be in a better position there. But digging in more, it seems like the actual per-folder diff is fairly similar to mbsync: a few messages missing here and there. Considering OfflineIMAP's instability and poor performance, I have not looked any deeper in those discrepancies.

Other projects to evaluate Those are all the options I have considered, in alphabetical order
  • doveadm-sync: requires dovecot on both ends, can tunnel over SSH, may have performance issues in incremental sync, written in C
  • fdm: fetchmail replacement, IMAP/POP3/stdin/Maildir/mbox,NNTP support, SOCKS support (for Tor), complex rules for delivering to specific mailboxes, adding headers, piping to commands, etc. discarded because no (real) support for keeping mail on the server, and written in C
  • getmail: fetchmail replacement, IMAP/POP3 support, supports incremental runs, classification rules, Python
  • interimap: syncs two IMAP servers, apparently faster than doveadm and offlineimap, but requires running an IMAP server locally, Perl
  • isync/mbsync: TLS client certs and SSH tunnels, fast, incremental, IMAP/POP/Maildir support, multiple mailbox, trash and recursion support, and generally has good words from multiple Debian and notmuch people (Arch tutorial), written in C, review above
  • mail-sync: notify support, happens over any piped transport (e.g. ssh), diff/patch system, requires binary on both ends, mentions UUCP in the manpage, mentions rsmtp which is a nice name for rsendmail. not evaluated because it seems awfully complex to setup, Haskell
  • nncp: treat the local spool as another mail server, not really compatible with my "multiple clients" setup, Golang
  • offlineimap3: requires IMAP, used the py2 version in the past, might just still work, first sync painful (IIRC), ways to tunnel over SSH, review above, Python
Most projects were not evaluated due to lack of time.

Conclusion I'm now using mbsync to sync my mail. I'm a little disappointed by the synchronisation times over the slow link, but I guess that's on par for the course if we use IMAP. We are bound by the network speed much more than with custom protocols. I'm also worried about the C implementation and the crashes I have witnessed, but I am encouraged by the fast upstream response. Time will tell if I will stick with that setup. I'm certainly curious about the promises of interimap and mail-sync, but I have ran out of time on this project.

17 November 2021

Christoph Berg: PostgreSQL and Undelete

pg_dirtyread Earlier this week, I updated pg_dirtyread to work with PostgreSQL 14. pg_dirtyread is a PostgreSQL extension that allows reading "dead" rows from tables, i.e. rows that have already been deleted, or updated. Of course that works only if the table has not been cleaned-up yet by a VACUUM command or autovacuum, which is PostgreSQL's garbage collection machinery. Here's an example of pg_dirtyread in action:
# create table foo (id int, t text);
CREATE TABLE
# insert into foo values (1, 'Doc1');
INSERT 0 1
# insert into foo values (2, 'Doc2');
INSERT 0 1
# insert into foo values (3, 'Doc3');
INSERT 0 1
# select * from foo;
 id    t
 
  1   Doc1
  2   Doc2
  3   Doc3
(3 rows)
# delete from foo where id < 3;
DELETE 2
# select * from foo;
 id    t
 
  3   Doc3
(1 row)
Oops! The first two documents have disappeared. Now let's use pg_dirtyread to look at the table:
# create extension pg_dirtyread;
CREATE EXTENSION
# select * from pg_dirtyread('foo') t(id int, t text);
 id    t
 
  1   Doc1
  2   Doc2
  3   Doc3
All three documents are still there, but only one of them is visible. pg_dirtyread can also show PostgreSQL's system colums with the row location and visibility information. For the first two documents, xmax is set, which means the row has been deleted:
# select * from pg_dirtyread('foo') t(ctid tid, xmin xid, xmax xid, id int, t text);
 ctid    xmin   xmax   id    t
 
 (0,1)   1577   1580    1   Doc1
 (0,2)   1578   1580    2   Doc2
 (0,3)   1579      0    3   Doc3
(3 rows)
Undelete Caveat: I'm not promising any of the ideas quoted below will actually work in practice. There are a few caveats and a good portion of intricate knowledge about the PostgreSQL internals might be required to succeed properly. Consider consulting your favorite PostgreSQL support channel for advice if you need to recover data on any production system. Don't try this at work. I always had plans to extend pg_dirtyread to include some "undelete" command to make deleted rows reappear, but never got around to trying that. But rows can already be restored by using the output of pg_dirtyread itself:
# insert into foo select * from pg_dirtyread('foo') t(id int, t text) where id = 1;
This is not a true "undelete", though - it just inserts new rows from the data read from the table. pg_surgery Enter pg_surgery, which is a new PostgreSQL extension supplied with PostgreSQL 14. It contains two functions to "perform surgery on a damaged relation". As a side-effect, they can also make delete tuples reappear. As I discovered now, one of the functions, heap_force_freeze(), works nicely with pg_dirtyread. It takes a list of ctids (row locations) that it marks "frozen", but at the same time as "not deleted". Let's apply it to our test table, using the ctids that pg_dirtyread can read:
# create extension pg_surgery;
CREATE EXTENSION
# select heap_force_freeze('foo', array_agg(ctid))
    from pg_dirtyread('foo') t(ctid tid, xmin xid, xmax xid, id int, t text) where id = 1;
 heap_force_freeze
 
(1 row)
Et voil , our deleted document is back:
# select * from foo;
 id    t
 
  1   Doc1
  3   Doc3
(2 rows)
# select * from pg_dirtyread('foo') t(ctid tid, xmin xid, xmax xid, id int, t text);
 ctid    xmin   xmax   id    t
 
 (0,1)      2      0    1   Doc1
 (0,2)   1578   1580    2   Doc2
 (0,3)   1579      0    3   Doc3
(3 rows)
Disclaimer Most importantly, none of the above methods will work if the data you just deleted has already been purged by VACUUM or autovacuum. These actively zero out reclaimed space. Restore from backup to get your data back. Since both pg_dirtyread and pg_surgery operate outside the normal PostgreSQL MVCC machinery, it's easy to create corrupt data using them. This includes duplicated rows, duplicated primary key values, indexes being out of sync with tables, broken foreign key constraints, and others. You have been warned. pg_dirtyread does not work (yet) if the deleted rows contain any toasted values. Possible other approaches include using pageinspect and pg_filedump to retrieve the ctids of deleted rows. Please make sure you have working backups and don't need any of the above.

16 November 2021

Paul Tagliamonte: Measuring the Power Output of my SDRs

Over the last few years, I ve often wondered what the true power output of my SDRs are. It s a question with a shocking amount of complexity in the response, due to a number of factors (mostly Frequency). The ranges given in spec sheets are often extremely vague, and if I m being honest with myself, not incredibly helpful for being able to determine what specific filters and amplifiers I ll need to get a clean signal transmitted.
Hey, heads up! - This post contains extremely unvalidated and back of the napkin quality work to understand how my equipment works. Hopefully this work can be of help to others, but please double check any information you need for your own work!
I was specifically interested in what gain output (in dBm) looks like across the frequency range in particular, how variable the output dBm is when I change frequencies. The second question I had was understanding how linear the output gain is when adjusting the requested gain from the radio. Does a 2 dB increase on a HackRF API mean 2 dB of gain in dBm, no matter what the absolute value of the gain stage is? I ve finally bit the bullet and undertaken work to characterize the hardware I do have, with some outdated laboratory equipment I found on eBay. Of course, if it s worth doing, it s worth overdoing, so I spent a bit of time automating a handful of components in order to collect the data that I need from my SDRs. I bought an HP 437B, which is the cutting edge of 30 years ago, but still accurate to within 0.01dBm. I paired this Power Meter with an Agilent 8481A Power Sensor (-30 dBm to 20 dBm from 10MHz to 18GHz). For some of my radios, I was worried about exceeding the 20 dBm mark, so I used a 20db attenuator while I waited for a higher power power sensor. Finally, I was able to find a GPIB to USB interface, and get that interface working with the GPIB Kernel driver on my system. With all that out of the way, I was able to write Go bindings to my HP 437B to allow for totally headless and automated control in sync with my SDR s RF output. This allowed me to script the transmission of a sine wave at a controlled amplitude across a defined gain range and frequency range and read the Power Sensor s measured dBm output to characterize the Gain across frequency and configured Gain.

HackRF Looking at configured Gain against output power, the requested gain appears to have a fairly linear relation to the output signal power. The measured dBm ranged between the sensor noise floor to approx +13dBm. The average standard deviation of all tested gain values over the frequency range swept was +/-2dBm, with a minimum standard deviation of +/-0.8dBm, and a maximum of +/-3dBm. When looking at output power over the frequency range swept, the HackRF contains a distinctive (and frankly jarring) ripple across the Frequency range, with a clearly visible jump in gain somewhere around 2.1GHz. I have no idea what is causing this massive jump in output gain, nor what is causing these distinctive ripples. I d love to know more if anyone s familiar with HackRF s RF internals!

PlutoSDR The power output is very linear when operating above -20dB TX channel gain, but can get quite erratic the lower the output power is configured. The PlutoSDR s output power is directly related to the configured power level, and is generally predictable once a minimum power level is reached. The measured dBm ranged from the noise floor to 3.39 dBm, with an average standard deviation of +/-1.98 dBm, a minimum standard deviation of +/-0.91 dBm and a maximum standard deviation of +/-3.37 dBm. Generally, the power output is quite stable, and looks to have very even and wideband gain control. There s a few artifacts, which I have not confidently isolated to the SDR TX gain, noise (transmit artifacts such as intermodulation) or to my test setup. They appear fairly narrowband, so I m not overly worried about them yet. If anyone has any ideas what this could be, I d very much appreciate understanding why they exist!

Ettus B210 The power output on the Ettus B210 is higher (in dBm) than any of my other radios, but it has a very odd quirk where the power becomes nonlinear somewhere around -55dB TX channel gain. After that point, adding gain has no effect on the measured signal output in dBm up to 0 dB gain. The measured dBm ranged from the noise floor to 18.31 dBm, with an average standard deviation of +/-2.60 dBm, a minimum of +/-1.39 dBm and a maximum of +/-5.82 dBm. When the Gain is somewhere around the noise floor, the measured gain is incredibly erratic, which throws the maximum standard deviation significantly. I haven t isolated that to my test setup or the radio itself. I m inclined to believe it s my test setup. The radio has a fairly even and wideband gain, and so long as you re operating between -70dB to -55dB, fairly linear as well.

Summary Of all my radios, the Ettus B210 has the highest output (in dBm) over the widest frequency range, but the HackRF is a close second, especially after the gain bump kicks in around 2.1GHz. The Pluto SDR feels the most predictable and consistent, but also a very low output, comparatively - right around 0 dBm.
Name Max dBm stdev dBm stdev min dBm stdev max dBm
HackRF +12.6 +/-2.0 +/-0.8 +/-3.0
PlutoSDR +3.3 +/-2.0 +/-0.9 +/-3.7
B210 +18.3 +/-2.6 +/-1.4 +/-6.0

15 November 2021

Vincent Bernat: Git as a source of truth for network automation

The first step when automating a network is to build the source of truth. A source of truth is a repository of data that provides the intended state: the list of devices, the IP addresses, the network protocols settings, the time servers, etc. A popular choice is NetBox. Its documentation highlights its usage as a source of truth:
NetBox intends to represent the desired state of a network versus its operational state. As such, automated import of live network state is strongly discouraged. All data created in NetBox should first be vetted by a human to ensure its integrity. NetBox can then be used to populate monitoring and provisioning systems with a high degree of confidence.
When introducing Jerikan, a common feedback we got was: you should use NetBox for this. Indeed, Jerikan s source of truth is a bunch of YAML files versioned with Git.

Why Git? If we look at how things are done with servers and services, in a datacenter or in the cloud, we are likely to find users of Terraform, a tool turning declarative configuration files into infrastructure. Declarative configuration management tools like Salt, Puppet,1 or Ansible take care of server configuration. NixOS is an alternative: it combines package management and configuration management with a functional language to build virtual machines and containers. When using a Kubernetes cluster, people use Kustomize or Helm, two other declarative configuration management tools. Tapped together, these tools implement the infrastructure as code paradigm.
Infrastructure as code is an approach to infrastructure automation based on practices from software development. It emphasizes consistent, repeatable routines for provisioning and changing systems and their configuration. You make changes to code, then use automation to test and apply those changes to your systems. Kief Morris, Infrastructure as Code, O Reilly.
A version control system is a central tool for infrastructure as code. The usual candidate is Git with a source code management system like GitLab or GitHub. You get:
Traceability and visibility
Git keeps a log of all changes: what, who, why, and when. With a bit of discipline, each change is explained and self-contained. It becomes part of the infrastructure documentation. When the support team complains about a degraded experience for some customers over the last two months or so, you quickly discover this may be related to a change to an incoming policy in New York.
Rolling back
If a change is defective, it can be reverted quickly, safely, and without much effort, even if other changes happened in the meantime. The policy change at the origin of the problem spanned over three routers. Reverting this specific change and deploying the configuration let you solve the situation until you find a better fix.
Branching, reviewing, merging
When working on a new feature or refactoring some part of the infrastructure, a team member creates a branch and works on their change without interfering with the work of other members. Once the branch is ready, a pull request is created and the change is ready to be reviewed by the other team members before merging. You discover the issue was related to diverting traffic through an IX where one ISP was connected without enough capacity. You propose and discuss a fix that includes a change of the schema and the templates used to declare policies to be able to handle this case.
Continuous integration
For each change, automated tests are triggered. They can detect problems and give more details on the effect of a change. Branches can be deployed to a test infrastructure where regression tests are executed. The results can be synthesized as a comment in the pull request to help the review. You check your proposed change does not modify the other existing policies.

Why not NetBox? NetBox does not share these features. It is a database with a REST and a GraphQL API. Traceability is limited: changes are not grouped into a transaction and they are not documented. You cannot fork the database. Usually, there is one staging database to test modifications before applying them to the production database. It does not scale well and reviews are difficult. Applying the same change to the production database can be hazardous. Rolling back a change is non-trivial.

Update (2021-11) Nautobot, a fork of NetBox, will soon address this point by using Dolt, an SQL database engine allowing you to clone, branch, and merge, like a Git repository. Dolt is compatible with MySQL clients. See Nautobots, Roll Back! for a preview of this feature.

Moreover, NetBox is not usually the single source of truth. It contains your hardware inventory, the IP addresses, and some topology information. However, this is not the place you put authorized SSH keys, syslog servers, or the BGP configuration. If you also use Ansible, this information ends in its inventory. The source of truth is therefore fragmented between several tools with different workflows. Since NetBox 2.7, you can append additional data with configuration contexts. This mitigates this point. The data is arranged hierarchically but the hierarchy cannot be customized.2 Nautobot can manage configuration contexts in a Git repository, while still allowing to use of the API to fetch them. You get some additional perks, thanks to Git, but the remaining data is still in a database with a different lifecycle. Lastly, the schema used by NetBox may not fit your needs and you cannot tweak it. For example, you may have a rule to compute the IPv6 address from the IPv4 address for dual-stack interfaces. Such a relationship cannot be easily expressed and enforced in NetBox. When changing the IPv4 address, you may forget the IPv6 address. The source of truth should only contain the IPv4 address but you also want the IPv6 address in NetBox because this is your IPAM and you need it to update your DNS entries.

Why not Git? There are some limitations when putting your source of truth in Git:
  1. If you want to expose a web interface to allow an external team to request a change, it is more difficult to do it with Git than with a database. Out-of-the-box, NetBox provides a nice web interface and a permission system. You can also write your own web interface and interact with NetBox through its API.
  2. YAML files are more difficult to query in different ways. For example, looking for a free IP address is complex if they are scattered in multiple places.
In my opinion, in most cases, you are better off putting the source of truth in Git instead of NetBox. You get a lot of perks by doing that and you can still use NetBox as a read-only view, usable by other tools. We do that with an Ansible module. In the remaining cases, Git could still fit the bill. Read-only access control can be done through submodules. Pull requests can restrict write access: a bot can check the changes only modify allowed files before auto-merging. This still requires some Git knowledge, but many teams are now comfortable using Git, thanks to its ubiquity.

  1. Wikimedia manages its infrastructure with Puppet. They publish everything on GitHub. Creative Commons uses Salt. They also publish everything on GitHub. Thanks to them for doing that! I wish I could provide more real-life examples.
  2. Being able to customize the hierarchy is key to avoiding repetition in the data. For example, if switches are paired together, some data should be attached to them as a group and not duplicated on each of them. Tags can be used to partially work around this issue but you lose the hierarchical aspect.

9 November 2021

Joachim Breitner: How to audit an Internet Computer canister

I was recently called upon by Origyn to audit the source code of some of their Internet Computer canisters ( canisters are services or smart contracts on the Internet Computer), which were written in the Motoko programming language. Both the application model of the Internet Computer as well as Motoko bring with them their own particular pitfalls and possible sources for bugs. So given that I was involved in the creation of both, they reached out to me. In the course of that audit work I collected a list of things to watch out for, and general advice around them. Origyn generously allowed me to share that list here, in the hope that it will be helpful to the wider community.

Inter-canister calls The Internet Computer system provides inter-canister communication that follows the actor model: Inter-canister calls are implemented via two asynchronous messages, one to initiate the call, and one to return the response. Canisters process messages atomically (and roll back upon certain error conditions), but not complete calls. This makes programming with inter-canister calls error-prone. Possible common sources for bugs, vulnerabilities or simply unexpected behavior are:
  • Reading global state before issuing an inter-canister call, and assuming it to still hold when the call comes back.
  • Changing global state before issuing an inter-canister call, changing it again in the response handler, but assuming nothing else changes the state in between (reentrancy).
  • Changing global state before issuing an inter-canister call, and not handling failures correctly, e.g. when the code handling the callback rolls backs.
If you find such pattern in your code, you should analyze if a malicious party can trigger them, and assess the severity that effect These issues apply to all canisters, and are not Motoko-specific.

Rollbacks Even in the absence of inter-canister calls the behavior of rollbacks can be surprising. In particular, rejecting (i.e. throw) does not rollback state changes done before, while trapping (e.g. Debug.trap, assert , out of cycle conditions) does. Therefore, one should check all public update call entry points for unwanted state changes or unwanted rollbacks. In particular, look for methods (or rather, messages, i.e. the code between commit points) where a state change is followed by a throw. This issues apply to all canisters, and are not Motoko-specific, although other CDKs may not turn exceptions into rejects (which don t roll back).

Talking to malicious canisters Talking to untrustworthy canisters can be risky, for the following (likely incomplete) reasons:
  • The other canister can withhold a response. Although the bidirectional messaging paradigm of the Internet Computer was designed to guarantee a response eventually, the other party can busy-loop for as long as they are willing to pay for before responding. Worse, there are ways to deadlock a canister.
  • The other canister can respond with invalidly encoded Candid. This will cause a Motoko-implemented canister to trap in the reply handler, with no easy way to recover. Other CDKs may give you better ways to handle invalid Candid, but even then you will have to worry about Candid cycle bombs that will cause your reply handler to trap.
Many canisters do not even do inter-canister calls, or only call other trustwothy canisters. For the others, the impact of this needs to be carefully assessed.

Canister upgrade: overview For most services it is crucial that canisters can be upgraded reliably. This can be broken down into the following aspects:
  1. Can the canister be upgraded at all?
  2. Will the canister upgrade retain all data?
  3. Can the canister be upgraded promptly?
  4. Is three a recovery plan for when upgrading is not possible?

Canister upgradeability A canister that traps, for whatever reason, in its canister_preupgrade system method is no longer upgradeable. This is a major risk. The canister_preupgrade method of a Motoko canister consists of the developer-written code in any system func preupgrade() block, followed by the system-generated code that serializes the content of any stable var into a binary format, and then copies that to stable memory. Since the Motoko-internal serialization code will first serialize into a scratch space in the main heap, and then copy that to stable memory, canisters with more than 2GB of live data will likely be unupgradeable. But this is unlikely the first limit: The system imposes an instruction limit on upgrading a canister (spanning both canister_preupgrade and canister_postupgrade). This limit is a subnet configuration value, and sepearate (and likely higher) than the normal per-message limit, and not easily determined. If the canister s live data becomes too large to be serialized within this limit, the canister becomes non-upgradeable. This risk cannot be eliminated completely, as long as Motoko and Stable Variables are used. It can be mitigated by appropriate load testing: Install a canister, fill it up with live data, and exercise the upgrade. If this succeeds with a live data set exceeding the expected amount of data by a margin, this risk is probably acceptable. Bonus points for adding functionality that will prevent the canister s live data to increase above a certain size. If this testing is to be done on a local replica, extra care needs to be taken to make sure the local replica actually performs instruction counting and has the same resource limits as the production subnet. An alternative mitigation is to avoid canister_pre_upgrade as much as possible. This means no use of stable var (or restricted to small, fixed-size configuration data). All other data could be
  • mirrored off the canister (possibly off chain), and manually re-hydrated after an upgrade.
  • stored in stable memory manually, during each update call, using the ExperimentalStableMemory API. While this matches what high-assurance Rust canisters (e.g. the Internet Identity) do, This requires manual binary encoding of the data, and is marked experimental, so I cannot recommend this at the moment.
  • not put into a Motoko canister until Motoko has a scalable solution for stable variable (for example keeping them in stable memory permanently, with smart caching in main memory, and thus obliterating the need for pre-upgrade code.)

Data retention on upgrades Obviously, all live data ought to be retained during upgrades. Motoko automatically ensures this for stable var data. But often canisters want to work with their data in a different format (e.g. in objects that are not shared and thus cannot be put in stable vars, such as HashMap or Buffer objects), and thus may follow following idiom:
stable var fooStable =  ;
var foo = fooFromStable(fooStable);
system func preupgrade()   fooStable := fooToStable(foo);  )
system func postupgrade()   fooStable := (empty);  )
In this case, it is important to check that
  • All non-stable global vars, or global lets with mutable values, have a stable companion.
  • The assignments to foo and fooStable are not forgotten.
  • The fooToStable and fooFromStable form bijections.
An example would be HashMaps stored as arrays via Iter.toArray( .entries()) and HashMap.fromIter( .vals()). It is worth pointiong out that a code view will only look at a single version of the code, but cannot check whether code changes will preserve data on upgrade. This can easily go wrong if the names and types of stable variables are changed in incompatible way. The upgrade may fail loudly in this cases, but in bad cases, the upgrade may even succeed, losing data along the way. This risk needs to be mitigated by thorough testing, and possibly backups (see below).

Prompt upgrades Motoko and Rust canisters cannot be safely upgraded when they are still waiting for responses to inter-canister calls (the callback would eventually reach the new instance, and because of infelicities of the IC s System API, could possibly call arbitrary internal functions). Therefore, the canister needs to be stopped before upgrading, and started again. If the inter-canister calls take a long time, this mean that upgrading may take a long time, which may be undesirable. Again, this risk is reduced if all calls are made to trustworthy canisters, and elevated when possibly untrustworthy canisters are called, directly or indirectly.

Backup and recovery Because of the above risk around upgrades it is advisable to have a disaster recovery strategy. This could involve off-chain backups of all relevant data, so that it is possible to reinstall (not upgrade) the canister and re-upload all data. Note that reinstall has the same issue as upgrade described above in prompt upgrades : It ought to be stopped first to be safe. Note that the instruction limit for messages, as well as the message size limit, limit the amount of data returned. If the canister needs to hold more data than that, the backup query method might have to return chunks or deltas, with all the extra complexity that entails, e.g. state changes between downloading chunks. If large data load testing is performed (as Irecommend anyways to test upgradeability), one can test whether the backup query method works within the resource limits.

Time is not strictly monotonic The timestamps for current time that the Internet Computer provides to its canisters is guaranteed to be monotonic, but not strictly monotonic. It can return the same values, even in the same messages, as long as they are processed in the same block. It should therefore not be used to detect happens-before relations. Instead of using and comparing time stamps to check whether Y has been performed after X happened last, introduce an explicit var y_done : Bool state, which is set to False by X and then to True by Y. When things become more complex, it will be easier to model that state via an enumeration with speaking tag names, and update this state machine along the way. Another solution to this problem is to introduce a var v : Nat counter that you bump in every update method, and after each await. Now v is your canister s state counter, and can be used like a timestamp in many ways. While we are talking about time: The system time (typically) changes across an await. So if you do let now = Time.now() and then await, the value in now may no longer be what you want.

Wrapping arithmetic The Nat64 data type, and the other fixed-width numeric types provide opt-in wrapping arithmetic (e.g. +%, fromIntWrap). Unless explicitly required by the current application, this should be avoided, as usually a too large or negatie value is a serious, unrecoverable logic error, and trapping is the best one can do.

Cycle balance drain attacks Because of the IC s canister pays model, all canisters are prone to DoS attacks by draining their cycle balance, and this risk needs to be taken into account. The most elementary mitigation strategy is to monitor the cycle balance of canisters and keep it far from the (configurable) freezing threshold. On the raw IC-level, further mitigation strategies are possible:
  • If all update calls are authenticated, perform this authentication as quickly as possible, possibly before decoding the caller s argument. This way, a cycle drain attack by an unauthenticated attacker is less effective (but still possible).
  • Additionally, implementing the canister_inspect_message system method allows the above checks to be performed before the message even is accepted by the Internet Computer. But it does not defend against inter-canister messages and is therefore not a complete solution.
  • If an attack from an authenticated user (e.g. a stakeholder) is to be expected, the above methods are not effective, and an effective defense might require relatively involved additional program logic (e.g. per-caller statistics) to detect such an attack, and react (e.g. rate-limiting).
  • Such defenses are pointless if there is only a single method where they do not apply (e.g. an unauthenticated user registration method). If the application is inherently attackable this way, it is not worth the bother to raise defenses for other methods.
Related: A justification why the Internet Identity does not use canister_inspect_message) A motoko-implemented canister currently cannot perform most of these defenses: Argument decoding happens unconditionally before any user code that may reject a message based on the caller, and canister_inspect_message is not supported. Furthermore, Candid decoding is not very cycle defensive, and one should assume that it is possible to construct Candid messages that require many instructions to decode, even for simple argument type signatures. The conclusion for the audited canisters is to rely on monitoring to keep the cycle blance up, even during an attack, if the expense can be born, and maybe pray for IC-level DoS protections to kick in.

Large data attacks Another DoS attack vector exists if public methods allow untrustworthy users to send data of unlimited size that is persisted in the canister memory. Because of the translation of async-await code into multiple message handlers, this applies not only to data that is obviously stored in global state, but also local data that is live across an await point. The effectiveness of such attacks is limited by the Internet Computer s message size limit, which is in the order of a few megabytes, but many of those also add up. The problem becomes much worse if a method has an argument type that allows a Candid space bomb: It is possible to encode very large vectors with all values null in Candid, so if any method has an argument of type [Null] or [?t], a small message will expand to a large value in the Motoko heap. Other types to watch out:
  • Nat and Int: This is an unbounded natural number, and thus can be arbitrarily large. The Motoko representation will however not be much larger than the Candid encoding (so this does not qualify as a space bomb).
It is still advisable to check if the number is reasonable in size before storing it or doing an await. For example, when it denotes an index in an array, throw early if it exceeds the size of the array; if it denotes a token amount to transfer, check it against the available balance, if it denotes time, check it against reasonable bounds.
  • Principal: A Principal is effectively a Blob. The Interface specification says that principals are at most 29 bytes in length, but the Motoko Candid decoder does not check that currently (fixed in the next version of Motoko). Until then, a Principal passed as an argument can be large (the principal in msg.caller is system-provided and thus safe). If you cannot wait for the fix to reach you, manually check the size of the princial (via Principal.toBlob) before doing the await.

Shadowing of msg or caller Don t use the same name for the message context of the enclosing actor and the methods of the canister: It is dangerous to write shared(msg) actor, because now msg is in scope across all public methods. As long as these also use public shared(msg) func , and thus shadow the outer msg, it is correct, but it if one accidentially omits or mis-types the msg, no compiler error would occur, but suddenly msg.caller would now be the original controller, likely defeating an important authorization step. Instead, write shared(init_msg) actor or shared( caller = controller ) actor to avoid using msg.

Conclusion If you write a serious canister, whether in Motoko or not, it is worth to go through the code and watch out for these patterns. Or better, have someone else review your code, as it may be hard to spot issues in your own code. Unfortunately, such a list is never complete, and there are surely more ways to screw up your code in addition to all the non-IC-specific ways in which code can be wrong. Still, things get done out there, so best of luck!

31 October 2021

Joachim Breitner: A mostly allocation-free optional type

The Motoko programming language has a built-in data type for optional values, named ?t with values null and ?v (for v : t); this is the equivalent of Haskell s Maybe, Ocaml s option or Rust s Option. In this post, I explain how Motoko represents such optional values (almost) without allocation. I neither claim nor expect that any of this is novel; I just hope it s interesting.

Uniform representation The Motoko programming language, designed by Andreas Rossberg and implemented by a pretty cool team at DFINITY is a high-level language with strict semantics and a strong, structural, equi-recursive type system that compiles down to WebAssembly. Because the type system supports polymorphism, it represents all values uniformly. Simplified for the purpose of this blog post, they are always pointers into a heap object where the first word of the heap object, the heap tag, contains information about the value:
 
  tag    
 
The tag is something like array, int64, blob, variant, record, , and it has two purposes:
  • The garbage collector uses it to understand what kind of object it is looking at, so that it can move it around and follow pointers therein. Variable size objects such as arrays include the object size in a subsequent word of the heap object.
  • Some types have values that may have different shapes on the heap. For example, the ropes used in our text representation can either be a plain blob, or a concatenation node of two blobs. For these types, the tag of the heap object is inspected.

The optional type, naively The optional type (?t) is another example of such a type: Its values can either be null, or ?v for some value v of type t, and the primitive operations on this type are the two introduction forms, an analysis function, and a projection for non-null values:
null : () -> ?t
some : t -> ?t
is_some : ?t -> bool
project : ?t -> t     // must only be used if is_some holds
It is natural to use the heap tag to distinguish the two kind of values:
  • The null value is a simple one-word heap object with just a tag that says that this is null:
     
      null  
     
  • The other values are represented by a two-word object: The tag some, indicating that it is a ?v, and then the payload, which is the pointer that represents the value v:
     
      some   payload  
     
With this, the four operations can be implemented as follows:
def null():
  ptr <- alloc(1)
  ptr[0] = NULL_TAG
  return ptr
def some(v):
  ptr <- alloc(2)
  ptr[0] = SOME_TAG
  ptr[1] = v
  return ptr
def is_some(p):
  return p[0] == SOME_TAG
def project(p):
  return p[1]
The problem with this implementation is that null() and some(v) both allocate memory. This is bad if they are used very often, and very liberally, and this is the case in Motoko: For example the iterators used for the for (x in e) construct have type
type Iter<T> =   next : () -> ?T  
and would unavoidably allocate a few words on each iteration. Can we avoid this?

Static values It is quite simple to avoid this for for null: Just statically create a single null value and use it every time:
static_null = [NULL_TAG]
def null():
  return static_null
This way, at least null() doesn t allocate. But we gain more: Now every null value is represented by the same pointer, and since the pointer points to static memory, it does not change even with a moving garbage collector, so we can speed up is_some:
def is_some(p):
  return p != static_null
This is not a very surprising change so far, and most compilers out there can and will do at least the static allocation of such singleton constructors. For example, in Haskell, there is only a single empty list ([]) and a single Nothing value in your program, as you can see in my videos exploring the Haskell heap. But can we get rid of the allocation in some(v) too?

Unwrapped optional values If we don t want to allocate in some(v), what can we do? How about simply
def some(v):
  return v
That does not allocate! But it is also broken. At type ??Int, the values null, ?null and ??null are distinct values, but here this breaks. Or, more formally, the following laws should hold for our four primitive operations:
  1. is_some(null()) = false
  2. v. is_some(some(v)) = true
  3. p. project(some(p)) = p
But with the new definition of some, we d get is_some(some(null())) = false. Not good! But note that we only have a problem if we are wrapping a value that is null or some(v). So maybe take the shortcut only then, and write the following:
def some(v):
  if v == static_null   v[0] == SOME_TAG:
    ptr <- alloc(2)
    ptr[0] = SOME_TAG
    ptr[1] = v
    return ptr
  else:
    return v
The definition of is_some can stay as it is: It is still the case that all null values are represented by static_null. But the some values are now no longer all of the same shape, so we have to change project():
def project(p):
  if p[0] == SOME_TAG:
    return p[1]
  else:
    return p
Now things fall into place: A value ?v can, in many cases, be represented the same way as v, and no allocation is needed. Only when v is null or ?null or ??null or ???null etc. we need to use the some heap object, and thus have to allocate. In fact, it doesn t cost much to avoid allocation even for ?null:
static_some_null = [SOME_TAG, static_null]
def some(v):
  if v == static_null:
    return static_some_null
  else if v[0] == SOME_TAG:
    ptr <- alloc(2)
    ptr[0] = SOME_TAG
    ptr[1] = v
    return ptr
  else:
    return v
So unless one nests the ? type two levels deep, there is no allocation whatsoever, and the only cost is a bit more branching in some and project. That wasn t hard, but quite rewarding, as one can now use idioms like the iterator shown above with greater ease.

Examples The following tables shows the representation of various values before and after. Here [ ] is a pointed-to dynamically allocated heap object, a statically allocated heap object, N = NULL_TAG and S = SOME_TAG.
type value before after
Null null N N
?Int null N N
?Int ?23 [S,23] 23
??Int null N N
??Int ?null [S, N ] S, N
??Int ??23 [S,[S,23]] 23
???Int null N N
???Int ?null [S, N ] S, N
???Int ??null [S,[S, N ]] [S, S, N ]
???Int ???23 [S,[S,[S,23]]] 23

Concluding remarks
  • If you know what parametric polymorphism is, and wonder how this encoding can work in a language that has that, note that this representation is on the low-level of the untyped run-time value representation: We don t need to know the type of v in some(v), merely its heap representation.
  • The uniform representation in Motoko is a bit more involved: The pointers are tagged (by subtracting 1) and some scalar values are represented directly in place (shifted left by 1 bit). But this is luckily orthogonal to what I showed here.
  • Could Haskell do this for its Maybe type? Not so easily:
    • The Maybe type is not built-in, but rather a standard library-defined algebraic data type. But the compiler could feasible detect that this is option-like?
    • Haskell is lazy, so at runtime, the type Maybe could be Nothing, or Just v, or, and this is crucial, a yet to be evaluated expression, also called a thunk. And one definitely needs to distinguish between a thunk t :: Maybe a that may evaluate to Nothing, and a value Just t :: Maybe a that definitely is Just, but contains a value, which may be a thunk.
    Something like this may work for a strict Maybe type or unlifted sums like (# (# #) a #), but may clash with other ticks in GHC, such as pointer tagging.
  • As I said above, I don t expect this to be novel, and I am happy to add references to prior art here.
  • Given that a heap object with tag SOME_TAG now always encodes a tower ? null for n>0, one could try to optimize that even more by just storing the n:
     
      some    n   
     
    But that seems unadvisable: It is only a win if you have deep towers, which is probably rare. Worse, now the project function would have to return such a heap object with n decremented, so now projection might have to allocate, which goes against the cost model expected by the programmer.
  • If you rather want to see code than blog posts, feel free to check out Motoko PR #2115.
  • Does this kind of stuff excite you? Motoko is open source, so your contributions may be welcome!

Next.

Previous.