• Bepasty for sharing files,
• Ikiwiki for taking notes,
• Quassel for staying connected to IRC chat servers,
• Radicale for synchronizing my calendar and tasks,
• Shaarli for sharing bookmarks, and
• Tiny Tiny RSS for reading other people s blogs.

• It is not part of my educational or professional background (which mainly consists of embedded software engineering).
• I think that I lack the kind of discipline needed to be a sysadmin. For example, taking notes of the commands that you are running, or testing things before you do them on a production server. I m likely to just run commands, and lose track of my notes.
• And finally a type of laziness. I just don t want to spend a lot of time and effort setting up or maintaining some services.

• I do have a server running in my home, but I don t consider myself a sysadmin.
• I have a different kind of discipline than the one I described above.
• I m willing to spend time and effort to improve things, but I want to do it in a different way than usual.

• In the various Debian packages of the services I have installed.
• In FreedomBox configuration and integration of those packages.
• Or even directly to the upstream software development.

• The software I am running is very standardized, specifically to what is available in Debian and FreedomBox. This limits the number of services available: for example, I could not easily run a Mastodon server with this approach. Also, I am not maintaining any custom configurations for my own services.
• I am presenting these ideas as they occur to me, and specific to my situation. But there are similar concepts out there, for example Infrastructure as Code .

podman pull registry.access.redhat.com/ubi9/openjdk-11
podman pull registry.access.redhat.com/ubi9/openjdk-17
podman pull registry.access.redhat.com/ubi9/openjdk-11-runtime
podman pull registry.access.redhat.com/ubi9/openjdk-17-runtime

• Some kind of 45 leaning prong to dry bottles and flasks on
• A tea tray and coasters
• Small tins to keep loose-leaf tea in

A vinyl record.

a replacement prop arm/foot for my computer keyboard

replacement toy bolt

replacement bits for an Early Learning Centre Build It Deluxe Set

Little kids trinkets. Pacman ghosts

The Pacman ghost family so far

• Lego storage/sorters
• DIY bits-and-bobs sorter/storage (nuts and bolts etc)

A free-standing inclined vinyl record display stand

A bracket to install a Gotek drive in my Amiga 500

To Throw or not to throw Here is our little toy language to analyze: It has variables, lambdas and applications, non-recursive (lazy) let bindings and, so that we have something to analyze, a way to throw and to catch exceptions:

type Var = String

data Exp
  = Var Var
    Throw
    Catch Exp
    Lam Var Exp
    App Exp Exp
    Let Var Exp Exp

Given such an expression, we would like to know whether it might throw an exception. Such an analysis is easy to write: We traverse the syntax tree, remembering in the env which of the variables may throw an exception:

canThrow1 :: Exp -> Bool
canThrow1 = go M.empty
  where
    go :: M.Map Var Bool -> Exp -> Bool
    go env (Var v)       = env M.! v
    go env Throw         = True
    go env (Catch e)     = False
    go env (Lam v e)     = go (M.insert v False env) e
    go env (App e1 e2)   = go env e1   go env e2
    go env (Let v e1 e2) = go env' e2
      where
        env_bind = M.singleton v (go env e1)
        env' = M.union env_bind env

The most interesting case is the one for Let, where we extend the environment env with the information about the additional variable env_bind, which is calculated from analyzing the right-hand side e1. So far so good:

ghci> someVal = Lam "y" (Var "y")
ghci> canThrow1 $ Throw
True
ghci> canThrow1 $ Let "x" Throw someVal
False
ghci> canThrow1 $ Let "x" Throw (App (Var "x") someVal)
True

Let it rec To spice things up, let us add a recursive let to the language:

data Exp
   
    LetRec [(Var, Exp)] Exp

How can we support this new constructor in canThrow1? Let use naively follow the pattern used for Let: Calculate the analysis information for the variables in env_bind, extend the environment with that, and pass it down:

    go env (LetRec binds e) = go env' e
      where
        env_bind = M.fromList [ (v, go env' e)   (v,e) <- binds ]
        env' = M.union env_bind env

Note that, crucially, we use env', and not just env, when analyzing the right-hand sides. It has to be that way, as all the variables are in scope in all the right-hand sides. In a strict language, such a mutually recursive definition, where env_bind uses env' which uses env_bind is basically unthinkable. But in a lazy language like Haskell, it might just work. Unfortunately, it works only as long as the recursive bindings are not actually recursive, or if they are recursive, they are not used:

ghci> canThrow1 $ LetRec [("x", Throw)] (Var "x")
True
ghci> canThrow1 $ LetRec [("x", App (Var "y") someVal), ("y", Throw)] (Var "x")
True
ghci> canThrow1 $ LetRec [("x", App (Var "x") someVal), ("y", Throw)] (Var "y")
True

But with genuine recursion, it does not work, and simply goes into a recursive cycle:

ghci> canThrow1 $ LetRec [("x", App (Var "x") someVal), ("y", Throw)] (Var "x")
^CInterrupted.

That is disappointing! Do we really have to toss that code and somehow do an explicit fixed-point calculation here? Obscuring our nice declarative code? And possibly having to repeat work (such as traversing the syntax tree) many times that we should only have to do once?

rec-def to the rescue Not with rec-def! Using RBool from Data.Recursive.Bool instead of Bool, we can write the exact same code, as follows:

import qualified Data.Recursive.Bool as RB

canThrow2 :: Exp -> Bool
canThrow2 = RB.get . go M.empty
  where
    go :: M.Map Var RBool -> Exp -> R Bool
    go env (Var v)       = env M.! v
    go env Throw         = RB.true
    go env (Catch e)     = RB.false
    go env (Lam v e)     = go (M.insert v RB.false env) e
    go env (App e1 e2)   = go env e1 RB.  go env e2
    go env (Let v e1 e2) = go env' e2
      where
        env_bind = M.singleton v (go env e1)
        env' = M.union env_bind env
    go env (LetRec binds e) = go env' e
      where
        env_bind = M.fromList [ (v, go env' e)   (v,e) <- binds ]
        env' = M.union env_bind env

And it works!

ghci> canThrow2 $ LetRec [("x", App (Var "x") someVal), ("y", Throw)] (Var "x")
False
ghci> canThrow2 $ LetRec [("x", App (Var "x") Throw), ("y", Throw)] (Var "x")
True

I find this much more pleasing than the explicit naive fix-pointing you might do otherwise, where you stabilize the result at each LetRec independently: Not only is all that extra work hidden from the programmer, but now also a single traversal of the syntax tree creates, thanks to the laziness, a graph of RBool values, which are then solved under the hood .

The issue with x=x There is one downside worth mentioning: canThrow2 fails to produce a result in case we hit x=x:

ghci> canThrow2 $ LetRec [("x", Var "x")] (Var "x")
^CInterrupted.

This is, after all the syntax tree has been processed and all the map lookups have been resolved, equivalent to

ghci> let x = x in RB.get (x :: RBool)
^CInterrupted.

which also does not work. The rec-def machinery can only kick in if at least one of its function is used on any such cycle, even if it is just a form of identity (which I ~ought to add to the library~ since have added to the library):

ghci> idR x = RB.false   x
ghci> let x = idR x in getR (x :: R Bool)
False

And indeed, if I insert a call to idR in the line

        env_bind = M.fromList [ (v, idR (go env' e))   (v,e) <- binds ]

then our analyzer will no longer stumble over these nasty recursive equations:

ghci> canThrow2 $ LetRec [("x", Var "x")] (Var "x")
False

It is a bit disappointing to have to do that, but I do not see a better way yet. I guess the def-rec library expects the programmer to have a similar level of sophistication as other tie-the-know tricks with laziness (where you also have to ensure that your definitions are productive and that the sharing is not accidentally lost).

$ mkdir yakuake
$ wget https://steamdeck-packages.steamos.cloud/archlinux-mirror/extra/os/x86_64/yakuake-21.12.1-1-x86_64.pkg.tar.zst
$ tar -C yakuake -xaf yakuake-*.tar.zst usr

$ mkdir -p yakuake/usr/lib/extension-release.d/
$ echo ID=steamos > yakuake/usr/lib/extension-release.d/extension-release.yakuake
$ echo VERSION_ID=3.3.1 >> yakuake/usr/lib/extension-release.d/extension-release.yakuake

$ mksquashfs yakuake yakuake.raw

(deck@steamdeck ~)$ mkdir extensions
(deck@steamdeck ~)$ scp user@host:/path/to/yakuake.raw extensions/
(deck@steamdeck ~)$ sudo ln -s $PWD/extensions /var/lib/extensions

(deck@steamdeck ~)$ sudo systemctl enable systemd-sysext
(deck@steamdeck ~)$ sudo systemctl start systemd-sysext

$ systemd-sysext status
HIERARCHY EXTENSIONS SINCE
/opt      none       -
/usr      yakuake    Tue 2022-09-13 18:21:53 CEST

1. Using extensions is easy (you put them in the directory and voil !). However, creating extensions is not necessarily always easy. To begin with, any libraries, files, etc., that your extensions may need should be either present in the root filesystem or provided by the extension itself. You may need to combine files from different sources or packages into a single extension, or compile them yourself.
2. In particular, if the extension contains binaries they should probably come from the Steam Deck repository or they should be built to work with those packages. If you need to build your own binaries then having a SteamOS virtual machine can be handy. There you can install all development files and also test that everything works as expected. One could also create a Steam Deck SDK extension with all the necessary files to develop directly on the Deck
3. Extensions are not distribution packages, they don t have dependency information and therefore they should be self-contained. They also lack triggers and other features available in packages. For desktop applications I still recommend using a system like Flatpak when possible.
4. Extensions are tied to a particular version of the OS and, as explained above, the ID and VERSION_ID of each extension must match the values from /etc/os-release. If the fields don t match then the extension will be ignored. This is to be expected because there s no guarantee that a particular extension is going to work with a different version of the OS. This can happen after a system update. In the best case one simply needs to update the extension s VERSION_ID, but in some cases it might be necessary to create the extension again with different/updated files.
5. Extensions only install files in /usr and /opt. Any other file in the image will be ignored. This can be a problem if a particular piece of software needs files in other directories.
6. When extensions are enabled the /usr and /opt directories become read-only because they are now part of an overlayfs. They will remain read-only even if you run steamos-readonly disable !!. If you really want to make the rootfs read-write you need to disable the extensions (systemd-sysext unmerge) first.
7. Unlike Flatpak or Podman (including toolbox / distrobox), this is (by design) not meant to isolate the contents of the extension from the rest of the system, so you should be careful with what you re installing. On the other hand, this lack of isolation makes systemd-sysext better suited to some use cases than those container-based systems.

(The picture is of the previous edition.) Almost two years after the previous Norwegian Bokm l translation of the "The Debian Administrator's Handbook" was published, a new edition is finally being prepared. The english text is updated, and it is time to start working on the translations. Around 37 percent of the strings have been updated, one way or another, and the translations starting from a complete Debian Buster edition now need to bring their translation up from 63% to 100%. The complete book is licensed using a Creative Commons license, and has been published in several languages over the years. The translations are done by volunteers to bring Linux in their native tongue. The last time I checked, it complete text was available in English, Norwegian Bokm l, German, Indonesian, Brazil Portuguese and Spanish. In addition, work has been started for Arabic (Morocco), Catalan, Chinese (Simplified), Chinese (Traditional), Croatian, Czech, Danish, Dutch, French, Greek, Italian, Japanese, Korean, Persian, Polish, Romanian, Russian, Swedish, Turkish and Vietnamese. The translation is conducted on the hosted weblate project page. Prospective translators are recommeded to subscribe to the translators mailing list and should also check out the instructions for contributors. I am one of the Norwegian Bokm l translators of this book, and we have just started. Your contribution is most welcome. As usual, if you use Bitcoin and want to show your support of my activities, please send Bitcoin donations to my address 15oWEoG9dUPovwmUL9KWAnYRtNJEkP1u1b.

Changes in RcppArmadillo version 0.11.2.4.0 (2022-09-09)

• Upgraded to Armadillo release 11.2.4 (Classic Roast)
  • fix handling of std::move() involving matrices constructed with auxiliary memory
• In the fastLm() examples, use arma::dot() to compute to the inner product (as proposed by Conrad), plus small edits
• Support optional #define named RCPPARMADILLO_FORCE_DEPRECATE to suppress use of ARMA_IGNORE_DEPRECATED_MARKER permitting use and development under deprecation

This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. Please report excessive re-aggregation in third-party for-profit settings.

Politics I have been reading books, both fiction and non-fiction for a long long time. My first book was a comic most probably when I was down with Malaria when I was a kid. I must be around 4-5 years old. Over the years, books have given me great joy and I continue to find nuggets of useful information, both in fiction as well as non-fiction books. So here s to sharing something and how that can lead you to a rabbit hole. This entry would be a bit NSFW as far as language is concerned. NYPD Red 5 by James Patterson First of all, have no clue as to why James Patterson s popularity has been falling. He used to be right there with Lee Child and others, but not so much now. While I try to be mysterious about books, I would give a bit of heads-up so people know what to expect. This is probably more towards the Adult crowd as there is a bit of sex as well as quite a few grey characters. The NYPD Red is a sort of elite police task force that basically is for celebrities. In the book series, they do a lot of ass-kissing (figuratively more than literally). Now the reason I have always liked fiction is that however wild the assumption or presumption is, it does have somewhere a grain of truth. And each and every time I read a book or two, that gets cemented. One of the statements in the book told something about how 9/11 took a lot of police personnel out of the game. First, there were a number of policemen who were patrolling the Two Towers, so they perished literally during the explosion. Then there were policemen who were given the cases to close the cases (bring the cases to conclusion). When you are investigating your own brethren or even civilians who perished 9/11 they must have experienced emotional trauma and no outlet. Mental health even in cops is the same and given similar help as you and me (i.e. next to none.) But both of these were my assumptions. The only statement that was in the book was they lost a lot of bench strength. Even NYFD (New York Fire Department). This led me to me to With Crime At Record Lows, Should NYC Have Fewer Cops? This is more right-wing sentiment and in fact, there have been calls to defund the police. This led me to https://cbcny.org/ and one specific graph. Unfortunately, this tells the story from 2010-2022 but not before. I was looking for data from around 1999 to 2005 because that will tell whether or not it happened. Then I remembered reading in newspapers the year or two later how 9/11 had led NYC to recession. I looked up online and for sure NY was booming before 9/11. One can argue that NYC could come down and that is pretty much possible, everything that goes up comes down, it s a law of nature but it would have been steady rather than abrupt. And once you are in recession, the first thing to go is personnel. So people both from NYPD and NYFD were let go, even though they were needed the most then. As you can see, a single statement in a book can take you to places & time literally. Edit: Addition 11th September There were quite a few people who also died from New York Port Authority and they also lost quite a number of people directly and indirectly and did a lot of patrolling of the water bodies near NYC. Later on, even in their department, there were a lot of early retirements.

Kosovo A couple of days back I had a look at the Debconf 2023 BOF that was done in Kosovo. One of the interesting things that happened during the BOF is when a woman participant chimed in and asks India to recognize Kosovo. Immediately it triggered me and I opened the Kosovo Wikipedia page to get some understanding of the topic. Reading up on it, came to know Russia didn t agree and doesn t recognize Kosovo. Mr. Modi likes Putin and India imports a lot of its oil from Russia. Unrelatedly, but still useful, we rejected to join IPEF. Earlier, we had rejected China s BRI. India has never been as vulnerable as she is now. Our foreign balance has reached record lows. Now India has been importing quite a bit of Russian crude and has been buying arms and ammunition from them. We are also scheduled to buy a couple of warships and submarines etc. We even took arms and ammunition from them on lease. So we can t afford that they are displeased with India. Even though Russia has more than friendly relations with both China and Pakistan. At the same time, the U.S. is back to aiding Pakistan which the mainstream media in India refuses to even cover. And to top all of this, we have the Chip 4 Alliance but that needs its own article, truth be told but we will do with a paragraph Edit Addition 11th September Seems Kosovo isn t unique in that situation, there are 3-4 states like that. A brief look at worldpopulationreview tells you there are many more.

Chip 4 Alliance For almost a decade I have been screaming about this on my blog as well as everywhere that chip fabrication is a national security thing. And for years, most people deny it. And now we have chip 4 alliance. Now to understand this, you have to understand that China for almost a decade, somewhere around 2014 or so came up with something called the big fund . Now one can argue one way or the other how successful the fund has been, but it has, without doubt, created ripples so strong that the U.S., Taiwan, Japan, and probably South Korea will join and try to stem the tide. Interestingly, in this grouping, South Korea is the weakest in the statements and what they have been saying. Within the group itself, there is a lot of tension and China would use that and there are a number of unresolved issues between the three countries that both China & Russia would exploit. For e.g. the Comfort women between South Korea and Japan. Or the 1985 Accord Agreement between Japan and the U.S. Now people need to understand this, this is not just about China but also about us. If China has 5-6x times India s GDP and their research budget is at the very least 100x times what India spends, how do you think we will be self-reliant? Whom are we fooling? Are we not tired of fooling ourselves In diplomacy, countries use leverage. Sadly, we let go of some of our most experienced negotiators in 2014 and since then have been singing in the wind

Accessibility, Jitsi, IRC, Element-Desktop The Wikipedia page on Accessibility says the following Accessibility is the design of products, devices, services, vehicles, or environments so as to be usable by people with disabilities. The concept of accessible design and practice of accessible development ensures both direct access (i.e. unassisted) and indirect access meaning compatibility with a person s assistive technology. Now IRC or Internet Relay Chat has been accessible for a long time. I know of even blind people who have been able to navigate IRC quite effortlessly as there has been a lot of work done to make sure all the joints speak to each other so people with one or more disabilities still can use, and contribute without an issue. It does help that IRC and many clients have been there since the 1970s so most of them have had more than enough time to get all the bugs fixed and both text-to-speech and speech-to-text work brilliantly on IRC. Newer software like Jitsi or for that matter Telegram is lacking those features. A few days ago, discovered on Telegram I was shared that Samsung Voice input is also able to do the same. The Samsung Voice Input works wonder as it translates voice to text, I have not yet tried the text-to-speech but perhaps somebody can and they can share whatever the results can be one way or the other. I have tried element-desktop both on the desktop as well as mobile phone and it has been disappointing, to say the least. On the desktop, it is unruly and freezes once in a while, and is buggy. The mobile version is a little better but that s not saying a lot. I prefer the desktop version as I can use the full-size keyboard. The bug I reported has been there since its Riot days. I had put up a bug report even then. All in all, yesterday was disappointing

"Look at it this way, then," she said, and took a deep mental breath. "Wherever people are obtuse and absurd... and wherever they have, by even the most generous standards, the attention span of a small chicken in a hurricane and the investigative ability of a one-legged cockroach... and wherever people are inanely credulous, pathetically attached to the certainties of the nursery and, in general, have as much grasp of the physical universe as an oyster has of mountaineering... yes, Twyla: there is a Hogfather.

"All right," said Susan. "I'm not stupid. You're saying humans need... fantasies to make life bearable." REALLY? AS IF IT WAS SOME KIND OF PINK PILL? NO. HUMANS NEED FANTASY TO BE HUMAN. TO BE THE PLACE WHERE THE FALLING ANGEL MEETS THE RISING APE. "Tooth fairies? Hogfathers? Little " YES. AS PRACTICE. YOU HAVE TO START OUT LEARNING TO BELIEVE THE LITTLE LIES. "So we can believe the big ones?" YES. JUSTICE. MERCY. DUTY. THAT SORT OF THING. "They're not the same at all!" YOU THINK SO? THEN TAKE THE UNIVERSE AND GRIND IT DOWN TO THE FINEST POWDER AND SIEVE IT THROUGH THE FINEST SIEVE AND THEN SHOW ME ONE ATOM OF JUSTICE, ONE MOLECULE OF MERCY. AND YET Death waved a hand. AND YET YOU ACT AS IF THERE IS SOME IDEAL ORDER IN THE WORLD, AS IF THERE IS SOME...SOME RIGHTNESS IN THE UNIVERSE BY WHICH IT MAY BE JUDGED. "Yes, but people have got to believe that, or what's the point " MY POINT EXACTLY.

let s1 = RS.insert 23 s2
    s2 = RS.insert 42 s1
in RS.get s1

insert :: Ord a => a -> RSet a -> RSet a
get :: RSet a -> Set a

IO-infested propagator interfaces This suggests that an implementation of such a propagator network could provide an interface with the following three operations:

• Functions to declare cells
• Functions to declare relations
• Functions to read values off cells

This is clearly an imperative interface, so we ll see monads, and we ll simply use IO. So concretely for our small example above, we might expect

data Cell a
newCell :: IO (Cell a)
-- declareInsert x c1 c2 declares that the value of c1 ought to be S.insert x c2
declareInsert :: Ord a => a -> Cell a -> Cell a -> IO ()
getCell :: Cell a -> IO (Set a)

There is no need for an explicit solve function: solving can happen when declareInsert or getCell is called as a User I do not care about that. You might be curious about the implementation of newCell, declareInsert and getCell, but I have to disappoint you: This is not the topic of this article. Instead, I want to discuss how to turn this IO-infested interface into the pure interface seen above?

Pure, but too strict Obviously, we have to get rid of the IO somehow, and have to use unsafePerformIO :: IO a -> a somehow. This dangerous function creates a pure-looking value that, when used the first time, will run the IO-action and turn into that action s result. So maybe we can simply write the following:

type RSet a = Cell a
insert x c2 = unsafePerformIO $ do
  c1 <- newCell
  declareInsert x c1 c2
  return c1
get c = unsafePerformIO $
  getCell c

Indeed, the types line up, but if we try to use that code, nothing will happen. Our insert is too strict to be used recursively: It requires the value of c2 (as it is passed to declareInsert, which we assume to be strict in its arguments) before it can return c1, so the recursive example at the top of this post will not make any progress.

Pure, lazy, but forgetful To work around this, maybe it suffices if we do not run declareInsert right away, but just remember that we have to do it eventually? So let s introduce a new data type for RSet a that contains not just the cell (Cell a), but also an action that we still have to run before getting a value:

data RSet a = RSet (Cell a) (IO ())
insert x r2 = unsafePerformIO $ do
  c1 <- newCell
  let todo = do
        let (RSet c2 _) = r2
        declareInsert x c1 c2
  return (RSet c1 todo)
get (RSet c todo) = unsafePerformIO $ do
  todo
  getCell c

This is better: insert is now lazy in its arguments (for this it is crucial to pattern-match on RSet only inside the todo code, not in the pattern of insert!) This means that our recursive code above does not get stuck right away.

Pure, lazy, but runs in circles But it is still pretty bad: Note that we do not run get s2 in the example above, so that cell s todo, which would declareInsert 42, will never run. This cannot work! We have to (eventually) run the declaration code from all involved cells before we can use getCell! We can try to run the todo action of all the dependencies as part of a cell s todo action:

data RSet a = RSet (Cell a) (IO ())
insert x r2 = unsafePerformIO $ do
  c1 <- newCell
  let todo = do
        let (RSet c2 todo2) = r2
        declareInsert x c1 c2
        todo2
  return (RSet c1 todo)
get (RSet c todo) = unsafePerformIO $ do
  todo
  getCell c

Now we certainly won t forget to run the second cell s todo action, so that is good. But that cell s todo action will run the first cell s todo action, and that again the second cell s, and so on.

Pure, lazy, terminating, but not thread safe This is silly: We only need (and should!) run that code once! So let s keep track of whether we ran it already:

data RSet a = RSet (Cell a) (IO ()) (IORef Bool)
insert x r2 = unsafePerformIO $ do
  c1 <- newCell
  done <- newIORef False
  let todo = do
        is_done <- readIORef done
        unless is_done $ do
          writeIORef done True
          let (RSet c2 todo2 _) = r2
          declareInsert x c1 c2
          todo2
  return (RSet c1 todo done)
get (RSet c todo _) = unsafePerformIO $ do
  todo
  getCell c

Ah, much better: It works! Our call to get c1 will trigger the first cell s todo action, which will mark it as done before calling the second cell s todo action. When that now invokes the first cell s todo action, it is already marked done and we break the cycle, and by the time we reach getCell, all relations have been correctly registered. In a single-threaded world, this would be all good and fine, but we have to worry about multiple threads running get concurrently, on the same or on different cells. In fact, because we use unsafePerformIO, we have to worry about this even when the program is not using threads. And the above code has problems. Imagine a second call to get c1 while the first one has already marked it as done, but has not finished processing all the dependencies yet: It will call getCell before all relations are registered, which is bad.

Recursive do-once IO actions Making this thread-safe seems to be possible, but would clutter both the code and this blog post. So let s hide that problem behind a nice and clean interface. Maybe there will be a separate blog post about its implementation (let me know if you are curious), or you can inspect the code in System.IO.RecThunk module yourself). The interface is simply

data Thunk
thunk :: IO [Thunk] -> IO Thunk
force :: Thunk -> IO ()

and the idea is that thunk act will defer the action act until the thunk is passed to force for the first time, and force will not return until the action has been performed (possibly waiting if another thread is doing that at the moment), and also until the actions of all the thunks returned by act have performed, recursively, without running into cycles. We can use this in our definition of RSet and get to the final, working solution:

data RSet a = RSet (Cell a) Thunk
insert x r2 = unsafePerformIO $ do
  c1 <- newCell
  t1 <- thunk $ do
    let (RSet c2 t2) = r2
    declareInsert x c1 c2
    return [t2]
  return (RSet c1 t1)
get (RSet c t) = unsafePerformIO $ do
  force t
  getCell c

This snippet captures the essential ideas behind rec-def:

• Use laziness to allow recursive definition to describe the propagator graph naturally
• Use a form of explicit thunk to register the propagator graph relations at the right time (not too early/strict, not too late)

And that s all? The actual implementation in rec-def has a few more moving parts. In particular, it tries to support different value types (not just sets), possibly with different implementations, and even mixing them (e.g. in member :: Ord a => a -> RSet a -> RBool), so the generic code is in Data.Propagator.Purify, and supports various propagators underneath. The type RSet is then just a newtype around that, defined in Data.Recursive.Internal to maintain the safety of the abstraction, I went back and forth on a few variants of the design here, including one where there was a generic R type constructor (R (Set a), R Bool etc.), but then monomorphic interface seems simpler.

Does it really work? The big remaining question is certainly: Is this really safe and pure? Does it still behave like Haskell? The answer to these questions certainly depends on the underlying propagator implementation. But it also depends on what we actually mean by safe and pure ? For example, do we expect the Static Argument Transformation be semantics preserving? Or is it allowed to turn undefined values into defined ones (as it does here)? I am unsure myself yet, so I ll defer this discussion to a separate blog post, after I hopefully had good discussions about this here at ICFP 2022 in Ljubljana. If you are around and want to discuss, please hit me up!

Community news As announced last month, registration is currently open for our in-person summit this year which is due to be held between November 1st November 3rd. The event will take place in Venice (Italy). Very soon we intend to pick a venue reachable via the train station and an international airport. However, the precise venue will depend on the number of attendees. Please see the announcement email for information about how to register.

---

The US National Security Agency (NSA), Cybersecurity and Infrastructure Security Agency (CISA) and the Office of the Director of National Intelligence (ODNI) have released a document called Securing the Software Supply Chain: Recommended Practices Guide for Developers (PDF) as part of their Enduring Security Framework (ESF) work. The document expressly recommends having reproducible builds as part of advanced recommended mitigations, along with hermetic builds. Page 31 (page 35 in the PDF) says:

Reproducible builds provide additional protection and validation against attempts to compromise build systems. They ensure the binary products of each build system match: i.e., they are built from the same source, regardless of variable metadata such as the order of input files, timestamps, locales, and paths. Reproducible builds are those where re-running the build steps with identical input artifacts results in bit-for-bit identical output. Builds that cannot meet this must provide a justification why the build cannot be made reproducible.

The full press release is available online.

---

On our mailing list this month, Marc Prud hommeaux posted a feature request for diffoscope which additionally outlines a project called The App Fair, an autonomous distribution network of free and open-source macOS and iOS applications, where validated apps are then signed and submitted for publication .

---

Author/blogger Cory Doctorow posted published a provocative blog post this month titled Your computer is tormented by a wicked god . Touching on Ken Thompson s famous talk, Reflections on Trusting Trust , the early goals of Secure Computing and UEFI firmware interfaces:

This is the core of a two-decade-old debate among security people, and it s one that the benevolent God faction has consistently had the upper hand in. They re the curated computing advocates who insist that preventing you from choosing an alternative app store or side-loading a program is for your own good because if it s possible for you to override the manufacturer s wishes, then malicious software may impersonate you to do so, or you might be tricked into doing so. [..] This benevolent dictatorship model only works so long as the dictator is both perfectly benevolent and perfectly competent. We know the dictators aren t always benevolent. [ ] But even if you trust a dictator s benevolence, you can t trust in their perfection. Everyone makes mistakes. Benevolent dictator computing works well, but fails badly. Designing a computer that intentionally can t be fully controlled by its owner is a nightmare, because that is a computer that, once compromised, can attack its owner with impunity.

---

Lastly, Chengyu HAN updated the Reproducible Builds website to correct an incorrect Git command. [ ]

Debian In Debian this month, the essential and required package sets became 100% reproducible in Debian bookworm on the amd64 and arm64 architectures. These two subsets of the full Debian archive refer to Debian package priority levels as described in the 2.5 Priorities section of the Debian Policy there is no canonical minimal installation package set in Debian due to its diverse methods of installation. As it happens, these package sets are not reproducible on the i386 architecture because the ncurses package on that architecture is not yet reproducible, and the sed package currently fails to build from source on armhf too. The full list of reproducible packages within these package sets can be viewed within our QA system, such as on the page of required packages in amd64 and the list of essential packages on arm64, both for Debian bullseye.

---

It recently has become very easy to install reproducible Debian Docker containers using podman on Debian bullseye:

$ sudo apt install podman
$ podman run --rm -it debian:bullseye bash

The (pre-built) image used is itself built using debuerrotype, as explained on docker.debian.net. This page also details how to build the image yourself and what checksums are expected if you do so.

---

Related to this, it has also become straightforward to reproducibly bootstrap Debian using mmdebstrap, a replacement for the usual debootstrap tool to create Debian root filesystems:

$ SOURCE_DATE_EPOCH=$(date --utc --date=2022-08-29 +%s) mmdebstrap unstable > unstable.tar

This works for (at least) Debian unstable, bullseye and bookworm, and is tested automatically by a number of QA jobs set up by Holger Levsen (unstable, bookworm and bullseye)

---

Work has also taken place to ensure that the canonical debootstrap and cdebootstrap tools are also capable of bootstrapping Debian reproducibly, although it currently requires a few extra steps:

1. Clamping the modification time of files that are newer than $SOURCE_DATE_EPOCH to be not greater than SOURCE_DATE_EPOCH.
2. Deleting a few files. For debootstrap, this requires the deletion of /etc/machine-id, /var/cache/ldconfig/aux-cache, /var/log/dpkg.log, /var/log/alternatives.log and /var/log/bootstrap.log, and for cdebootstrap we also need to delete the /var/log/apt/history.log and /var/log/apt/term.log files as well.

This process works at least for unstable, bullseye and bookworm and is now being tested automatically by a number of QA jobs setup by Holger Levsen [ ][ ][ ][ ][ ][ ]. As part of this work, Holger filed two bugs to request a better initialisation of the /etc/machine-id file in both debootstrap [ ] and cdebootstrap [ ].

---

Elsewhere in Debian, 131 reviews of Debian packages were added, 20 were updated and 27 were removed this month, adding to our extensive knowledge about identified issues. Chris Lamb added a number of issue types, including: randomness_in_browserify_output [ ], haskell_abi_hash_differences [ ], nondeterministic_ids_in_html_output_generated_by_python_sphinx_panels [ ]. Lastly, Mattia Rizzolo removed the deterministic flag from the captures_kernel_variant flag [ ].

Other distributions Vagrant Cascadian posted an update of the status of Reproducible Builds in GNU Guix, writing that:

Ignoring the pesky unknown packages, it is more like ~93% reproducible and ~7% unreproducible... that feels a bit better to me! These numbers wander around over time, mostly due to packages moving back into an "unknown" state while the build farms catch up with each other... although the above numbers seem to have been pretty consistent over the last few days.

The post itself contains a lot more details, including a brief discussion of tooling. Elsewhere in GNU Guix, however, Vagrant updated a number of packages such as itpp [ ], perl-class-methodmaker [ ], libnet [ ], directfb [ ] and mm-common [ ], as well as updated the version of reprotest to 0.7.21 [ ]. In openSUSE, Bernhard M. Wiedemann published his usual openSUSE monthly report.

diffoscope diffoscope is our in-depth and content-aware diff utility. Not only can it locate and diagnose reproducibility issues, it can provide human-readable diffs from many kinds of binary formats. This month, Chris Lamb prepared and uploaded versions 220 and 221 to Debian, as well as made the following changes:

• Update external_tools.py to reflect changes to xxd and the vim-common package. [ ]
• Depend on the dedicated xxd package now, not the vim-common package. [ ]
• Don t crash if we can open a PDF file using the PyPDF library, but cannot subsequently parse the annotations within. [ ]

In addition, Vagrant Cascadian updated diffoscope in GNU Guix, first to to version 220 [ ] and later to 221 [ ].

Community news The Reproducible Builds project aims to fix as many currently-unreproducible packages as possible as well as to send all of our patches upstream wherever appropriate. This month we created a number of patches, including:

• Bernhard M. Wiedemann:
  • at-spi-sharp (build failure when build on a multiprocessor machine).
  • borgbackup (fails to build in 2038, fix)
  • buzztrax (parallelism-related issue)
  • chart-testing (date-related issue)
  • memcached (fails to build in 2038)
  • nim (fails to build in 2038)
  • perl-Time-Moment (fails to build in 2038)
  • python-bson (fails to build in 2038)
  • python-heatclient (fails to build in 2038)
  • python3.8 (fails to build in 2038)
  • reproducible-faketools
  • s3fs (date-related issue)
  • systemd (date-related issue)
• Chris Lamb:
  • #1016486 filed against wayfire.
  • #1016583 filed against multipath-tools.
  • #1016584 filed against node-canvas-confetti.
  • #1017473 filed against psi (forwarded upstream).
  • #1017475 filed against sphinx-panels (forwarded upstream).
  • #1017920 filed against sysfsutils.
  • #1017945 filed against geeqie.
• Vagrant Cascadian:
  • #1017073 filed against python-suntime.
  • #1017372 filed against plymouth.
  • #1017373 filed against tiemu.
  • #1017421 filed against fltk1.3.
  • #1018802 filed against localechooser.
  • uboot (Rasmus Villemoes proposed fixing gcc instead [ ][ ])

Testing framework The Reproducible Builds project runs a significant testing framework at tests.reproducible-builds.org, to check packages and other artifacts for reproducibility. This month, Holger Levsen made the following changes:

• Debian-related changes:
  • Temporarily add Debian unstable deb-src lines to enable test builds a Non-maintainer Upload (NMU) campaign targeting 708 sources without .buildinfo files found in Debian unstable, including 475 in bookworm. [ ][ ]
  • Correctly deal with the Debian Edu packages not being installable. [ ]
  • Finally, stop scheduling stretch. [ ]
  • Make sure all Ubuntu nodes have the linux-image-generic kernel package installed. [ ]
• Health checks & view:
  • Detect SSH login problems. [ ]
  • Only report the first uninstallable package set. [ ]
  • Show new bootstrap jobs. [ ] and debian-live jobs. [ ] in the job health view.
  • Fix regular expression to detect various zombie jobs. [ ]
• New jobs:
  • Add a new job to test reproducibility of mmdebstrap bootstrapping tool. [ ][ ][ ][ ]
  • Run our new mmdebstrap job remotely [ ][ ]
  • Improve the output of the mmdebstrap job. [ ][ ][ ]
  • Adjust the mmdebstrap script to additionally support debootstrap as well. [ ][ ][ ]
  • Work around mmdebstrap and debootstrap keeping logfiles within their artifacts. [ ][ ][ ]
  • Add support for testing cdebootstrap too and add such a job for unstable. [ ][ ][ ]
  • Use a reproducible value for SOURCE_DATE_EPOCH for all our new bootstrap jobs. [ ]
• Misc changes:
  • Send the create_meta_pkg_sets notification to #debian-reproducible-changes instead of #debian-reproducible. [ ]

In addition, Roland Clobus re-enabled the tests for live-build images [ ] and added a feature where the build would retry instead of give up when the archive was synced whilst building an ISO [ ], and Vagrant Cascadian added logging to report the current target of the /bin/sh symlink [ ].

Contact As ever, if you are interested in contributing to the Reproducible Builds project, please visit our Contribute page on our website. However, you can get in touch with us via:

• IRC: #reproducible-builds on irc.oftc.net.
• Twitter: @ReproBuilds
• Mailing list: rb-general@lists.reproducible-builds.org

Aug 8 04:04] list_del corruption. prev->next should be ffff90c96e9c2090,
but was ffff90c94e9c2090