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

• The SG2000 SoC is a Dual-architecture beast. A hardware switch controls whether the CPU is an ARM or a RISC-V.
  • Not only that: It has a second (albeit lesser) RISC-V core that can run independently. They mention this computer can run simultaneously Linux and FreeRTOS!
• 512MB RAM
• Sweet form factor (4.2 4.2cm)
• Peeking around their Web site, it is one of the most open and well documented computers in their hardware range.
  • Schematics at different levels of detail
  • Datasheet. It s preliminary, and it s written in Mandarin, but the information it contains is quite descriptive and clear.
  • Full specifications (all 710 pages of it!) Again, it has a lot of Mandarin in it, but it conveys a lot of useful information.

• Milk- Duo S booting with Buildroot (24K)
• Milk- Duo S booting with Debian (49K)

A Case Study This post could have been written any time in the past well, decade or so, really. But the trigger for my sitting down and writing this post is the recent breach of wallet-finding and criminal-harassment-enablement platform Tile. As reported by Engadget, a statement attributed to Life360 CEO Chris Hulls says

The potentially impacted data consists of information such as names, addresses, email addresses, phone numbers, and Tile device identification numbers.

But don t worry though; even though your home address is now public information

It does not include more sensitive information, such as credit card numbers

Aaaaaand here is where I get salty.

Why Credit Card Numbers Don t Matter Describing credit card numbers as more sensitive information is somewhere between disingenuous and a flat-out lie. It was probably included in the statement because it s part of the standard playbook. Why is it part of the playbook, though? Not being a disaster comms specialist, I can t say for sure, but my hunch is that the post-breach playbook includes this line because (a) credit cards are less commonly breached these days (more on that later), and (b) it s a way to insinuate that all your financial data is safe, no need to worry without having to say that (because that statement would absolutely be a lie). The thing that not nearly enough people realise about credit card numbers is:

1. The credit card holder is not usually liable for most fraud done via credit card numbers; and
2. In terms of actual, long-term damage to individuals, credit card fraud barely rates a mention. Identity fraud, Business Email Compromise, extortion, and all manner of other unpleasantness is far more damaging to individuals.

Why Credit Card Numbers Do Matter Losing credit card numbers in a data breach is a huge deal but not for the users of the breached platform. Instead, it s a problem for the company that got breached. See, going back some years now, there was a wave of huge credit card data breaches. If you ve been around a while, names like Target and Heartland will bring back some memories. Because these breaches cost issuing banks and card brands a lot of money, the Payment Card Industry Security Standards Council (PCI-SSC) and the rest of the ecosystem went full goblin mode. Now, if you lose credit card numbers in bulk, it will cost you big. Massive fines for breaches (typically levied by the card brands via the acquiring bank), increased transaction fees, and even the Credit Card Death Penalty (being banned from charging credit cards), are all very big sticks.

Now Comes the Finding Out In news that should not be surprising, when there are actual consequences for failing to do something, companies take the problem seriously. Which is why no credit card numbers were disclosed is such an interesting statement. Consider why no credit card numbers were disclosed. It s not that credit card numbers aren t valuable to criminals because they are. Instead, it s because the company took steps to properly secure the credit card data. Next, you ll start to consider why, if the credit card numbers were secured, why wasn t the personal information that did get disclosed similarly secured? Information that is far more damaging to the individuals to whom that information relates than credit card numbers. The only logical answer is that it wasn t deemed financially beneficial to the company to secure that data. The consequences of disclosure for that information isn t felt by the company which was breached. Instead, it s felt by the individuals who have to spend weeks of their life cleaning up from identity fraud committed against them. It s felt by the victim of intimate partner violence whose new address is found in a data dump, letting their ex find them again. Until there are real, actual consequences for the companies which hemorrhage our personal data (preferably ones that have percentage of global revenue at the end), data breaches will continue to happen. Not because they re inevitable because as credit card numbers show, data can be secured but because there s no incentive for companies to prevent our personal data from being handed over to whoever comes along.

Support my Salt My salty takes are powered by refreshing beverages. If you d like to see more of the same, buy me one.

Debian LTS contributors In May, 17 contributors have been paid to work on Debian LTS, their reports are available:

• Adrian Bunk did 34.25h (out of 24.0h assigned and 22.0h from previous period), thus carrying over 11.75h to the next month.
• Bastien Roucari s did 20.0h (out of 20.0h assigned).
• Ben Hutchings did 16.0h (out of 24.0h assigned), thus carrying over 8.0h to the next month.
• Chris Lamb did 18.0h (out of 18.0h assigned).
• Daniel Leidert did 8.0h (out of 10.0h assigned), thus carrying over 2.0h to the next month.
• Emilio Pozuelo Monfort did 35.5h (out of 46.0h assigned), thus carrying over 10.5h to the next month.
• Guilhem Moulin did 13.0h (out of 14.75h assigned and 5.25h from previous period), thus carrying over 7.0h to the next month.
• Lee Garrett did 11.0h (out of 37.25h assigned and 8.75h from previous period), thus carrying over 35.0h to the next month.
• Lucas Kanashiro did 10.0h (out of 20.0h assigned), thus carrying over 10.0h to the next month.
• Markus Koschany did 40.0h (out of 40.0h assigned).
• Ola Lundqvist did 6.5h (out of 22.5h assigned and 1.5h from previous period), thus carrying over 17.5h to the next month.
• Roberto C. S nchez did 7.75h (out of 11.0h assigned and 1.0h from previous period), thus carrying over 4.25h to the next month.
• Santiago Ruano Rinc n did 8.0h (out of 16.0h assigned), thus carrying over 8.0h to the next month.
• Sean Whitton did 5.5h (out of 5.5h assigned and 0.5h from previous period), thus carrying over 0.5h to the next month.
• Sylvain Beucler did 10.5h (out of 0.75h assigned and 45.25h from previous period), thus carrying over 35.5h to the next month.
• Thorsten Alteholz did 14.0h (out of 14.0h assigned).
• Tobias Frost did 7.75h (out of 10.0h assigned and 2.0h from previous period), thus carrying over 4.25h to the next month.

Evolution of the situation In May, we have released 20 DLAs. Notable security updates in May included:

• apache2: multiple vulnerabilities which may result in HTTP response splitting, denial of service, or authorization bypass (by Bastien Roucari s, in collaboration with apache2 maintainer Yadd)
• bind9: two vulnerabilities, called KeyTrap and NSEC3, which may result in denial of service (by Santiago Ruano Rinc n)
• python-pymysql: potential SQL injection attack (by Chris Lamb)

The aforementioned apache2 was prepared by its Debian maintainer Yadd. This update also involved work on the package test suite in buster, which contributor Bastien Roucari s then forwarded to the apache2 package in unstable. More importantly, a regression in fossil was reported, and Bastien prepared a fix for it. Bastien coordinated the upload of both packages to minimize the introduction of regressions. Contributor Daniel Leidert also prepared an upload of runc to Debian 11 in order fix a number of CVEs still affecting that package. Finally, contributor Thorsten Alteholz prepared uploads for qtbase-opensource-src, libjwt, and libmicrohttpd in Debian 11. Note that Debian 11 will pass into the LTS phase of support in August and these updates will improve the state and long-term supportability of Debian 11. Debian 10 is presently in its final month of LTS support (as announced on the debian-lts-announce mailing list, support will end on June 30th), after which no new security updates will be made available on security.debian.org. However, Freexian and its team of paid Debian contributors will continue to maintain Debian 10 going forward for the customers of the Extended LTS offer. Subscribe right away if you sill have Debian 10 which must be kept secure (and which cannot yet be upgraded).

Thanks to our sponsors Sponsors that joined recently are in bold.

• Platinum sponsors:
  • TOSHIBA (for 105 months)
  • Civil Infrastructure Platform (CIP) (for 73 months)
  • VyOS Inc (for 37 months)
• Gold sponsors:
  • Roche Diagnostics International AG (for 116 months)
  • Linode (for 110 months)
  • Babiel GmbH (for 99 months)
  • Plat Home (for 99 months)
  • CINECA (for 73 months)
  • University of Oxford (for 55 months)
  • Deveryware (for 42 months)
  • EDF SA (for 26 months)
  • CERN
• Silver sponsors:
  • Domeneshop AS (for 120 months)
  • Nantes M tropole (for 115 months)
  • Univention GmbH (for 106 months)
  • Universit Jean Monnet de St Etienne (for 106 months)
  • Ribbon Communications, Inc. (for 100 months)
  • Exonet B.V. (for 90 months)
  • Leibniz Rechenzentrum (for 84 months)
  • Minist re de l Europe et des Affaires trang res (for 67 months)
  • Cloudways by DigitalOcean (for 57 months)
  • Dinahosting SL (for 55 months)
  • Bauer Xcel Media Deutschland KG (for 49 months)
  • Platform.sh SAS (for 49 months)
  • Moxa Inc. (for 43 months)
  • sipgate GmbH (for 40 months)
  • OVH US LLC (for 38 months)
  • Tilburg University (for 38 months)
  • GSI Helmholtzzentrum f r Schwerionenforschung GmbH (for 30 months)
  • Soliton Systems K.K. (for 27 months)
  • THINline s.r.o.
• Bronze sponsors:
  • Evolix (for 121 months)
  • Seznam.cz, a.s. (for 121 months)
  • Intevation GmbH (for 118 months)
  • Linuxhotel GmbH (for 118 months)
  • Daevel SARL (for 116 months)
  • Bitfolk LTD (for 115 months)
  • Megaspace Internet Services GmbH (for 115 months)
  • NUMLOG (for 115 months)
  • Greenbone AG (for 114 months)
  • WinGo AG (for 114 months)
  • Ecole Centrale de Nantes - LHEEA (for 110 months)
  • Entr ouvert (for 105 months)
  • Adfinis AG (for 102 months)
  • GNI MEDIA (for 97 months)
  • Laboratoire LEGI - UMR 5519 / CNRS (for 97 months)
  • Tesorion (for 97 months)
  • Bearstech (for 88 months)
  • LiHAS (for 88 months)
  • Catalyst IT Ltd (for 83 months)
  • Supagro (for 78 months)
  • Demarcq SAS (for 77 months)
  • Universit Grenoble Alpes (for 63 months)
  • TouchWeb SAS (for 55 months)
  • SPiN AG (for 52 months)
  • CoreFiling (for 47 months)
  • Institut des sciences cognitives Marc Jeannerod (for 42 months)
  • Observatoire des Sciences de l Univers de Grenoble (for 39 months)
  • Tem Innovations GmbH (for 34 months)
  • WordFinder.pro (for 33 months)
  • CNRS DT INSU R sif (for 32 months)
  • Alter Way (for 25 months)
  • Institut Camille Jordan (for 14 months)

• [#1070154] bullseye-pu: qtbase-opensource-src/5.15.2+dfsg-9+deb11u1 package upload
• [#1064550] bullseye-pu: libjwt 1.10.2-1+deb11u1 has been marked for accept
• [#1067544] bullseye-pu: libmicrohttpd 0.9.72-2+deb11u1 has been marked for accept

• [ELA-1104-1-1]nghttp2 security update for one CVEs to fix an DoS resulting from bad handling of CONTINUATION frames in Stretch

• cups-bjnp

• astap
• astap-cli
• c-munipack
• libahp-xc
• planetary-system-stacker
• fxload
• virtualgps

• pywws
• openzwave
• openoverlayrouter

• dateutils
• chktex
• liburjtag
• pollinate
• usb-modeswitch
• greylistd
• entropybroker

1. A peek into build provenance for Homebrew
2. Distribution news
3. Mailing list news
4. Miscellaneous news
5. Two new academic papers
6. diffoscope
7. Website updates
8. Upstream patches
9. Reproducibility testing framework

A peek into build provenance for Homebrew Joe Sweeney and William Woodruff on the Trail of Bits blog wrote an extensive post about build provenance for Homebrew, the third-party package manager for MacOS. Their post details how each bottle (i.e. each release):

[ ] built by Homebrew will come with a cryptographically verifiable statement binding the bottle s content to the specific workflow and other build-time metadata that produced it. [ ] In effect, this injects greater transparency into the Homebrew build process, and diminishes the threat posed by a compromised or malicious insider by making it impossible to trick ordinary users into installing non-CI-built bottles.

The post also briefly touches on future work, including work on source provenance:

Homebrew s formulae already hash-pin their source artifacts, but we can go a step further and additionally assert that source artifacts are produced by the repository (or other signing identity) that s latent in their URL or otherwise embedded into the formula specification.

Distribution news In Debian this month, Johannes Schauer Marin Rodrigues (aka josch) noticed that the Debian binary package bash version 5.2.15-2+b3 was uploaded to the archive twice. Once to bookworm and once to sid but with differing content. This is problem for reproducible builds in Debian due its assumption that the package name, version and architecture triplet is unique. However, josch highlighted that

This example with bash is especially problematic since bash is Essential:yes, so there will now be a large portion of .buildinfo files where it is not possible to figure out with which of the two differing bash packages the sources were compiled.

In response to this, Holger Levsen performed an analysis of all .buildinfo files and found that this needs almost 1,500 binNMUs to fix the fallout from this bug. Elsewhere in Debian, Vagrant Cascadian posted about a Non-Maintainer Upload (NMU) sprint to take place during early June, and it was announced that there is now a #debian-snapshot IRC channel on OFTC to discuss the creation of a new source code archiving service to, perhaps, replace snapshot.debian.org. Lastly, 11 reviews of Debian packages were added, 15 were updated and 48 were removed this month adding to our extensive knowledge about identified issues. A number of issue types have been updated by Chris Lamb as well. [ ][ ]
Elsewhere in the world of distributions, deep within a larger announcement from Colin Percival about the release of version 14.1-BETA2, it was mentioned that the FreeBSD kernels are now built reproducibly.
In Fedora, however, the change proposal mentioned in our report for April 2024 was approved, so, per the ReproduciblePackageBuilds wiki page, the add-determinism tool is now running in new builds for Fedora 41 ( rawhide ). The add-determinism tool is a Rust program which, as its name suggests, adds determinism to files that are given as input by attempting to standardize metadata contained in binary or source files to ensure consistency and clamping to $SOURCE_DATE_EPOCH in all instances . This is essentially the Fedora version of Debian s strip-nondeterminism. However, strip-nondeterminism is written in Perl, and Fedora did not want to pull Perl in the buildroot for every package. The add-determinism tool eliminates many causes of non-determinism and work is ongoing to continue the scope of packages it can operate on.

Mailing list news On our mailing list this month, regular contributor kpcyrd wrote to the list with an update on their source code indexing project, whatsrc.org. The whatsrc.org project, which was launched last month in response to the XZ Utils backdoor, now contains and indexes almost 250,000 unique source code archives. In their post, kpcyrd gives an example of its intended purpose, noting that it shown that whilst there seems to be consensus about [the] source code for zsh 5.9 in various Linux distributions, it does not align with the contents of the zsh Git repository . Holger Levsen also posted to the list with a pre-announcement of sorts for the 2024 Reproducible Builds summit. In particular:

[Whilst] the dates and location are not fixed yet, however if you don help us with finding a suitable location soon, it is very likely that we ll meet again in Hamburg in the 2nd half of September 2024 [ ].

Lastly, Frederic-Emmanuel Picca wrote to the list asking for help understanding the non-reproducible status of the Debian silx package and received replies from both Vagrant Cascadian and Chris Lamb.

Miscellaneous news strip-nondeterminism is our tool to remove specific non-deterministic results from a completed build. This month strip-nondeterminism version 1.14.0-1 was uploaded to Debian unstable by Chris Lamb chiefly to incorporate a change from Alex Muntada to avoid a dependency on Sub::Override to perform monkey-patching and break circular dependencies related to debhelper [ ]. Elsewhere in our tooling, Jelle van der Waa modified reprotest because the pipes module will be removed in Python version 3.13 [ ].
It was also noticed that a new blog post by Daniel Stenberg detailing How to verify a Curl release mentions the SOURCE_DATE_EPOCH environment variable. This is because:

The [curl] release tools document also contains another key component: the exact time stamp at which the release was done using integer second resolution. In order to generate a correct tarball clone, you need to also generate the new version using the old version s timestamp. Because the modification date of all files in the produced tarball will be set to this timestamp.

Furthermore, Fay Stegerman filed a bug against the Signal messenger app for Android to report that their reproducible builds cannot, in fact, be reproduced. However, Fay is quick to note that she has:

found zero evidence of any kind of compromise. Some differences are yet unexplained but everything I found seems to be benign. I am disappointed that Reproducible Builds have been broken for months but I have zero reason to doubt Signal s security in any way.

Lastly, it was observed that there was a concise and diagrammatic overview of supply chain threats on the SLSA website.

Two new academic papers Two new scholarly papers were published this month. Firstly, Mathieu Acher, Beno t Combemale, Georges Aaron Randrianaina and Jean-Marc J z quel of University of Rennes on Embracing Deep Variability For Reproducibility & Replicability. The authors describe their approach as follows:

In this short [vision] paper we delve into the application of software engineering techniques, specifically variability management, to systematically identify and explicit points of variability that may give rise to reproducibility issues (e.g., language, libraries, compiler, virtual machine, OS, environment variables, etc.). The primary objectives are: i) gaining insights into the variability layers and their possible interactions, ii) capturing and documenting configurations for the sake of reproducibility, and iii) exploring diverse configurations to replicate, and hence validate and ensure the robustness of results. By adopting these methodologies, we aim to address the complexities associated with reproducibility and replicability in modern software systems and environments, facilitating a more comprehensive and nuanced perspective on these critical aspects.

(A PDF of this article is available.)
Secondly, Ludovic Court s, Timothy Sample, Simon Tournier and Stefano Zacchiroli have collaborated to publish a paper on Source Code Archiving to the Rescue of Reproducible Deployment. Their paper was motivated because:

The ability to verify research results and to experiment with methodologies are core tenets of science. As research results are increasingly the outcome of computational processes, software plays a central role. GNU Guix is a software deployment tool that supports reproducible software deployment, making it a foundation for computational research workflows. To achieve reproducibility, we must first ensure the source code of software packages Guix deploys remains available.

(A PDF of this article is also available.)

diffoscope diffoscope is our in-depth and content-aware diff utility that can locate and diagnose reproducibility issues. This month, Chris Lamb made a number of changes such as uploading versions 266, 267, 268 and 269 to Debian, making the following changes:

• New features:
  • Use xz --list to supplement output when comparing .xz archives; essential when metadata differs. (#1069329)
  • Include xz --verbose --verbose (ie. double) output. (#1069329)
  • Strip the first line from the xz --list output. [ ]
  • Only include xz --list --verbose output if the xz has no other differences. [ ]
  • Actually append the xz --list after the container differences, as it simplifies a lot. [ ]
• Testing improvements:
  • Allow Debian testing to fail right now. [ ]
  • Drop apktool from Build-Depends; we can still test APK functionality via autopkgtests. (#1071410)
  • Add a versioned dependency for at least version 5.4.5 for the xz tests as they fail under (at least) version 5.2.8. (#374)
  • Fix tests for 7zip 24.05. [ ][ ]
  • Fix all tests after additon of xz --list. [ ][ ]
• Misc:
  • Update copyright years. [ ]

In addition, James Addison fixed an issue where the HTML output showed only the first difference in a file, while the text output shows all differences [ ][ ][ ], Sergei Trofimovich amended the 7zip version test for older 7z versions that include the string [64] [ ][ ] and Vagrant Cascadian relaxed the versioned dependency to allow version 5.4.1 for the xz tests [ ] and proposed updates to guix for versions 267, 268 and pushed version 269 to Guix. Furthermore, Eli Schwartz updated the diffoscope.org website in order to explain how to install diffoscope on Gentoo [ ].

Website updates There were a number of improvements made to our website this month, including Chris Lamb making the print CSS stylesheet nicer [ ]. Fay Stegerman made a number of updates to the page about the SOURCE_DATE_EPOCH environment variable [ ][ ][ ] and Holger Levsen added some of their presentations to the Resources page. Furthermore, IOhannes zm lnig stipulated support for SOURCE_DATE_EPOCH in clang version 16.0.0+ [ ], Jan Zerebecki expanded the Formal definition page and fixed a number of typos on the Buy-in page [ ] and Simon Josefsson fixed the link to Trisquel GNU/Linux on the Projects page [ ].

Upstream patches This month, we wrote a number of patches to fix specific reproducibility issues, including:

• Bernhard M. Wiedemann:
  • nauty (CPU-detection issue)
  • emacs (ASLR)
• Chris Lamb:
  • #1070754 filed against gensio.
  • #1071064 filed against tkgate.
  • #1072094 filed against ruby-pgplot.

Reproducibility testing framework The Reproducible Builds project operates a comprehensive testing framework running primarily at tests.reproducible-builds.org in order to check packages and other artifacts for reproducibility. In May, a number of changes were made by Holger Levsen:

• Debian-related changes:
  • Enable the rebuilder-snapshot API on osuosl4. [ ]
  • Schedule the i386 architecture a bit more often. [ ]
  • Adapt cleanup_nodes.sh to the new way of running our build services. [ ]
  • Add 8 more workers for the i386 architecture. [ ]
  • Update configuration now that the infom07 and infom08 nodes have been reinstalled as real i386 systems. [ ]
  • Make diffoscope timeouts more visible on the #debian-reproducible-changes IRC channel. [ ]
  • Mark the cbxi4a-armhf node as down. [ ][ ]
  • Only install the hdmi2usb-mode-switch package only on Debian bookworm and earlier [ ] and only install the haskell-platform package on Debian bullseye [ ].
• Misc:
  • Install the ntpdate utility as we need it later. [ ]
  • Document the progress on the i386 architecture nodes at Infomaniak. [ ]
  • Drop an outdated and unnoticed notice. [ ]
  • Add live_setup_schroot to the list of so-called zombie jobs. [ ]

In addition, Mattia Rizzolo reinstalled the infom07 and infom08 nodes [ ] and Vagrant Cascadian marked the cbxi4a node as online [ ].

---

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
• Mastodon: @reproducible_builds@fosstodon.org
• Mailing list: rb-general@lists.reproducible-builds.org

DebConf Bursary updates, by Utkarsh Gupta Utkarsh is the bursaries team lead for DebConf 24. Bursary requests are dispatched to a team of volunteers to review. The results are collated, adjusted and merged to produce priority lists of requests to fund. Utkarsh raised the team, coordinated the review, and issued bursaries to attendees.

/usr-move, by Helmut Grohne More and more, the /usr-move transition is being carried out by multiple contributors and many performed around a hundred of the requested uploads. Of these, Helmut contributed five patches and two uploads. As a result, there are less than 350 packages left to be converted, and all of the non-trivial cases have patches. We started with three times that number. Thanks to everyone involved for supporting this effort. For people interested in background information of this transition, Helmut gave a presentation at MiniDebConf Berlin 2024 (slides).

sbuild, by Helmut Grohne While unshare mode of sbuild has existed for quite a while, it is now getting significant use in Debian, and new problems are popping up. Helmut looked into an apparmor-related failure and provided a diagnosis. While relevant code would detect the chroot nature of a schroot backend and skip apparmor tests, the unshare environment would be just good enough to run and fail the test. As sbuild exposes fewer special kernel filesystems, the tests will be skipped again. Another problem popped up when gobject-introspection added a dependency on the host architecture Python interpreter in a cross build environment. sbuild would prefer installing (and failing) a host architecture Python to installing the qemu alternative. Attempts to fix this would result in systemd killing sbuild. ischroot as used by libc6.postinst would not classify the unshare environment as a chroot. Therefore libc6.postinst would run telinit which would kill the build process. This is a complex interaction problem that shall eventually be solved by providing triggers from libc6 to be implemented by affected init systems.

Salsa CI updates, by Santiago Ruano Rinc n Several issues arose about Salsa CI last month, and it is probably worth mentioning part of the challenges of defining its framework in YAML. With the upcoming end-of-support of Debian 10 buster as LTS, armel was removed from deb.debian.org, making the jobs that build images for buster/armel to fail. While the removal of buster/armel from the repositories is a natural change, it put some light on the flaws in the Salsa CI design regarding the support of the different Debian releases. Currently, the images are defined like these (from .images-debian.yml):

.all-supported-releases: &all-supported-releases
  - stretch
  - stretch-backports
  - buster
  - bullseye
  - bullseye-backports
  - bookworm
  - bookworm-backports
  - trixie
  - sid
  - experimental

And from them, different images are built according to the different jobs and how they are supported, for example:

images-prod-arm:
  stage: build
  extends: .build_template
  tags:
    - $SALSA_CI_ARM_RUNNER_TAG
  parallel:
    matrix:
      # Base image, all releases, all arches
      - IMAGE_NAME: base
        ARCH:
          - arm32v5
          - arm32v7
          - arm64v8
        RELEASE: *all-supported-releases

The removal of buster/armel could be easily reflected as:

images-prod-arm:
  stage: build
  extends: .build_template
  tags:
    - $SALSA_CI_ARM_RUNNER_TAG
  parallel:
    matrix:
      # Base image, fully supported releases, all arches
      - IMAGE_NAME: base
        ARCH:
          - arm32v5
          - arm32v7
          - arm64v8
        RELEASE:
          - stretch
          - buster
          - bullseye
          - bullseye-backports
          - bookworm
          - bookworm-backports
          - trixie
          - sid
          - experimental
      # buster only supports armhf and arm64
      - IMAGE_NAME: base
        ARCH:
          - arm32v7
          - arm64v8
        RELEASE: buster

Evidently, this increases duplication of the release support data, which is of course not optimal and it is error prone when changing the data about supported releases. A better approach would be to have two different YAML lists, such as:

# releases that have partial support. E.g.: buster is transitioning to
# Debian LTS, and buster armel is no longer found in deb.debian.org
.old-releases: &old-releases
  - stretch
  - buster

.currently-supported-releases: &currently-supported-releases
  - bullseye
  - bullseye-backports
  - bookworm
  - bookworm-backports
  - trixie
  - sid
  - experimental

and then a unified list:

.all-supported-releases: &all-supported-releases
  - *old-releases
  - *currently-supported-releases

that could be used in the matrix of the jobs that build all the images available in the pipeline container registry. However, due to limitations in GitLab, it is not possible to expand the variables or mapping values in a parallel:matrix context. At least not in an elegant fashion. This is the kind of issue that recently arose and that Santiago is currently working to solve, in the simplest possible way. Astute readers would notice that stretch is listed in the fully supported releases. And there is no problem with stretch, because it is built from archive.debian.org. Otto actually has tried to fix the broken image build job doing the same, but it is still incorrect, because the security repository is not (yet) available in archive.debian.org. Additionally, Santiago has also worked on other merge requests, such as:

1. support branch/tags as target head in the test projects,
2. build autopkgtest image on top of stable
3. Add .yamllint and make it happy in the autopkgtest-lxc project
4. enable FF_SCRIPT_SECTIONS to log multiline commands, among others.

Archiving DebConf Websites, by Stefano Rivera DebConf, the annual Debian conference, has its own new website every year. These are typically complex dynamic web applications (featuring registration, call for papers, scheduling, etc.) Once the conference is over, there is no need to keep maintaining these applications, so we archive the sites off as static HTML, and serve them from Debian s static CDN. Stefano archived the websites for the last two DebConfs. The schedule system behind DebConf 14 and 15 s websites was a derivative of Canonical s summit system. This was only used for a couple of years before migrating to wafer, the current system. Archiving summit content has been on the nice to have list for years, but nobody has ever tackled it. The machine that served the sites went away a couple of years ago. After much digging, a backup of the database was found, and Stefano got this code running on an ancient Python 2.7. Recently Stefano put this all together and hooked in an archive export to finally get this content preserved.

Python 3.x and pypy3 security bug triage, by Stefano Rivera Stefano Rivera triaged all the open security bugs against the Python 3.x and PyPy3 packages for Debian s stable and LTS releases. Several had been fixed but this wasn t recorded in the security tracker.

Linux livepatching support for Debian, by Santiago Ruano Rinc n In collaboration with Emmanuel Arias, Santiago filed ITP bug #1070494. As stated in the bug, more than an Intent to Package, it is an Intent to Design and Implement live patching support for the Linux kernel in Debian. For now, Emmanuel and Santiago have done exploratory work and they are working to understand the different possibilities to implement livepatching. One possible direction is to rely on kpatch, and the other is to package the modules using regular packaging tools. Also, it is needed to evaluate if it is possible to rely on distributing the modules via packages, or instead as a service, as it is done by some commercial distributions.

Miscellaneous contributions

• Thorsten Alteholz uploaded cups-bjnp to improve packaging.
• Colin Watson tracked down a baffling CI issue in openssh to unblock several merge requests, removed the user_readenv=1 option from its PAM configuration, and started on the first stage of his plan to split out GSS-API key exchange support to separate packages.
• Colin did his usual routine work on the Python team, upgrading 26 packages to new upstream versions, and cherry-picking an upstream PR to fix a pytest 8 incompatibility in ipywidgets.
• Colin NMUed a couple of packages to reduce the need for explicit overrides in Packages-arch-specific, and removed some other obsolete entries from there.
• Emilio managed various library transitions, and helped finish a few of the remaining t64 transitions.
• Helmut sent a patch for enabling piuparts to work as a regular user building on earlier work.
• Helmut sent patches for 7 cross build failures, 6 other debian bugs and fixed an infrastructure problem in crossqa.debian.net.
• Nicholas worked on a sponsored package upload, and discovered the blhc tool for diagnosing build hardening.
• Stefano Rivera started and completed the re2 transition. The release team suggested moving to a virtual package scheme that includes the absl ABI (as re2 now depends on it). Adopted this.
• Stefano continued to work on DebConf 24 planning.
• Santiago continued to work on DebConf24 Content tasks as well as Debconf25 organisation.

• I did a little work on improving debbugs autopkgtest status.
• openssh:
  • I fixed an OpenSSL version mismatch error in openssh-ssh1.
  • I finally tracked down a baffling CI issue in openssh, unblocking several contributed merge requests that I d been sitting on until I could get CI to pass for them. (Special thanks to Andreas Hasenack; GSS-API integration tests will make my life much easier.)
  • I removed the user_readenv=1 option from openssh s PAM configuration, and did some work on release notes to document this change for affected users.
  • I started work on the first stage of my plan to split out GSS-API key exchange support to separate packages.
• Python team:
  • I upgraded bitstruct, flufl.enum, flufl.testing, gunicorn, langtable, psycopg3, pygresql, pylint-flask, python-click-didyoumean, python-gssapi, python-httplib2, python-json-log-formatter, python-persistent, python-pgspecial, python-pyld, python-repoze.tm2, python-serializable, python-tenacity, python-typing-extensions, python-unidiff, responses, shortuuid (including an upstream packaging tweak), sqlparse, vulture, zc.lockfile, and zope.interface to new upstream versions.
  • I cherry-picked an upstream PR to fix a pytest 8 incompatibility in ipywidgets.
• I decided that fixing my old troffcvt package to support groff 1.23.0 wasn t worth the time investment, and filed a removal request instead.
• I NMUed bidentd and linuxtv-dvb-apps to declare Architecture: linux-any (and in the latter case also to fix a build failure due to 64-bit time), and worked with the buildd team to remove several of the other remaining entries from Packages-arch-specific.

# passwd jacob

Changing Root s Password with Sudo This one is probably obvious, but if you have a user with the ability to use sudo, then you can change root s password without access to the root account by running:

$ sudo passwd

which will reset the password for the root account without requiring the existing password.

Boot Directly to a Shell Getting root access to any Linux machine you have physical access to is surprisingly simple. You can just boot the machine directly into a root shell without any access control, i.e. passwords.

Why You Should Always Encrypt Your Storage¹ To boot directly to a shell you need to append the following text to the kernel command line:

init=/bin/sh

(You could use pretty much any program here, but you re putting your system into a weird state doing this, and so I d recommend the simplest approach.)

GRUB GRUB will allow you to edit boot parameters on startup using the e key. You ll then be presented with a editor² that you can use to change the kernel command line by appending to the linux line. E.g. If your editor looks like this:

        load_video
        insmod gzio
        if [ x$grub_platform = xxen ]; then insmod xzio; insmod lzopio; fi
        insmod part_gpt
        insmod ext2
        search --no-floppy --fs-uuid --set=root abcd1234-5678-0910-1112-abcd12345678
        echo    'Loading Linux 6.1.0-21-amd64 ...'
        linux   /boot/vmlinuz-6.1.0-21-amd64 root=UUID=abcd1234-5678-0910-1112-abcd12345678 ro  quiet
        echo    'Loading initial ramdisk ...'
        initrd  /boot/initrd.img-6.1.0-21-amd64

Then you would add init=/bin/sh like so:

        load_video
        insmod gzio
        if [ x$grub_platform = xxen ]; then insmod xzio; insmod lzopio; fi
        insmod part_gpt
        insmod ext2
        search --no-floppy --fs-uuid --set=root abcd1234-5678-0910-1112-abcd12345678
        echo    'Loading Linux 6.1.0-21-amd64 ...'
        linux   /boot/vmlinuz-6.1.0-21-amd64 root=UUID=abcd1234-5678-0910-1112-abcd12345678 ro  quiet init=/bin/sh
        echo    'Loading initial ramdisk ...'
        initrd  /boot/initrd.img-6.1.0-21-amd64

Once you ve edited it you can start your machine with Ctrl+x, as you can see from the prompt text under the editor.

Raspberry Pi cmdline.txt On a Raspberry Pi you ll want to append the above to only line of the cmdline.txt file on the boot partition of the SD card. This is the first partition of the disk, the one that is FAT32. You ll need to do this on another machine, since if you had root access to edit cmdline.txt you could also just change your password. As it is a FAT32 partition on an SD card, it should be editable on any other machine that supports SD cards. E.g. If your cmdline.txt looks like this

console=serial0,115200 console=tty1 root=PARTUUID=fb33757d-02 rootfstype=ext4 fsck.repair=yes rootwait quiet

Then you would add init=/bin/sh like so:

console=serial0,115200 console=tty1 root=PARTUUID=fb33757d-02 rootfstype=ext4 fsck.repair=yes rootwait quiet init=/bin/sh

Mount Read / Write Since you re replacing the init process of the machine with a shell, no other processes will be running. Also, your root filesystem will be mounted read-only, as init is expected to remount it read-write as needed.

# mount -o remount,rw /

Change Root Password Once you ve remounted the root filesystem, all that s needed is to run the passwd command.

# passwd

Since you re running the command as root you won t need to provide your existing password, and will only need to type a new password twice. Now of course you simply need to remember that password in order to ensure you don t need to do this again.

Reboot Safely You now cannot follow the standard reboot process here, as you re only running one process. Therefore it s important to put your root filesystem back into read-only before powering off your machine:

# mount -o remount,ro /

Once you ve done that you just need to hold down the power button until the machine completely powers off or pull the plug. And then you re done! Boot the computer again and you ll have everything working as normal, with a root password you remember.

1. Not that this is the only reason, anyone with physical access to your machine could also boot it into another operating system they control, or just remove your storage device and put it into another computer, or probably other things I m not thinking of now. You should always encrypt your devices.
2. The editor uses emacs-like keybindings. The manual includes a list of all the options available.

• [1] https://tinyurl.com/yu2fk9gd
• [2] https://boehs.org/node/npm-everything
• [3] https://tinyurl.com/25oowm6n
• [4] https://joeyh.name/blog/entry/the_vulture_in_the_coal_mine/
• [5] https://tinyurl.com/2bp9ysrj
• [6] https://tinyurl.com/2bqoewzr
• [7] https://tinyurl.com/2yz3qrvu
• [8] https://tinyurl.com/y8gmut3n
• [9] https://tinyurl.com/27kn8a3y
• [10] https://lutrasecurity.com/en/articles/kobold-letters/

A Modest Proposal So, here s my thought. What if instead of the repo tabs looking like the above, they instead looked like this: My conception of this is that it would, essentially, be a kind of yearbook , that people who used and liked the software could scribble their thoughts on. With some fairly straightforward affordances elsewhere to encourage its use, it could be a powerful way to show maintainers that they are, in fact, valued and appreciated. There are a number of software packages I ve used recently, that I d really like to say a general thanks, this is awesome! to. However, I m not about to make the Issues tab look even scarier by creating an issue to say thanks, and digging up an email address is often surprisingly difficult, and wouldn t be a public show of my gratitude, which I believe is a valuable part of the interaction.

You Can t Pay Your Rent With Kudos Absolutely you cannot. A means of expressing appreciation in no way replaces the pressing need to figure out a way to allow open source developers to pay their rent. Conversely, however, the need to pay open source developers doesn t remove the need to also show those people that their work is appreciated and valued by many people around the world. Anyway, who knows a senior exec at GitHub? I ve got an idea I d like to run past them

Commit Mono This time I seem to be settling on either Commit Mono or Space Mono. For now I'm using Commit Mono because it's a little more compressed than Fira and does have a italic version. I don't like how Space Mono's parenthesis (()) is "squarish", it feels visually ambiguous with the square brackets ([]), a big no-no for my primary use case (code). So here I am using a new font, again. It required changing a bunch of configuration files in my home directory (which is in a private repository, sorry) and Emacs configuration (thankfully that's public!). One gotcha is I realized I didn't actually have a global font configuration in Emacs, as some Faces define their own font family, which overrides the frame defaults. This is what it looks like, before:

Fira Mono

After:

Commit Mono

(Notice how those screenshots are not sharp? I'm surprised too. The originals look sharp on my display, I suspect this is something to do with the Wayland transition. I've tried with both grim and flameshot, for what its worth.) They are pretty similar! Commit Mono feels a bit more vertically compressed maybe too much so, actually -- the line height feels too low. But it's heavily customizable so that's something that's relatively easy to fix, if it's really a problem. Its weight is also a little heavier and wider than Fira which I find a little distracting right now, but maybe I'll get used to it. All characters seem properly distinguishable, although, if I'd really want to nitpick I'd say the and are too different, with the latter (REGISTERED SIGN) being way too small, basically unreadable here. Since I see this sign approximately never, it probably doesn't matter at all. I like how the ampersand (&) is more traditional, although I'll miss the exotic one Fira produced... I like how the back quotes ( , GRAVE ACCENT) drop down low, nicely aligned with the apostrophe. As I mentioned before, I like how the bar on the "f" aligns with the other top of letters, something in Fira mono that really annoys me now that I've noticed it (it's not aligned!).

A UTF-8 test file Here's the test sheet I've made up to test various characters. I could have sworn I had a good one like this lying around somewhere but couldn't find it so here it is, I guess.

US keyboard coverage:
abcdefghijklmnopqrstuvwxyz 1234567890-=[]\;',./
ABCDEFGHIJKLMNOPQRSTUVWXYZ~!@#$%^&*()_+ :"<>?
latin1 coverage:  
EURO SIGN, TRADE MARK SIGN:  
ambiguity test:
e coC0ODQ iI71lL! 
b6G&0B83  []() /\. 
zs$S52Z%   '" 
all characters in a sentence, uppercase:
the quick fox jumps over the lazy dog
THE QUICK FOX JUMPS OVER THE LAZY DOG
same, in french:
Portez ce vieux whisky au juge blond qui fume.
d s no l, o  un z phyr ha  me v t de gla ons w rmiens, je d ne
d exquis r tis de b uf au kir,   l a  d ge m r, &c tera.
D S NO L, O  UN Z PHYR HA  ME V T DE GLA ONS W RMIENS, JE D NE
D EXQUIS R TIS DE B UF AU KIR,   L A  D GE M R, &C TERA.
Ligatures test:
-<< -< -<- <-- <--- <<- <- -> ->> --> ---> ->- >- >>-
=<< =< =<= <== <=== <<= <= => =>> ==> ===> =>= >= >>=
<-> <--> <---> <----> <=> <==> <===> <====> :: ::: __
<~~ </ </> /> ~~> == != /= ~= <> === !== !=== =/= =!=
<: := *= *+ <* <*> *> <  < >  > <. <.> .> +* =* =: :>
(* *) /* */ [   ]     ++ +++ \/ /\  - -  <!-- <!---
Box drawing alignment tests:
                                                                    
                                 
                           
                                                   
                                       
                                                 
                               
                                
Dashes alignment test:
HYPHEN-MINUS, MINUS SIGN, EN, EM DASH, HORIZONTAL BAR, LOW LINE
--------------------------------------------------
 
 
 
 
__________________________________________________

Update: here is another such sample sheet, it's pretty good and has support for more languages while being still relatively small. So there you have it, got completely nerd swiped by typography again. Now I can go back to writing a too-long proposal again. Sources and inspiration for the above:

• the unicode(1) command, to lookup individual characters to disambiguate, for example, - (U+002D HYPHEN-MINUS, the minus sign next to zero on US keyboards) and (U+2212 MINUS SIGN, a math symbol)
• searchable list of characters and their names - roughly equivalent to the unicode(1) command, but in one page, amazingly the /usr/share/unicode database doesn't have any one file like this
• bits/UTF-8-Unicode-Test-Documents - full list of UTF-8 characters
• UTF-8 encoded plain text file - nice examples of edge cases, curly quotes example and box drawing alignment test which, incidentally, showed me I needed specific faces customisation in Emacs to get the Markdown code areas to display properly, also the idea of comparing various dashes
• sample sentences in many languages - unused, "Sentences that contain all letters commonly used in a language"
• UTF-8 sampler - unused, similar

Other fonts In my previous blog post about fonts, I had a list of alternative fonts, but it seems people are not digging through this, so I figured I would redo the list here to preempt "but have you tried Jetbrains mono" kind of comments. My requirements are:

• no ligatures: yes, in the previous post, I wanted ligatures but I have changed my mind. after testing this, I find them distracting, confusing, and they often break the monospace nature of the display
• monospace: this is to display code
• italics: often used when writing Markdown, where I do make use of italics... Emacs falls back to underlining text when lacking italics which is hard to read
• free-ish, ultimately should be packaged in Debian

Here is the list of alternatives I have considered in the past and why I'm not using them:

• agave: recommended by tarzeau, not sure I like the lowercase a, a bit too exotic, packaged as fonts-agave
• Cascadia code: optional ligatures, multilingual, not liking the alignment, ambiguous parenthesis (look too much like square brackets), new default for Windows Terminal and Visual Studio, packaged as fonts-cascadia-code
• Fira Code: ligatures, was using Fira Mono from which it is derived, lacking italics except for forks, interestingly, Fira Code succeeds the alignment test but Fira Mono fails to show the X signs properly! packaged as fonts-firacode
• Hack: no ligatures, very similar to Fira, italics, good alternative, fails the X test in box alignment, packaged as fonts-hack
• Hermit: no ligatures, smaller, alignment issues in box drawing and dashes, packaged as fonts-hermit somehow part of cool-retro-term
• IBM Plex: irritating website, replaces Helvetica as the IBM corporate font, no ligatures by default, italics, proportional alternatives, serifs and sans, multiple languages, partial failure in box alignment test (X signs), fancy curly braces contrast perhaps too much with the rest of the font, packaged in Debian as fonts-ibm-plex
• Inconsolata: no ligatures, maybe italics? more compressed than others, feels a little out of balance because of that, packaged in Debian as fonts-inconsolata
• Intel One Mono: nice legibility, no ligatures, alignment issues in box drawing, not packaged in Debian
• Iosevka: optional ligatures, italics, multilingual, good legibility, has a proportional option, serifs and sans, line height issue in box drawing, fails dash test, not in Debian
• Jetbrains Mono: (mandatory?) ligatures, good coverage, originally rumored to be not DFSG-free (Debian Free Software Guidelines) but ultimately packaged in Debian as fonts-jetbrains-mono
• Monoid: optional ligatures, feels much "thinner" than Jetbrains, not liking alignment or spacing on that one, ambiguous 2Z, problems rendering box drawing, packaged as fonts-monoid
• Mononoki: no ligatures, looks good, good alternative, suggested by the Debian fonts team as part of fonts-recommended, problems rendering box drawing, em dash bigger than en dash, packaged as fonts-mononoki
• Source Code Pro: italics, looks good, but dash metrics look whacky, not in Debian
• spleen: bitmap font, old school, spacing issue in box drawing test, packaged as fonts-spleen
• sudo: personal project, no ligatures, zero originally not dotted, relied on metrics for legibility, spacing issue in box drawing, not in Debian

So, if I get tired of Commit Mono, I might probably try, in order:

1. Hack
2. Jetbrains Mono
3. IBM Plex Mono

Iosevka, Monoki and Intel One Mono are also good options, but have alignment problems. Iosevka is particularly disappointing as the EM DASH metrics are just completely wrong (much too wide). This was tested using the Programming fonts site which has all the above fonts, which cannot be said of Font Squirrel or Google Fonts, amazingly. Other such tools:

• Coding Font (broken in Firefox as of 2024-05-30)
• dev fonts comparator
• Font Squirrel
• Google Fonts
• Programming fonts

• learn about nix
• install software that might not be available in Debian
• install software without root access
• declare software necessary for a user s environment inside $HOME/.config

with (import <nixpkgs>  );
 
  packageOverrides = pkgs: with pkgs;  
    thk-emacsWithPackages = (pkgs.emacsPackagesFor emacs-gtk).emacsWithPackages (
      epkgs:
      (with epkgs.elpaPackages; [
        ace-window
        company
        org
        use-package
      ]) ++ (with epkgs.melpaPackages; [
        editorconfig
        flycheck
        haskell-mode
        magit
        nix-mode
        paredit
        rainbow-delimiters
        treemacs
        visual-fill-column
        yasnippet-snippets
      ]) ++ [    # From main packages set
      ]
    );

    userPackages = buildEnv  
      extraOutputsToInstall = [ "doc" "info" "man" ];
      name = "user-packages";
      paths = [
        ghc
        git
        (pkgs.haskell-language-server.override   supportedGhcVersions = [ "94" ];  )
        nix
        stack
        thk-emacsWithPackages
        tmux
        vcsh
        virtiofsd
      ];
     ;
   ;
 

nix-env --install --attr nixpkgs.userPackages --remove-all

[Service]
ExecStart=
ExecStart=@/home/thk/.local/state/nix/profile/bin/nix-daemon nix-daemon --daemon

1. We can mark packages as install or reject
2. We can order actions/clauses. When backtracking the action that came later will be the first we try to backtrack on
3. We can order the choices of a dependency - we try them left to right.

apt upgrade is hard The apt upgrade commands implements a safe version of dist-upgrade that essentially calculates the dist-upgrade, and then undoes anything that would cause a package to be removed, but it (unlike its apt-get counterpart) allows the solver to install new packages. Now, consider the following package is installed:

X Depends: A (= 1)   B

An upgrade from A=1 to A=2 is available. What should happen? The classic solver would choose to remove X in a dist-upgrade, and then upgrade A, so it s answer is quite clear: Keep back the upgrade of A. The new solver however sees two possible solutions:

1. Install B to satisfy X Depends A (= 1) B.
2. Keep back the upgrade of A

Which one does it pick? This depends on the order in which it sees the upgrade action for A and the dependency, as it will backjump chronologically. So

1. If it gets to the dependency first, it marks A=1 for install to satisfy A (= 1). Then it gets to the upgrade request, which is just A Depends A (= 2) A (= 1) and sees it is satisfied already and is content.
2. If it gets to the upgrade request first, it marks A=2 for install to satisfy A (= 2). Then later it gets to X Depends: A (= 1) B, sees that A (= 1) is not satisfiable, and picks B.

We have two ways to approach this issue:

1. We always order upgrade requests last, so they will be kept back in case of conflicting dependencies
2. We require that, for apt upgrade a currently satisfied dependency must be satisfied by currently installed packages, hence eliminating B as a choice.

Recommends are hard too See if you have a X Recommends: A (= 1) and a new version of A, A (= 2), the solver currently will silently break the Recommends in some cases. But let s explore what the behavior of a X Recommends: A (= 1) in combination with an available upgrade of A (= 2) should be. We could say the rule should be:

• An upgrade should keep back A instead of breaking the Recommends
• A dist-upgrade should either keep back A or remove X (if it is obsolete)

This essentially leaves us the same choices as for the previous problem, but with an interesting twist. We can change the ordering (and we already did), but we could also introduce a new rule, promotions :

A Recommends in an installed package, or an upgrade to that installed package, where the Recommends existed in the installed version, that is currently satisfied, must continue to be satisfied, that is, it effectively is promoted to a Depends.

This neatly solves the problem for us. We will never break Recommends that are satisfied. Likewise, we already have a Recommends demotion rule:

A Recommends in an installed package, or an upgrade to that installed package, where the Recommends existed in the installed version, that is currently unsatisfied, will not be further evaluated (it is treated like a Suggests is in the default configuration).

Whether we should be allowed to break Suggests with our decisions or not (the old autoremover did not, for instance) is a different decision. Should we promote currently satisified Suggests to Depends as well? Should we follow currently satisified Suggests so the solver sees them and doesn t autoremove them, but treat them as optional?

tightening of versioned dependencies Another case of versioned dependencies with alternatives that has complex behavior is something like

X Depends: A (>= 2)   B
X Recommends: A (>= 2)   B

In both cases, installing X should upgrade an A < 2 in favour of installing B. But a naive SAT solver might not. If your request to keep A installed is encoded as A (= 1) A (= 2), then it first picks A (= 1). When it sees the Depends/Recommends it will switch to B. We can solve this again as in the previous example by ordering the keep A installed requests after any dependencies. Notably, we will enqueue the common dependencies of all A versions first before selecting a version of A, so something may select a version for us.

version narrowing instead of version choosing A different approach to dealing with the issue of version selection is to not select a version until the very last moment. So instead of selecting a version to satisfy A (>= 2) we instead translate

Depends: A (>= 2)

into two rules:

1. The package selection rule:

    Depends: A
   

   This ensures that any version of A is installed (i.e. it adds a version choice clause, A (= 1) A (= 2) in an example with two versions for A.
2. The version narrowing rule:

    Conflicts: A (<< 2)
   

   This outright would reject a choice of A (= 1).

So now we have 3 kinds of clauses:

1. package selection
2. version narrowing
3. version selection

If we process them in that order, we should surely be able to find the solution that best matches the semantics of our Debian dependency model, i.e. selecting earlier choices in a dependency before later choices in the face of version restrictions. This still leaves one issue: What if our maintainer did not use Depends: A (>= 2) B but e.g. Depends: A (= 3) B A (= 2). He d expect us to fall back to B if A (= 3) is not installable, and not to B. But we d like to enqueue A and reject all choices other than 3 and 2. I think it s fair to say: Don t do that, then here.

Implementing strict pinning correctly APT knows a single candidate version per package, this makes the solver relatively deterministic: It will only ever pick the candidate, or an installed version. This also happens to significantly reduce the search space which is good - less backtracking. An uptodate system will only ever have one version per package that can be installed, so we never actually have to choose versions. But of course, APT allows you to specify a non-candidate version of a package to install, for example:

apt install foo/oracular-proposed

The way this works is that the core component of the previous solver, which is the pkgDepCache maintains what essentially amounts to an overlay of the policy that you could see with apt-cache policy. The solver currently however validates allowed version choices against the policy directly, and hence finds these versions are not allowed and craps out. This is an interesting problem because the solver should not be dependent on the pkgDepCache as the pkgDepCache initialization (Building dependency tree...) accounts for about half of the runtime of APT (until the Y/n prompt) and I d really like to get rid of it. But currently the frontend does go via the pkgDepCache. It marks the packages in there, building up what you could call a transaction, and then we translate it to the new solver, and once it is done, it translates the result back into the pkgDepCache. The current implementation of allowed version is implemented by reducing the search space, i.e. every dependency, we outright ignore any non-allowed versions. So if you have a version 3 of A that is ignored a Depends: A would be translated into A (= 2) A (= 1). However this has two disadvantages. (1) It means if we show you why A could not be installed, you don t even see A (= 3) in the list of choices and (2) you would need to keep the pkgDepCache around for the temporary overrides. So instead of actually enforcing the allowed version rule by filtering, a more reasonable model is that we apply the allowed version rule by just marking every other version as not allowed when discovering the package in the from depcache translation layer. This doesn t really increase the search space either but it solves both our problem of making overrides work and giving you a reasonable error message that lists all versions of A.

pulling up common dependencies to minimize backtracking cost One of the common issues we have is that when we have a dependency group

 A   B   C   D 

we try them in order, and if one fails, we undo everything it did, and move on to the next one. However, this isn t perhaps the best choice of operation. I explained before that one thing we do is queue the common dependencies of a package (i.e. dependencies shared in all versions) when marking a package for install, but we don t do this here: We have already lowered the representation of the dependency group into a list of versions, so we d need to extract the package back out of it. This can of course be done, but there may be a more interesting solution to the problem, in that we simply enqueue all the common dependencies. That is, we add n backtracking levels for n possible solutions:

1. We enqueue the common dependencies of all possible solutions deps(A)&deps(B)&deps(C)&deps(D)
2. We decide (adding a decision level) not to install D right now and enqueue deps(A)&deps(B)&deps(C)
3. We decide (adding a decision level) not to install C right now and enqueue deps(A)&deps(B)
4. We decide (adding a decision level) not to install B right now and enqueue A

Now if we need to backtrack from our choice of A we hopefully still have a lot of common dependencies queued that we do not need to redo. While we have more backtracking levels, each backtracking level would be significantly cheaper, especially if you have cheap backtracking (which admittedly we do not have, yet anyway). The caveat though is: It may be pretty expensive to find the common dependencies. We need to iterate over all dependency groups of A and see if they are in B, C, and D, so we have a complexity of roughly #A * (#B+#C+#D) Each dependency group we need to check i.e. is X Y in B meanwhile has linear cost: We need to compare the memory content of two pointer arrays containing the list of possible versions that solve the dependency group. This means that X Y and Y X are different dependencies of course, but that is to be expected they are. But any dependency of the same order will have the same memory layout. So really the cost is roughly N^4. This isn t nice. You can apply various heuristics here on how to improve that, or you can even apply binary logic:

1. Enqueue common dependencies of A B C D
2. Move into the left half, enqueue of A B
3. Again divide and conquer and select A.

This has a significant advantage in long lists of choices, and also in the common case, where the first solution should be the right one. Or again, if you enqueue the package and a version restriction instead, you already get the common dependencies enqueued for the chosen package at least.

• make RedHatSignedRpmScanner distribution agnostic allowed Leapp to load GPG keys for alternative distributions, thus correctly identifying their packages
• call correct mkdir when trying to create /etc/rhsm/facts made Leapp not crash on systems without /etc/rhsm, like those not subscribed to Red Hat
• default to NO_RHSM mode when subscription-manager is not found made Leapp work better out of the box on systems without subscriptions
• load all substitutions from etc made Leapp work properly on distributions that ship additional repository URL substitutions
• be less strict when figuring out major version in initrd made Leapp work with CentOS Stream initrds that have non X.Y versions

# dnf copr enable evgeni/leapp
# dnf install leapp leapp-upgrade-el8toel9

# vim /etc/leapp/files/leapp_upgrade_repositories.repo
[c9-baseos]
name=CentOS Stream $releasever - BaseOS
metalink=https://mirrors.centos.org/metalink?repo=centos-baseos-9-stream&arch=$basearch&protocol=https,http
gpgkey=file:///etc/pki/rpm-gpg/RPM-GPG-KEY-centosofficial
gpgcheck=1
repo_gpgcheck=0
metadata_expire=6h
countme=1
enabled=1
[c9-appstream]
name=CentOS Stream $releasever - AppStream
metalink=https://mirrors.centos.org/metalink?repo=centos-appstream-9-stream&arch=$basearch&protocol=https,http
gpgkey=file:///etc/pki/rpm-gpg/RPM-GPG-KEY-centosofficial
gpgcheck=1
repo_gpgcheck=0
metadata_expire=6h
countme=1
enabled=1

# leapp preupgrade
 
============================================================
                      REPORT OVERVIEW                       
============================================================
Reports summary:
    Errors:                      0
    Inhibitors:                  0
    HIGH severity reports:       0
    MEDIUM severity reports:     0
    LOW severity reports:        3
    INFO severity reports:       3
Before continuing consult the full report:
    A report has been generated at /var/log/leapp/leapp-report.json
    A report has been generated at /var/log/leapp/leapp-report.txt
============================================================
                   END OF REPORT OVERVIEW                   
============================================================

Error starting domain: unsupported configuration: domain configuration does not support video model 'qxl'

    <channel type='spicevmc'>
      <target type='virtio' name='com.redhat.spice.0'/>
      <address type='virtio-serial' controller='0' bus='0' port='2'/>
    </channel>
    <graphics type='spice' autoport='yes'>
      <listen type='address'/>
    </graphics>
    <audio id='1' type='spice'/>
    <video>
      <model type='qxl' ram='65536' vram='65536' vgamem='16384' heads='1' primary='yes'/>
      <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x0'/>
    </video>
    <redirdev bus='usb' type='spicevmc'> 
      <address type='usb' bus='0' port='2'/> 
    </redirdev> 
    <redirdev bus='usb' type='spicevmc'> 
      <address type='usb' bus='0' port='3'/> 
    </redirdev>

    <graphics type='vnc' port='-1' autoport='yes'>
      <listen type='address'/>
    </graphics>
    <audio id='1' type='none'/>
    <video>
      <model type='cirrus' vram='16384' heads='1' primary='yes'/>
      <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x0'/>
    </video>

$ echo ghp_token   podman login ghcr.io -u <user> --password-stdin

• 06 salas em paralelo no s bado;
• 02 audit rios em paralelo na segunda e ter a-feira;
• 30 palestras/BoFs de todos os n veis;
• 05 oficinas para atividades do tipo m o na massa;
• 09 lightning talks sobre temas diversos;
• 01 performance Live Eletronics com Software Livre;
• Install fest para instalar Debian nos notebook dos(as) participantes;
• BSP (Bug Squashing Party - Festa de Ca a Bugs);
• Uploads de pacotes novos ou atualizados.

• Total de pessoas inscritas: 399
• Total de pessoas presentes: 224

• Alimenta o: 69
• Hospedagem: 20
• Passagens: 18

• Youtube
• Peertube
• video.debian.net

• Google photos
• Nextcloud

• Collabora

• Jedai.ai
• Toradex Brasil
• Globo.com
• Policorp Tecnologia

• BRDSoft - Tecnologias para TI e Telecomunica es
• EITA - Cooperativa de Trabalho Educa o, Informa o e Tecnologia para Autogest o

• ICTL - Instituto para Conserva o de Tecnologias Livres
• Nazinha Alimentos
• DACOMPSI

• Projeto Debian
• Comunidade Debian Brasil
• Comunidade Debian MG
• DCC/UFMG - Departamento de Ci ncia da Computa o da Universidade Federal de Minas Gerais