Note: I've been awfully silent here for the past ... (checks notes) oh dear, 3 months! But that's not because I've been idle, quite the contrary, I've been very busy but just didn't have time to write about anything. So I've taken it upon myself to write something about my work this week, and published this post on the Tor blog which I copy here for a broader audience. Let me know if you like this or not.

https://git.torproject.org/

https://gitlab.torproject.org/

Where is the code? Normally, nothing should be lost. All repositories in gitolite have been either explicitly migrated by their owners, forcibly migrated by the sysadmin team (TPA), or explicitly destroyed at their owner's request. An exhaustive rewrite map translates gitolite projects to GitLab projects. Some of those projects actually redirect to their parent in cases of empty repositories that were obvious forks. Destroyed repositories redirect to the GitLab front page. Because the migration happened progressively, it's technically possible that commits pushed to gitolite were lost after the migration. We took great care to avoid that scenario. First, we adopted a proposal (TPA-RFC-36) in June 2023 to announce the transition. Then, in March 2024, we locked down all repositories from any further changes. Around that time, only a handful of repositories had changes made after the adoption date, and we examined each repository carefully to make sure nothing was lost. Still, we built a diff of all the changes in the git references that archivists can peruse to check for data loss. It's large (6MiB+) because a lot of repositories were migrated before the mass migration and then kept evolving in GitLab. Many other repositories were rebuilt in GitLab from parent to rebuild a fork relationship which added extra references to those clones. A note to amateur archivists out there, it's probably too late for one last crawl now. The Git repositories now all redirect to GitLab and are effectively unavailable in their original form. That said, the GitWeb site was crawled into the Internet Archive in February 2024, so at least some copy of it is available in the Wayback Machine. At that point, however, many developers had already migrated their projects to GitLab, so the copies there were already possibly out of date compared with the repositories in GitLab. Software Heritage also has a copy of all repositories hosted on Gitolite since June 2023 and have continuously kept mirroring the repositories, where they will be kept hopefully in eternity. There's an issue where the main website can't find the repositories when you search for gitweb.torproject.org, instead search for git.torproject.org. In any case, if you believe data is missing, please do let us know by opening an issue with TPA.

Why? This is an old project in the making. The first discussion about migrating from gitolite to GitLab started in 2020 (almost 4 years ago). But going further back, the first GitLab experiment was in 2016, almost a decade ago. The current GitLab server dates from 2019, replacing Trac for issue tracking in 2020. It was originally supposed to host only mirrors for merge requests and issue trackers but, naturally, one thing led to another and eventually, GitLab had grown a container registry, continuous integration (CI) runners, GitLab Pages, and, of course, hosted most Git repositories. There were hesitations at moving to GitLab for code hosting. We had discussions about the increased attack surface and ways to mitigate that, but, ultimately, it seems the issues were not that serious and the community embraced GitLab. TPA actually migrated its most critical repositories out of shared hosting entirely, into specific servers (e.g. the Puppet Git repository is just on the Puppet server now), leveraging Git's decentralized nature and removing an entire attack surface from our infrastructure. Some of those repositories are mirrored back into GitLab, but the authoritative copy is not on GitLab. In any case, the proposal to migrate from Gitolite to GitLab was effectively just formalizing a fait accompli.

How to migrate from Gitolite / cgit to GitLab The progressive migration was a challenge. If you intend to migrate between hosting platforms, we strongly recommend to make a "flag day" during which you migrate all repositories at once. This ensures a smoother transition and avoids elaborate rewrite rules. When Gitolite access was shutdown, we had repositories on both GitLab and Gitolite, without a clear relationship between the two. A priori, the plan then was to import all the remaining Gitolite repositories into the legacy/gitolite namespace, but that seemed wasteful, particularly for large repositories like Tor Browser which uses nearly a gigabyte of disk space. So we took special care to avoid duplicating repositories. When the mass migration started, only 71 of the 538 Gitolite repositories were Migrated to GitLab in the gitolite.conf file. So, given that we had hundreds of repositories to migrate:, we developed some automation to "save time". We already automate similar ad-hoc tasks with Fabric, so we used that framework here as well. (Our normal configuration management tool is Puppet, which is a poor fit here.) So a relatively large amount of Python code was produced to basically do the following:

1. check if all on-disk repositories are listed in gitolite.conf (and vice versa) and either add missing repositories or delete them from disk if garbage
2. for each repository in gitolite.conf, if its category is marked Migrated to GitLab, skip, otherwise;
3. find a matching GitLab project by name, prompt the user for multiple matches
4. if a match is found, redirect if the repository is non-empty
   • we have GitLab projects that look like the real thing, but are only present to host migrated Trac issues
   • in such cases we cloned the Gitolite project locally and pushed to the existing repository instead
5. otherwise, a new repository is created in the legacy/gitolite namespace, using the "import" mechanism in GitLab to automatically import the repository from Gitolite, creating redirections and updating gitolite.conf to document the change

User repositories (those under the user/ directory in Gitolite) were handled specially. First, the existing redirection map was checked to see if a similarly named project was migrated (so that, e.g. user/dgoulet/tor is properly treated as a fork of tpo/core/tor). Then the parent project was forked in GitLab and the Gitolite project force-pushed to the fork. This allows us to show the fork relationship in GitLab and, more importantly, benefit from the "pool" feature in GitLab which deduplicates disk usage between forks. Sometimes, we found no such relationships. Then we simply imported multiple repositories with similar names in the legacy/gitolite namespace, sometimes creating forks between user repositories, on a first-come-first-served basis from the gitolite.conf order. The code used in this migration is now available publicly. We encourage other groups planning to migrate from Gitolite/GitWeb to GitLab to use (and contribute to) our fabric-tasks repository, even though it does have its fair share of hard-coded assertions. The main entry point is the gitolite.mass-repos-migration task. A typical migration job looked like:

anarcat@angela:fabric-tasks$ fab -H cupani.torproject.org gitolite.mass-repos-migration 
[...]
INFO: skipping project project/help/infra in category Migrated to GitLab
INFO: skipping project project/help/wiki in category Migrated to GitLab
INFO: skipping project project/jenkins/jobs in category Migrated to GitLab
INFO: skipping project project/jenkins/tools in category Migrated to GitLab
INFO: searching for projects matching fastlane
INFO: Successfully connected to https://gitlab.torproject.org
import gitolite project project/tor-browser/fastlane into gitlab legacy/gitolite/project/tor-browser/fastlane with desc 'Tor Browser app store and deployment configuration for Fastlane'? [Y/n] 
INFO: importing gitolite project project/tor-browser/fastlane into gitlab legacy/gitolite/project/tor-browser/fastlane with desc 'Tor Browser app store and deployment configuration for Fastlane'
INFO: building a new connect to cupani
INFO: defaulting name to fastlane
INFO: importing project into GitLab
INFO: Successfully connected to https://gitlab.torproject.org
INFO: loading group legacy/gitolite/project/tor-browser
INFO: archiving project
INFO: creating repository fastlane (fastlane) in namespace legacy/gitolite/project/tor-browser from https://git.torproject.org/project/tor-browser/fastlane into https://gitlab.torproject.org/legacy/gitolite/project/tor-browser/fastlane
INFO: migrating Gitolite repository project/tor-browser/fastlane to GitLab project legacy/gitolite/project/tor-browser/fastlane
INFO: uploading 399 bytes to /srv/git.torproject.org/repositories/project/tor-browser/fastlane.git/hooks/pre-receive
INFO: making /srv/git.torproject.org/repositories/project/tor-browser/fastlane.git/hooks/pre-receive executable
INFO: adding entry to rewrite_map /home/anarcat/src/tor/tor-puppet/modules/profile/files/git/gitolite2gitlab.txt
INFO: modifying gitolite.conf to add: "config gitweb.category = Migrated to GitLab"
INFO: rewriting gitolite config /home/anarcat/src/tor/gitolite-admin/conf/gitolite.conf to change project project/tor-browser/fastlane to category Migrated to GitLab
INFO: skipping project project/bridges/bridgedb-admin in category Migrated to GitLab
[...]

In the above, you can see migrated repositories skipped then the fastlane project being archived into GitLab. Another example with a later version of the script, processing only user repositories and showing the interactive prompt and a force-push into a fork:

$ fab -H cupani.torproject.org  gitolite.mass-repos-migration --include 'user/.*' --exclude '.*tor-?browser.*'
INFO: skipping project user/aagbsn/bridgedb in category Migrated to GitLab
[...]
INFO: skipping project user/phw/atlas in category Migrated to GitLab
INFO: processing project user/phw/obfsproxy (Philipp's obfsproxy repository) in category Users' development repositories (Attic)
INFO: Successfully connected to https://gitlab.torproject.org
INFO: user repository detected, trying to find fork phw/obfsproxy
WARNING: no existing fork found, entering user fork subroutine
INFO: found 6 GitLab projects matching 'obfsproxy' (https://gitweb.torproject.org/user/phw/obfsproxy.git)
0 legacy/gitolite/debian/obfsproxy
1 legacy/gitolite/debian/obfsproxy-legacy
2 legacy/gitolite/user/asn/obfsproxy
3 legacy/gitolite/user/ioerror/obfsproxy
4 tpo/anti-censorship/pluggable-transports/obfsproxy
5 tpo/anti-censorship/pluggable-transports/obfsproxy-legacy
select parent to fork from, or enter to abort: ^G4
INFO: repository is not empty: in-pack: 2104, packs: 1, size-pack: 414
fork project tpo/anti-censorship/pluggable-transports/obfsproxy into legacy/gitolite/user/phw/obfsproxy^G [Y/n] 
INFO: loading project tpo/anti-censorship/pluggable-transports/obfsproxy
INFO: forking project user/phw/obfsproxy into namespace legacy/gitolite/user/phw
INFO: waiting for fork to complete...
INFO: fork status: started, sleeping...
INFO: fork finished
INFO: cloning and force pushing from user/phw/obfsproxy to legacy/gitolite/user/phw/obfsproxy
INFO: deleting branch protection: <class 'gitlab.v4.objects.branches.ProjectProtectedBranch'> =>  'id': 2723, 'name': 'master', 'push_access_levels': [ 'id': 2864, 'access_level': 40, 'access_level_description': 'Maintainers', 'deploy_key_id': None ], 'merge_access_levels': [ 'id': 2753, 'access_level': 40, 'access_level_description': 'Maintainers' ], 'allow_force_push': False 
INFO: cloning repository git-rw.torproject.org:/srv/git.torproject.org/repositories/user/phw/obfsproxy.git in /tmp/tmp6orvjggy/user/phw/obfsproxy
Cloning into bare repository '/tmp/tmp6orvjggy/user/phw/obfsproxy'...
INFO: pushing to GitLab: https://gitlab.torproject.org/legacy/gitolite/user/phw/obfsproxy
remote: 
remote: To create a merge request for bug_10887, visit:        
remote:   https://gitlab.torproject.org/legacy/gitolite/user/phw/obfsproxy/-/merge_requests/new?merge_request%5Bsource_branch%5D=bug_10887        
remote: 
[...]
To ssh://gitlab.torproject.org/legacy/gitolite/user/phw/obfsproxy
 + 2bf9d09...a8e54d5 master -> master (forced update)
 * [new branch]      bug_10887 -> bug_10887
[...]
INFO: migrating repo
INFO: migrating Gitolite repository https://gitweb.torproject.org/user/phw/obfsproxy.git to GitLab project https://gitlab.torproject.org/legacy/gitolite/user/phw/obfsproxy
INFO: adding entry to rewrite_map /home/anarcat/src/tor/tor-puppet/modules/profile/files/git/gitolite2gitlab.txt
INFO: modifying gitolite.conf to add: "config gitweb.category = Migrated to GitLab"
INFO: rewriting gitolite config /home/anarcat/src/tor/gitolite-admin/conf/gitolite.conf to change project user/phw/obfsproxy to category Migrated to GitLab
INFO: processing project user/phw/scramblesuit (Philipp's ScrambleSuit repository) in category Users' development repositories (Attic)
INFO: user repository detected, trying to find fork phw/scramblesuit
WARNING: no existing fork found, entering user fork subroutine
WARNING: no matching gitlab project found for user/phw/scramblesuit
INFO: user fork subroutine failed, resuming normal procedure
INFO: searching for projects matching scramblesuit
import gitolite project user/phw/scramblesuit into gitlab legacy/gitolite/user/phw/scramblesuit with desc 'Philipp's ScrambleSuit repository'?^G [Y/n] 
INFO: checking if remote repo https://git.torproject.org/user/phw/scramblesuit exists
INFO: importing gitolite project user/phw/scramblesuit into gitlab legacy/gitolite/user/phw/scramblesuit with desc 'Philipp's ScrambleSuit repository'
INFO: importing project into GitLab
INFO: Successfully connected to https://gitlab.torproject.org
INFO: loading group legacy/gitolite/user/phw
INFO: creating repository scramblesuit (scramblesuit) in namespace legacy/gitolite/user/phw from https://git.torproject.org/user/phw/scramblesuit into https://gitlab.torproject.org/legacy/gitolite/user/phw/scramblesuit
INFO: archiving project
INFO: migrating Gitolite repository https://gitweb.torproject.org/user/phw/scramblesuit.git to GitLab project https://gitlab.torproject.org/legacy/gitolite/user/phw/scramblesuit
INFO: adding entry to rewrite_map /home/anarcat/src/tor/tor-puppet/modules/profile/files/git/gitolite2gitlab.txt
INFO: modifying gitolite.conf to add: "config gitweb.category = Migrated to GitLab"
INFO: rewriting gitolite config /home/anarcat/src/tor/gitolite-admin/conf/gitolite.conf to change project user/phw/scramblesuit to category Migrated to GitLab
[...]

Acute eyes will notice the bell used as a notification mechanism as well in this transcript. A lot of the code is now useless for us, but some, like "commit and push" or is-repo-empty live on in the git module and, of course, the gitlab module has grown some legs along the way. We've also found fun bugs, like a file descriptor exhaustion in bash, among other oddities. The retirement milestone and issue 41215 has a detailed log of the migration, for those curious. This was a challenging project, but it feels nice to have this behind us. This gets rid of 2 of the 4 remaining machines running Debian "old-old-stable", which moves a bit further ahead in our late bullseye upgrades milestone. Full transparency: we tested GPT-3.5, GPT-4, and other large language models to see if they could answer the question "write a set of rewrite rules to redirect GitWeb to GitLab". This has become a standard LLM test for your faithful writer to figure out how good a LLM is at technical responses. None of them gave an accurate, complete, and functional response, for the record. The actual rewrite rules as of this writing follow, for humans that actually like working answers provided by expert humans instead of artificial intelligence which currently seem to be, glorified, mansplaining interns.

git.torproject.org rewrite rules Those rules are relatively simple in that they rewrite a single URL to its equivalent GitLab counterpart in a 1:1 fashion. It relies on the rewrite map mentioned above, of course.

RewriteEngine on
# this RewriteMap connects the gitweb projects to their GitLab
# equivalent
RewriteMap gitolite2gitlab "txt:/etc/apache2/gitolite2gitlab.txt"
# if this becomes a performance bottleneck, convert to a DBM map with:
#
#  $ httxt2dbm -i mapfile.txt -o mapfile.map
#
# and:
#
# RewriteMap mapname "dbm:/etc/apache/mapfile.map"
#
# according to reports lavamind found online, we hit such a
# performance bottleneck only around millions of entries, which is not our case
# those two rules can go away once all the projects are
# migrated to GitLab
#
# this matches the request URI so we can check the RewriteMap
# for a match next
#
# WARNING: this won't match URLs without .git in them, which
# *do* work now. one possibility would be to match the request
# URI (without query string!) with:
#
# /git/(.*)(.git)?/(((branches hooks info objects/).*) git-.* upload-pack receive-pack HEAD config description)?.
#
# I haven't been able to figure out the actual structure of
# those URLs, so it's really hard to figure out the boundaries
# of the project name here. I stopped after pouring around the
# http-backend.c code in git
# itself. https://www.git-scm.com/docs/http-protocol is also
# kind of incomplete and unsatisfying.
RewriteCond % REQUEST_URI  ^/(git/)?(.*).git/.*$
# this makes the RewriteRule match only if there's a match in
# the rewrite map
RewriteCond $ gitolite2gitlab:%2 NOT_FOUND  !NOT_FOUND
RewriteRule ^/(git/)?(.*).git/(.*)$ https://gitlab.torproject.org/$ gitolite2gitlab:$2 .git/$3 [R=302,L]
# Fallback everything else to GitLab
RewriteRule (.*) https://gitlab.torproject.org [R=302,L]

gitweb.torproject.org rewrite rules Those are the vastly more complicated GitWeb to GitLab rewrite rules. Note that we say "GitWeb" but we were actually not running GitWeb but cgit, as the former didn't actually scale for us.

RewriteEngine on
# this RewriteMap connects the gitweb projects to their GitLab
# equivalent
RewriteMap gitolite2gitlab "txt:/etc/apache2/gitolite2gitlab.txt"
# special rule to process targets of the old spec.tpo site and
# bring them to the right redirect on the new spec.tpo site. that should turn, for example:
#
# https://gitweb.torproject.org/torspec.git/tree/address-spec.txt
#
# into:
#
# https://spec.torproject.org/address-spec
RewriteRule ^/torspec.git/tree/(.*).txt$ https://spec.torproject.org/$1 [R=302]
# list of endpoints taken from cgit's cmd.c
# those two RewriteCond are necessary because we don't move
# all repositories at once. once the migration is completed,
# they can be removed.
#
# and yes, they are copied all over the place below
#
# create a match for the project name to check if the project
# has been moved to GitLab
RewriteCond % REQUEST_URI  ^/(.*).git(/.*)?$
# this makes the RewriteRule match only if there's a match in
# the rewrite map
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
# main project page, like summary below
RewriteRule ^/(.*).git/?$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 / [R=302,L]
# summary
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteRule ^/(.*).git/summary/?$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 / [R=302,L]
# about
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteRule ^/(.*).git/about/?$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 / [R=302,L]
# commit
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteCond "% QUERY_STRING " "(.*(?:^ &))id=([^&]*)(&.*)?$"
RewriteRule ^/(.*).git/commit/? https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/commit/%2 [R=302,L,QSD]
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteRule ^/(.*).git/commit/? https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/commits/HEAD [R=302,L]
# diff, incomplete because can diff arbitrary refs and files in cgit but not in GitLab, hard to parse
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteCond % QUERY_STRING  id=([^&]*)
RewriteRule ^/(.*).git/diff/? https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/commit/%1 [R=302,L,QSD]
# patch
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteCond % QUERY_STRING  id=([^&]*)
RewriteRule ^/(.*).git/patch/? https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/commit/%1.patch [R=302,L,QSD]
# rawdiff, incomplete because can show only one file diff, which GitLab cannot
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteCond % QUERY_STRING  id=([^&]*)
RewriteRule ^/(.*).git/rawdiff/?$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/commit/%1.diff [R=302,L,QSD]
# log
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteCond % QUERY_STRING  h=([^&]*)
RewriteRule ^/(.*).git/log/?$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/commits/%1 [R=302,L,QSD]
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteRule ^/(.*).git/log/?$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/commits/HEAD [R=302,L]
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteRule ^/(.*).git/log(/?.*)$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/commits/HEAD$2 [R=302,L]
# atom
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteCond % QUERY_STRING  h=([^&]*)
RewriteRule ^/(.*).git/atom/?$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/commits/%1 [R=302,L,QSD]
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteRule ^/(.*).git/atom/?$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/commits/HEAD [R=302,L,QSD]
# refs, incomplete because two pages in GitLab, defaulting to "tags"
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteRule ^/(.*).git/refs/?$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/tags [R=302,L]
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteCond % QUERY_STRING  h=([^&]*)
RewriteRule ^/(.*).git/tag/? https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/tags/%1 [R=302,L,QSD]
# tree
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteCond % QUERY_STRING  id=([^&]*)
RewriteRule ^/(.*).git/tree(/?.*)$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/tree/%1$2 [R=302,L,QSD]
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteRule ^/(.*).git/tree(/?.*)$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/tree/HEAD$2 [R=302,L]
# /-/tree has no good default in GitLab, revert to HEAD which is a good
# approximation (we can't assume "master" here anymore)
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteRule ^/(.*).git/tree/?$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/tree/HEAD [R=302,L]
# plain
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteCond % QUERY_STRING  h=([^&]*)
RewriteRule ^/(.*).git/plain(/?.*)$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/raw/%1$2 [R=302,L,QSD]
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteRule ^/(.*).git/plain(/?.*)$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/raw/HEAD$2 [R=302,L]
# blame: disabled
#RewriteCond % REQUEST_URI  ^/(.*).git/.*$
#RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
#RewriteCond % QUERY_STRING  h=([^&]*)
#RewriteRule ^/(.*).git/blame(/?.*)$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/blame/%1$2 [R=302,L,QSD]
# same default as tree above
#RewriteCond % REQUEST_URI  ^/(.*).git/.*$
#RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
#RewriteRule ^/(.*).git/blame(/?.*)$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/blame/HEAD/$2 [R=302,L]
# stats
RewriteCond % REQUEST_URI  ^/(.*).git/.*$
RewriteCond $ gitolite2gitlab:%1 NOT_FOUND  !NOT_FOUND
RewriteRule ^/(.*).git/stats/?$ https://gitlab.torproject.org/$ gitolite2gitlab:$1 /-/graphs/HEAD [R=302,L]
# still TODO:
# repolist: once migration is complete
#
# cannot be done:
# atom: needs a feed token, user must be logged in
# blob: no direct equivalent
# info: not working on main cgit website?
# ls_cache: not working, irrelevant?
# objects: undocumented?
# snapshot: pattern too hard to match on cgit's side
# special case, we keep a copy of the main index on the archive
RewriteRule ^/?$ https://archive.torproject.org/websites/gitweb.torproject.org.html [R=302,L]
# Fallback: everything else to GitLab
RewriteRule .* https://gitlab.torproject.org [R=302,L]

The reference copy of those is available in our (currently private) Puppet git repository.

Me: Hi everyone, my name s Matt, and I m a mediocre programmer. Everyone: Hi, Matt. Facilitator: Are you an alcoholic, Matt? Me: No, not since I stopped reading Twitter. Facilitator: Then I think you re in the wrong room.

Neither a Borrower Nor a Lender Be If you know anything about Rust, you probably know about the dreaded borrow checker . It s the thing that makes sure you don t have two pieces of code trying to modify the same data at the same time, or using a value when it s no longer valid. While Rust s borrowing semantics allow excellent performance without compromising safety, for us mediocre programmers it gets very complicated, very quickly. So, the moment the compiler wants to start talking about explicit lifetimes , I shut it up by just using owned values instead. It s not that I never borrow anything; I have some situations that I know are borrow-safe for the mediocre programmer (I ll cover those later). But any time I m not sure how things will pan out, I ll go straight for an owned value. For example, if I need to store some text in a struct or enum, it s going straight into a String. I m not going to start thinking about lifetimes and &'a str; I ll leave that for smarter people. Similarly, if I need a list of things, it s a Vec<T> every time no &'b [T] in my structs, thank you very much.

Attack of the Clones Following on from the above, I ve come to not be afraid of .clone(). I scatter them around my code like seeds in a field. Life s too short to spend time trying to figure out who s borrowing what from whom, if I can just give everyone their own thing. There are warnings in the Rust book (and everywhere else) about how a clone can be expensive . While it s true that, yes, making clones of data structures consumes CPU cycles and memory, it very rarely matters. CPU cycles are (usually) plentiful and RAM (usually) relatively cheap. Mediocre programmer mental effort is expensive, and not to be spent on premature optimisation. Also, if you re coming from most any other modern language, Rust is already giving you so much more performance that you re probably ending up ahead of the game, even if you .clone() everything in sight. If, by some miracle, something I write gets so popular that the expense of all those spurious clones becomes a problem, it might make sense to pay someone much smarter than I to figure out how to make the program a zero-copy masterpiece of efficient code. Until then clone early and clone often, I say!

Derive Macros are Powerful Magicks If you start .clone()ing everywhere, pretty quickly you ll be hit with this error:

error[E0599]: no method named  clone  found for struct  Foo  in the current scope

This is because not everything can be cloned, and so if you want your thing to be cloned, you need to implement the method yourself. Well sort of. One of the things that I find absolutely outstanding about Rust is the derive macro . These allow you to put a little marker on a struct or enum, and the compiler will write a bunch of code for you! Clone is one of the available so-called derivable traits , so you add #[derive(Clone)] to your structs, and poof! you can .clone() to your heart s content. But there are other things that are commonly useful, and so I ve got a set of traits that basically all of my data structures derive:

#[derive(Clone, Debug, Default)]
struct Foo  
    // ...
 

Every time I write a struct or enum definition, that line #[derive(Clone, Debug, Default)] goes at the top. The Debug trait allows you to print a debug representation of the data structure, either with the dbg!() macro, or via the :? format in the format!() macro (and anywhere else that takes a format string). Being able to say what exactly is that? comes in handy so often, not having a Debug implementation is like programming with one arm tied behind your Aeron. Meanwhile, the Default trait lets you create an empty instance of your data structure, with all of the fields set to their own default values. This only works if all the fields themselves implement Default, but a lot of standard types do, so it s rare that you ll define a structure that can t have an auto-derived Default. Enums are easily handled too, you just mark one variant as the default:

#[derive(Clone, Debug, Default)]
enum Bar  
    Something(String),
    SomethingElse(i32),
    #[default]   // <== mischief managed
    Nothing,
 

Borrowing is OK, Sometimes While I previously said that I like and usually use owned values, there are a few situations where I know I can borrow without angering the borrow checker gods, and so I m comfortable doing it. The first is when I need to pass a value into a function that only needs to take a little look at the value to decide what to do. For example, if I want to know whether any values in a Vec<u32> are even, I could pass in a Vec, like this:

fn main()  
    let numbers = vec![0u32, 1, 2, 3, 4, 5];
    if has_evens(numbers)  
        println!("EVENS!");
     
 
fn has_evens(numbers: Vec<u32>) -> bool  
    numbers.iter().any( n  n % 2 == 0)
 

Howver, this gets ugly if I m going to use numbers later, like this:

fn main()  
    let numbers = vec![0u32, 1, 2, 3, 4, 5];
    if has_evens(numbers)  
        println!("EVENS!");
     
    // Compiler complains about "value borrowed here after move"
    println!("Sum:  ", numbers.iter().sum::<u32>());
 
fn has_evens(numbers: Vec<u32>) -> bool  
    numbers.iter().any( n  n % 2 == 0)
 

Helpfully, the compiler will suggest I use my old standby, .clone(), to fix this problem. But I know that the borrow checker won t have a problem with lending that Vec<u32> into has_evens() as a borrowed slice, &[u32], like this:

fn main()  
    let numbers = vec![0u32, 1, 2, 3, 4, 5];
    if has_evens(&numbers)  
        println!("EVENS!");
     
 
fn has_evens(numbers: &[u32]) -> bool  
    numbers.iter().any( n  n % 2 == 0)
 

The general rule I ve got is that if I can take advantage of lifetime elision (a fancy term meaning the compiler can figure it out ), I m probably OK. In less fancy terms, as long as the compiler doesn t tell me to put 'a anywhere, I m in the green. On the other hand, the moment the compiler starts using the words explicit lifetime , I nope the heck out of there and start cloning everything in sight. Another example of using lifetime elision is when I m returning the value of a field from a struct or enum. In that case, I can usually get away with returning a borrowed value, knowing that the caller will probably just be taking a peek at that value, and throwing it away before the struct itself goes out of scope. For example:

struct Foo  
    id: u32,
    desc: String,
 
impl Foo  
    fn description(&self) -> &str  
        &self.desc
     
 

Returning a reference from a function is practically always a mortal sin for mediocre programmers, but returning one from a struct method is often OK. In the rare case that the caller does want the reference I return to live for longer, they can always turn it into an owned value themselves, by calling .to_owned().

Avoid the String Tangle Rust has a couple of different types for representing strings String and &str being the ones you see most often. There are good reasons for this, however it complicates method signatures when you just want to take some sort of bunch of text , and don t care so much about the messy details. For example, let s say we have a function that wants to see if the length of the string is even. Using the logic that since we re just taking a peek at the value passed in, our function might take a string reference, &str, like this:

fn is_even_length(s: &str) -> bool  
    s.len() % 2 == 0
 

That seems to work fine, until someone wants to check a formatted string:

fn main()  
    // The compiler complains about "expected  &str , found  String "
    if is_even_length(format!("my string is  ", std::env::args().next().unwrap()))  
        println!("Even length string");
     
 

Since format! returns an owned string, String, rather than a string reference, &str, we ve got a problem. Of course, it s straightforward to turn the String from format!() into a &str (just prefix it with an &). But as mediocre programmers, we can t be expected to remember which sort of string all our functions take and add & wherever it s needed, and having to fix everything when the compiler complains is tedious. The converse can also happen: a method that wants an owned String, and we ve got a &str (say, because we re passing in a string literal, like "Hello, world!"). In this case, we need to use one of the plethora of available turn this into a String mechanisms (.to_string(), .to_owned(), String::from(), and probably a few others I ve forgotten), on the value before we pass it in, which gets ugly real fast. For these reasons, I never take a String or an &str as an argument. Instead, I use the Power of Traits to let callers pass in anything that is, or can be turned into, a string. Let us have some examples. First off, if I would normally use &str as the type, I instead use impl AsRef<str>:

fn is_even_length(s: impl AsRef<str>) -> bool  
    s.as_ref().len() % 2 == 0
 

Note that I had to throw in an extra as_ref() call in there, but now I can call this with either a String or a &str and get an answer. Now, if I want to be given a String (presumably because I plan on taking ownership of the value, say because I m creating a new instance of a struct with it), I use impl Into<String> as my type:

struct Foo  
    id: u32,
    desc: String,
 
impl Foo  
    fn new(id: u32, desc: impl Into<String>) -> Self  
        Self   id, desc: desc.into()  
     
 

We have to call .into() on our desc argument, which makes the struct building a bit uglier, but I d argue that s a small price to pay for being able to call both Foo::new(1, "this is a thing") and Foo::new(2, format!("This is a thing named name ")) without caring what sort of string is involved.

Always Have an Error Enum Rust s error handing mechanism (Results everywhere), along with the quality-of-life sugar surrounding it (like the short-circuit operator, ?), is a delightfully ergonomic approach to error handling. To make life easy for mediocre programmers, I recommend starting every project with an Error enum, that derives thiserror::Error, and using that in every method and function that returns a Result. How you structure your Error type from there is less cut-and-dried, but typically I ll create a separate enum variant for each type of error I want to have a different description. With thiserror, it s easy to then attach those descriptions:

#[derive(Clone, Debug, thiserror::Error)]
enum Error  
    #[error(" 0  caught fire")]
    Combustion(String),
    #[error(" 0  exploded")]
    Explosion(String),
 

I also implement functions to create each error variant, because that allows me to do the Into<String> trick, and can sometimes come in handy when creating errors from other places with .map_err() (more on that later). For example, the impl for the above Error would probably be:

impl Error  
    fn combustion(desc: impl Into<String>) -> Self  
        Self::Combustion(desc.into())
     
    fn explosion(desc: impl Into<String>) -> Self  
        Self::Explosion(desc.into())
     
 

It s a tedious bit of boilerplate, and you can use the thiserror-ext crate s thiserror_ext::Construct derive macro to do the hard work for you, if you like. It, too, knows all about the Into<String> trick.

Banish map_err (well, mostly) The newer mediocre programmer, who is just dipping their toe in the water of Rust, might write file handling code that looks like this:

fn read_u32_from_file(name: impl AsRef<str>) -> Result<u32, Error>  
    let mut f = File::open(name.as_ref())
        .map_err( e  Error::FileOpenError(name.as_ref().to_string(), e))?;
    let mut buf = vec![0u8; 30];
    f.read(&mut buf)
        .map_err( e  Error::ReadError(e))?;
    String::from_utf8(buf)
        .map_err( e  Error::EncodingError(e))?
        .parse::<u32>()
        .map_err( e  Error::ParseError(e))
 

This works great (or it probably does, I haven t actually tested it), but there are a lot of .map_err() calls in there. They take up over half the function, in fact. With the power of the From trait and the magic of the ? operator, we can make this a lot tidier. First off, assume we ve written boilerplate error creation functions (or used thiserror_ext::Construct to do it for us)). That allows us to simplify the file handling portion of the function a bit:

fn read_u32_from_file(name: impl AsRef<str>) -> Result<u32, Error>  
    let mut f = File::open(name.as_ref())
        // We've dropped the  .to_string()  out of here...
        .map_err( e  Error::file_open_error(name.as_ref(), e))?;
    let mut buf = vec![0u8; 30];
    f.read(&mut buf)
        // ... and the explicit parameter passing out of here
        .map_err(Error::read_error)?;
    // ...

If that latter .map_err() call looks weird, without the e and such, it s passing a function-as-closure, which just saves on a few characters typing. Just because we re mediocre, doesn t mean we re not also lazy. Next, if we implement the From trait for the other two errors, we can make the string-handling lines significantly cleaner. First, the trait impl:

impl From<std::string::FromUtf8Error> for Error  
    fn from(e: std::string::FromUtf8Error) -> Self  
        Self::EncodingError(e)
     
 
impl From<std::num::ParseIntError> for Error  
    fn from(e: std::num::ParseIntError) -> Self  
        Self::ParseError(e)
     
 

(Again, this is boilerplate that can be autogenerated, this time by adding a #[from] tag to the variants you want a From impl on, and thiserror will take care of it for you) In any event, no matter how you get the From impls, once you have them, the string-handling code becomes practically error-handling-free:

    Ok(
        String::from_utf8(buf)?
            .parse::<u32>()?
    )

The ? operator will automatically convert the error from the types returned from each method into the return error type, using From. The only tiny downside to this is that the ? at the end strips the Result, and so we ve got to wrap the returned value in Ok() to turn it back into a Result for returning. But I think that s a small price to pay for the removal of those .map_err() calls. In many cases, my coding process involves just putting a ? after every call that returns a Result, and adding a new Error variant whenever the compiler complains about not being able to convert some new error type. It s practically zero effort outstanding outcome for the mediocre programmer.

Just Because You re Mediocre, Doesn t Mean You Can t Get Better To finish off, I d like to point out that mediocrity doesn t imply shoddy work, nor does it mean that you shouldn t keep learning and improving your craft. One book that I ve recently found extremely helpful is Effective Rust, by David Drysdale. The author has very kindly put it up to read online, but buying a (paper or ebook) copy would no doubt be appreciated. The thing about this book, for me, is that it is very readable, even by us mediocre programmers. The sections are written in a way that really clicked with me. Some aspects of Rust that I d had trouble understanding for a long time such as lifetimes and the borrow checker, and particularly lifetime elision actually made sense after I d read the appropriate sections.

Finally, a Quick Beg I m currently subsisting on the kindness of strangers, so if you found something useful (or entertaining) in this post, why not buy me a refreshing beverage? It helps to know that people like what I m doing, and helps keep me from having to sell my soul to a private equity firm.

1. Voice input for dictation
2. Voice assistant like Google/Apple
3. Voice output
4. Full operation for visually impaired people

• [1] https://etbe.coker.com.au/2024/03/13/shape-computers/
• [2] https://www.youtube.com/watch?v=TitZV6k8zfA
• [3] https://etbe.coker.com.au/2023/10/11/pinephone-status/
• [4] https://wiki.debian.org/MobileApps
• [5] https://www.youtube.com/watch?v=0O2yTG3n1Vc
• [6] https://etbe.coker.com.au/2023/05/29/considering-convergence/
• [7] https://etbe.coker.com.au/2024/04/26/convergence-vs-transference/

#include <stdio.h>
int main()
 
  int   = 1,   = 2;
  printf(" =%d,  =%d\n",  ,  );
  return 0;
 

• [1] https://xkcd.com/1513/
• [2] https://www.w3schools.com/html/html_emojis.asp

There was no way of telling when the threshold would be reached. Perhaps not for hours, and perhaps the next moment. Biron remained standing helplessly, flashlight held loosely in his damp hands. Half an hour before, the visiphone had awakened him, and he had been at peace then. Now he knew he was going to die. Biron didn't want to die, but he was penned in hopelessly, and there was no place to hide.

• did some clerical work for the DebConf videoteam
• tried to book a plane ticket for DC24
• triaged #1067620 (dependency problem with whipper)
• closed #1067121 (flaky test in supysonic)
• closed #1065514 (qpdfview crossbuilding)

• organized his custom Application Manager stuff in a private repo
• did some coding for the New Maintainers website to make sure applicants agree with the Debian Code of Conduct alongside the SC/DFSG/DMUP
• closed #1057910 (new apticron upload)

• planned & brainstormed for the upcoming Debian usability tests
• mentored a student/new contributor (justin)
• babysat a future contributor!
• closed #1067649
• learnt about fabre.debian.net & element.debian.social (thanks, pollo!)

• uploaded puppet-strings 4.1.2-1 to unstable
• updated various services in personal infra
• cleaned vagrant-hostmanager and worked on packaging the new upstream release (1.8.10)
• extended GPG key expiry
• looked at options for a new backup machine

• updated puppetdb to 8.4.1

• opened #1068152 after a misfortune with #1068151
• created new contributor accounts (salsa & wiki)

 
  description = "Haskell dev MicroVM";
  inputs.impermanence.url = "github:nix-community/impermanence";
  inputs.microvm.url = "github:astro/microvm.nix";
  inputs.microvm.inputs.nixpkgs.follows = "nixpkgs";
  outputs =   self, impermanence, microvm, nixpkgs  :
    let
      persistencePath = "/persistent";
      system = "x86_64-linux";
      user = "thk";
      vmname = "haskell";
      nixosConfiguration = nixpkgs.lib.nixosSystem  
          inherit system;
          modules = [
            microvm.nixosModules.microvm
            impermanence.nixosModules.impermanence
            ( pkgs, ...  :  
            environment.persistence.$ persistencePath  =  
                hideMounts = true;
                users.$ user  =  
                  directories = [
                    "git" ".stack"
                  ];
                 ;
               ;
              environment.sessionVariables =  
                TERM = "screen-256color";
               ;
              environment.systemPackages = with pkgs; [
                ghc
                git
                (haskell-language-server.override   supportedGhcVersions = [ "94" ];  )
                htop
                stack
                tmux
                tree
                vcsh
                zsh
              ];
              fileSystems.$ persistencePath .neededForBoot = nixpkgs.lib.mkForce true;
              microvm =  
                forwardPorts = [
                    from = "host"; host.port = 2222; guest.port = 22;  
                    from = "guest"; host.port = 5432; guest.port = 5432;   # postgresql
                ];
                hypervisor = "qemu";
                interfaces = [
                    type = "user"; id = "usernet"; mac = "00:00:00:00:00:02";  
                ];
                mem = 4096;
                shares = [  
                  # use "virtiofs" for MicroVMs that are started by systemd
                  proto = "9p";
                  tag = "ro-store";
                  # a host's /nix/store will be picked up so that no
                  # squashfs/erofs will be built for it.
                  source = "/nix/store";
                  mountPoint = "/nix/.ro-store";
                   
                  proto = "virtiofs";
                  tag = "persistent";
                  source = "~/.local/share/microvm/vms/$ vmname /persistent";
                  mountPoint = persistencePath;
                  socket = "/run/user/1000/microvm-$ vmname -persistent";
                 
                ];
                socket = "/run/user/1000/microvm-control.socket";
                vcpu = 3;
                volumes = [];
                writableStoreOverlay = "/nix/.rwstore";
               ;
              networking.hostName = vmname;
              nix.enable = true;
              nix.nixPath = ["nixpkgs=$ builtins.storePath <nixpkgs> "];
              nix.settings =  
                extra-experimental-features = ["nix-command" "flakes"];
                trusted-users = [user];
               ;
              security.sudo =  
                enable = true;
                wheelNeedsPassword = false;
               ;
              services.getty.autologinUser = user;
              services.openssh =  
                enable = true;
               ;
              system.stateVersion = "24.11";
              systemd.services.loadnixdb =  
                description = "import hosts nix database";
                path = [pkgs.nix];
                wantedBy = ["multi-user.target"];
                requires = ["nix-daemon.service"];
                script = "cat $ persistencePath /nix-store-db-dump nix-store --load-db";
               ;
              time.timeZone = nixpkgs.lib.mkDefault "Europe/Berlin";
              users.users.$ user  =  
                extraGroups = [ "wheel" "video" ];
                group = "user";
                isNormalUser = true;
                openssh.authorizedKeys.keys = [
                  "ssh-rsa REDACTED"
                ];
                password = "";
               ;
              users.users.root.password = "";
              users.groups.user =  ;
             )
          ];
         ;
    in  
      packages.$ system .default = nixosConfiguration.config.microvm.declaredRunner;
     ;
 

[Unit]
Description=MicroVM for Haskell development
Requires=microvm-virtiofsd-persistent@.service
After=microvm-virtiofsd-persistent@.service
AssertFileNotEmpty=%h/.local/share/microvm/vms/%i/flake/flake.nix
[Service]
Type=forking
ExecStartPre=/usr/bin/sh -c "[ /nix/var/nix/db/db.sqlite -ot %h/.local/share/microvm/nix-store-db-dump ]   nix-store --dump-db >%h/.local/share/microvm/nix-store-db-dump"
ExecStartPre=ln -f -t %h/.local/share/microvm/vms/%i/persistent/ %h/.local/share/microvm/nix-store-db-dump
ExecStartPre=-%h/.local/state/nix/profile/bin/tmux new -s microvm -d
ExecStart=%h/.local/state/nix/profile/bin/tmux new-window -t microvm: -n "%i" "exec %h/.local/state/nix/profile/bin/nix run --impure %h/.local/share/microvm/vms/%i/flake"

[Unit]
Description=serve host persistent folder for dev VM
AssertPathIsDirectory=%h/.local/share/microvm/vms/%i/persistent
[Service]
ExecStart=%h/.local/state/nix/profile/bin/virtiofsd \
 --socket-path=$ XDG_RUNTIME_DIR /microvm-%i-persistent \
 --shared-dir=%h/.local/share/microvm/vms/%i/persistent \
 --gid-map :995:%G:1: \
 --uid-map :1000:%U:1:

Back to the Future Time to admit something. I really love Plan 9. It s just so good. So many ideas from Plan 9 are just so prescient, and everything just feels right. Not just right like, feels good like, correct. The bit that I ve always liked the most is 9p, the network protocol for serving a filesystem over a network. This leads to all sorts of fun programs, like the Plan 9 ftp client being a 9p server you mount the ftp server and access files like any other files. It s kinda like if fuse were more fully a part of how the operating system worked, but fuse is all running client-side. With 9p there s a single client, and different servers that you can connect to, which may be backed by a hard drive, remote resources over something like SFTP, FTP, HTTP or even purely synthetic. The interesting (maybe sad?) part here is that 9p wound up outliving Plan 9 in terms of adoption 9p is in all sorts of places folks don t usually expect. For instance, the Windows Subsystem for Linux uses the 9p protocol to share files between Windows and Linux. ChromeOS uses it to share files with Crostini, and qemu uses 9p (virtio-p9) to share files between guest and host. If you re noticing a pattern here, you d be right; for some reason 9p is the go-to protocol to exchange files between hypervisor and guest. Why? I have no idea, except maybe due to being designed well, simple to implement, and it s a lot easier to validate the data being shared and validate security boundaries. Simplicity has its value. As a result, there s a lot of lingering 9p support kicking around. Turns out Linux can even handle mounting 9p filesystems out of the box. This means that I can deploy a filesystem to my LAN or my localhost by running a process on top of a computer that needs nothing special, and mount it over the network on an unmodified machine unlike fuse, where you d need client-specific software to run in order to mount the directory. For instance, let s mount a 9p filesystem running on my localhost machine, serving requests on 127.0.0.1:564 (tcp) that goes by the name mountpointname to /mnt.

$ mount -t 9p \
-o trans=tcp,port=564,version=9p2000.u,aname=mountpointname \
127.0.0.1 \
/mnt

Linux will mount away, and attach to the filesystem as the root user, and by default, attach to that mountpoint again for each local user that attempts to use it. Nifty, right? I think so. The server is able to keep track of per-user access and authorization along with the host OS.

WHEREIN I STYX WITH IT Since I wanted to push myself a bit more with rust and tokio specifically, I opted to implement the whole stack myself, without third party libraries on the critical path where I could avoid it. The 9p protocol (sometimes called Styx, the original name for it) is incredibly simple. It s a series of client to server requests, which receive a server to client response. These are, respectively, T messages, which transmit a request to the server, which trigger an R message in response (Reply messages). These messages are TLV payload with a very straight forward structure so straight forward, in fact, that I was able to implement a working server off nothing more than a handful of man pages. Later on after the basics worked, I found a more complete spec page that contains more information about the unix specific variant that I opted to use (9P2000.u rather than 9P2000) due to the level of Linux specific support for the 9P2000.u variant over the 9P2000 protocol.

MR ROBOTO The backend stack over at zoo is rust and tokio running i/o for an HTTP and WebRTC server. I figured I d pick something fairly similar to write my filesystem with, since 9P can be implemented on basically anything with I/O. That means tokio tcp server bits, which construct and use a 9p server, which has an idiomatic Rusty API that partially abstracts the raw R and T messages, but not so much as to cause issues with hiding implementation possibilities. At each abstraction level, there s an escape hatch allowing someone to implement any of the layers if required. I called this framework arigato which can be found over on docs.rs and crates.io.

/// Simplified version of the arigato File trait; this isn't actually
/// the same trait; there's some small cosmetic differences. The
/// actual trait can be found at:
///
/// https://docs.rs/arigato/latest/arigato/server/trait.File.html
trait File  
/// OpenFile is the type returned by this File via an Open call.
 type OpenFile: OpenFile;
/// Return the 9p Qid for this file. A file is the same if the Qid is
 /// the same. A Qid contains information about the mode of the file,
 /// version of the file, and a unique 64 bit identifier.
 fn qid(&self) -> Qid;
/// Construct the 9p Stat struct with metadata about a file.
 async fn stat(&self) -> FileResult<Stat>;
/// Attempt to update the file metadata.
 async fn wstat(&mut self, s: &Stat) -> FileResult<()>;
/// Traverse the filesystem tree.
 async fn walk(&self, path: &[&str]) -> FileResult<(Option<Self>, Vec<Self>)>;
/// Request that a file's reference be removed from the file tree.
 async fn unlink(&mut self) -> FileResult<()>;
/// Create a file at a specific location in the file tree.
 async fn create(
&mut self,
name: &str,
perm: u16,
ty: FileType,
mode: OpenMode,
extension: &str,
) -> FileResult<Self>;
/// Open the File, returning a handle to the open file, which handles
 /// file i/o. This is split into a second type since it is genuinely
 /// unrelated -- and the fact that a file is Open or Closed can be
 /// handled by the  arigato  server for us.
 async fn open(&mut self, mode: OpenMode) -> FileResult<Self::OpenFile>;
 
/// Simplified version of the arigato OpenFile trait; this isn't actually
/// the same trait; there's some small cosmetic differences. The
/// actual trait can be found at:
///
/// https://docs.rs/arigato/latest/arigato/server/trait.OpenFile.html
trait OpenFile  
/// iounit to report for this file. The iounit reported is used for Read
 /// or Write operations to signal, if non-zero, the maximum size that is
 /// guaranteed to be transferred atomically.
 fn iounit(&self) -> u32;
/// Read some number of bytes up to  buf.len()  from the provided
 ///  offset  of the underlying file. The number of bytes read is
 /// returned.
 async fn read_at(
&mut self,
buf: &mut [u8],
offset: u64,
) -> FileResult<u32>;
/// Write some number of bytes up to  buf.len()  from the provided
 ///  offset  of the underlying file. The number of bytes written
 /// is returned.
 fn write_at(
&mut self,
buf: &mut [u8],
offset: u64,
) -> FileResult<u32>;
 

Thanks, decade ago paultag! Let s do it! Let s use arigato to implement a 9p filesystem we ll call debugfs that will serve all the debug files shipped according to the Packages metadata from the apt archive. We ll fetch the Packages file and construct a filesystem based on the reported Build-Id entries. For those who don t know much about how an apt repo works, here s the 2-second crash course on what we re doing. The first is to fetch the Packages file, which is specific to a binary architecture (such as amd64, arm64 or riscv64). That architecture is specific to a component (such as main, contrib or non-free). That component is specific to a suite, such as stable, unstable or any of its aliases (bullseye, bookworm, etc). Let s take a look at the Packages.xz file for the unstable-debug suite, main component, for all amd64 binaries.

$ curl \
https://deb.debian.org/debian-debug/dists/unstable-debug/main/binary-amd64/Packages.xz \
  unxz

This will return the Debian-style rfc2822-like headers, which is an export of the metadata contained inside each .deb file which apt (or other tools that can use the apt repo format) use to fetch information about debs. Let s take a look at the debug headers for the netlabel-tools package in unstable which is a package named netlabel-tools-dbgsym in unstable-debug.

Package: netlabel-tools-dbgsym
Source: netlabel-tools (0.30.0-1)
Version: 0.30.0-1+b1
Installed-Size: 79
Maintainer: Paul Tagliamonte <paultag@debian.org>
Architecture: amd64
Depends: netlabel-tools (= 0.30.0-1+b1)
Description: debug symbols for netlabel-tools
Auto-Built-Package: debug-symbols
Build-Ids: e59f81f6573dadd5d95a6e4474d9388ab2777e2a
Description-md5: a0e587a0cf730c88a4010f78562e6db7
Section: debug
Priority: optional
Filename: pool/main/n/netlabel-tools/netlabel-tools-dbgsym_0.30.0-1+b1_amd64.deb
Size: 62776
SHA256: 0e9bdb087617f0350995a84fb9aa84541bc4df45c6cd717f2157aa83711d0c60

So here, we can parse the package headers in the Packages.xz file, and store, for each Build-Id, the Filename where we can fetch the .deb at. Each .deb contains a number of files but we re only really interested in the files inside the .deb located at or under /usr/lib/debug/.build-id/, which you can find in debugfs under rfc822.rs. It s crude, and very single-purpose, but I m feeling a bit lazy.

Who needs dpkg?! For folks who haven t seen it yet, a .deb file is a special type of .ar file, that contains (usually) three files inside debian-binary, control.tar.xz and data.tar.xz. The core of an .ar file is a fixed size (60 byte) entry header, followed by the specified size number of bytes.

[8 byte .ar file magic]
[60 byte entry header]
[N bytes of data]
[60 byte entry header]
[N bytes of data]
[60 byte entry header]
[N bytes of data]
...

First up was to implement a basic ar parser in ar.rs. Before we get into using it to parse a deb, as a quick diversion, let s break apart a .deb file by hand something that is a bit of a rite of passage (or at least it used to be? I m getting old) during the Debian nm (new member) process, to take a look at where exactly the .debug file lives inside the .deb file.

$ ar x netlabel-tools-dbgsym_0.30.0-1+b1_amd64.deb
$ ls
control.tar.xz debian-binary
data.tar.xz netlabel-tools-dbgsym_0.30.0-1+b1_amd64.deb
$ tar --list -f data.tar.xz   grep '.debug$'
./usr/lib/debug/.build-id/e5/9f81f6573dadd5d95a6e4474d9388ab2777e2a.debug

Since we know quite a bit about the structure of a .deb file, and I had to implement support from scratch anyway, I opted to implement a (very!) basic debfile parser using HTTP Range requests. HTTP Range requests, if supported by the server (denoted by a accept-ranges: bytes HTTP header in response to an HTTP HEAD request to that file) means that we can add a header such as range: bytes=8-68 to specifically request that the returned GET body be the byte range provided (in the above case, the bytes starting from byte offset 8 until byte offset 68). This means we can fetch just the ar file entry from the .deb file until we get to the file inside the .deb we are interested in (in our case, the data.tar.xz file) at which point we can request the body of that file with a final range request. I wound up writing a struct to handle a read_at-style API surface in hrange.rs, which we can pair with ar.rs above and start to find our data in the .deb remotely without downloading and unpacking the .deb at all. After we have the body of the data.tar.xz coming back through the HTTP response, we get to pipe it through an xz decompressor (this kinda sucked in Rust, since a tokio AsyncRead is not the same as an http Body response is not the same as std::io::Read, is not the same as an async (or sync) Iterator is not the same as what the xz2 crate expects; leading me to read blocks of data to a buffer and stuff them through the decoder by looping over the buffer for each lzma2 packet in a loop), and tarfile parser (similarly troublesome). From there we get to iterate over all entries in the tarfile, stopping when we reach our file of interest. Since we can t seek, but gdb needs to, we ll pull it out of the stream into a Cursor<Vec<u8>> in-memory and pass a handle to it back to the user. From here on out its a matter of gluing together a File traited struct in debugfs, and serving the filesystem over TCP using arigato. Done deal!

A quick diversion about compression I was originally hoping to avoid transferring the whole tar file over the network (and therefore also reading the whole debug file into ram, which objectively sucks), but quickly hit issues with figuring out a way around seeking around an xz file. What s interesting is xz has a great primitive to solve this specific problem (specifically, use a block size that allows you to seek to the block as close to your desired seek position just before it, only discarding at most block size - 1 bytes), but data.tar.xz files generated by dpkg appear to have a single mega-huge block for the whole file. I don t know why I would have expected any different, in retrospect. That means that this now devolves into the base case of How do I seek around an lzma2 compressed data stream ; which is a lot more complex of a question. Thankfully, notoriously brilliant tianon was nice enough to introduce me to Jon Johnson who did something super similar adapted a technique to seek inside a compressed gzip file, which lets his service oci.dag.dev seek through Docker container images super fast based on some prior work such as soci-snapshotter, gztool, and zran.c. He also pulled this party trick off for apk based distros over at apk.dag.dev, which seems apropos. Jon was nice enough to publish a lot of his work on this specifically in a central place under the name targz on his GitHub, which has been a ton of fun to read through. The gist is that, by dumping the decompressor s state (window of previous bytes, in-memory data derived from the last N-1 bytes) at specific checkpoints along with the compressed data stream offset in bytes and decompressed offset in bytes, one can seek to that checkpoint in the compressed stream and pick up where you left off creating a similar block mechanism against the wishes of gzip. It means you d need to do an O(n) run over the file, but every request after that will be sped up according to the number of checkpoints you ve taken. Given the complexity of xz and lzma2, I don t think this is possible for me at the moment especially given most of the files I ll be requesting will not be loaded from again especially when I can just cache the debug header by Build-Id. I want to implement this (because I m generally curious and Jon has a way of getting someone excited about compression schemes, which is not a sentence I thought I d ever say out loud), but for now I m going to move on without this optimization. Such a shame, since it kills a lot of the work that went into seeking around the .deb file in the first place, given the debian-binary and control.tar.gz members are so small.

The Good First, the good news right? It works! That s pretty cool. I m positive my younger self would be amused and happy to see this working; as is current day paultag. Let s take debugfs out for a spin! First, we need to mount the filesystem. It even works on an entirely unmodified, stock Debian box on my LAN, which is huge. Let s take it for a spin:

$ mount \
-t 9p \
-o trans=tcp,version=9p2000.u,aname=unstable-debug \
192.168.0.2 \
/usr/lib/debug/.build-id/

And, let s prove to ourselves that this actually mounted before we go trying to use it:

$ mount   grep build-id
192.168.0.2 on /usr/lib/debug/.build-id type 9p (rw,relatime,aname=unstable-debug,access=user,trans=tcp,version=9p2000.u,port=564)

Slick. We ve got an open connection to the server, where our host will keep a connection alive as root, attached to the filesystem provided in aname. Let s take a look at it.

$ ls /usr/lib/debug/.build-id/
00 0d 1a 27 34 41 4e 5b 68 75 82 8E 9b a8 b5 c2 CE db e7 f3
01 0e 1b 28 35 42 4f 5c 69 76 83 8f 9c a9 b6 c3 cf dc E7 f4
02 0f 1c 29 36 43 50 5d 6a 77 84 90 9d aa b7 c4 d0 dd e8 f5
03 10 1d 2a 37 44 51 5e 6b 78 85 91 9e ab b8 c5 d1 de e9 f6
04 11 1e 2b 38 45 52 5f 6c 79 86 92 9f ac b9 c6 d2 df ea f7
05 12 1f 2c 39 46 53 60 6d 7a 87 93 a0 ad ba c7 d3 e0 eb f8
06 13 20 2d 3a 47 54 61 6e 7b 88 94 a1 ae bb c8 d4 e1 ec f9
07 14 21 2e 3b 48 55 62 6f 7c 89 95 a2 af bc c9 d5 e2 ed fa
08 15 22 2f 3c 49 56 63 70 7d 8a 96 a3 b0 bd ca d6 e3 ee fb
09 16 23 30 3d 4a 57 64 71 7e 8b 97 a4 b1 be cb d7 e4 ef fc
0a 17 24 31 3e 4b 58 65 72 7f 8c 98 a5 b2 bf cc d8 E4 f0 fd
0b 18 25 32 3f 4c 59 66 73 80 8d 99 a6 b3 c0 cd d9 e5 f1 fe
0c 19 26 33 40 4d 5a 67 74 81 8e 9a a7 b4 c1 ce da e6 f2 ff

Outstanding. Let s try using gdb to debug a binary that was provided by the Debian archive, and see if it ll load the ELF by build-id from the right .deb in the unstable-debug suite:

$ gdb -q /usr/sbin/netlabelctl
Reading symbols from /usr/sbin/netlabelctl...
Reading symbols from /usr/lib/debug/.build-id/e5/9f81f6573dadd5d95a6e4474d9388ab2777e2a.debug...
(gdb)

Yes! Yes it will!

$ file /usr/lib/debug/.build-id/e5/9f81f6573dadd5d95a6e4474d9388ab2777e2a.debug
/usr/lib/debug/.build-id/e5/9f81f6573dadd5d95a6e4474d9388ab2777e2a.debug: ELF 64-bit LSB shared object, x86-64, version 1 (SYSV), dynamically linked, interpreter *empty*, BuildID[sha1]=e59f81f6573dadd5d95a6e4474d9388ab2777e2a, for GNU/Linux 3.2.0, with debug_info, not stripped

The Bad Linux s support for 9p is mainline, which is great, but it s not robust. Network issues or server restarts will wedge the mountpoint (Linux can t reconnect when the tcp connection breaks), and things that work fine on local filesystems get translated in a way that causes a lot of network chatter for instance, just due to the way the syscalls are translated, doing an ls, will result in a stat call for each file in the directory, even though linux had just got a stat entry for every file while it was resolving directory names. On top of that, Linux will serialize all I/O with the server, so there s no concurrent requests for file information, writes, or reads pending at the same time to the server; and read and write throughput will degrade as latency increases due to increasing round-trip time, even though there are offsets included in the read and write calls. It works well enough, but is frustrating to run up against, since there s not a lot you can do server-side to help with this beyond implementing the 9P2000.L variant (which, maybe is worth it).

The Ugly Unfortunately, we don t know the file size(s) until we ve actually opened the underlying tar file and found the correct member, so for most files, we don t know the real size to report when getting a stat. We can t parse the tarfiles for every stat call, since that d make ls even slower (bummer). Only hiccup is that when I report a filesize of zero, gdb throws a bit of a fit; let s try with a size of 0 to start:

$ ls -lah /usr/lib/debug/.build-id/e5/9f81f6573dadd5d95a6e4474d9388ab2777e2a.debug
-r--r--r-- 1 root root 0 Dec 31 1969 /usr/lib/debug/.build-id/e5/9f81f6573dadd5d95a6e4474d9388ab2777e2a.debug
$ gdb -q /usr/sbin/netlabelctl
Reading symbols from /usr/sbin/netlabelctl...
Reading symbols from /usr/lib/debug/.build-id/e5/9f81f6573dadd5d95a6e4474d9388ab2777e2a.debug...
warning: Discarding section .note.gnu.build-id which has a section size (24) larger than the file size [in module /usr/lib/debug/.build-id/e5/9f81f6573dadd5d95a6e4474d9388ab2777e2a.debug]
[...]

This obviously won t work since gdb will throw away all our hard work because of stat s output, and neither will loading the real size of the underlying file. That only leaves us with hardcoding a file size and hope nothing else breaks significantly as a result. Let s try it again:

$ ls -lah /usr/lib/debug/.build-id/e5/9f81f6573dadd5d95a6e4474d9388ab2777e2a.debug
-r--r--r-- 1 root root 954M Dec 31 1969 /usr/lib/debug/.build-id/e5/9f81f6573dadd5d95a6e4474d9388ab2777e2a.debug
$ gdb -q /usr/sbin/netlabelctl
Reading symbols from /usr/sbin/netlabelctl...
Reading symbols from /usr/lib/debug/.build-id/e5/9f81f6573dadd5d95a6e4474d9388ab2777e2a.debug...
(gdb)

Much better. I mean, terrible but better. Better for now, anyway.

Kilroy was here Do I think this is a particularly good idea? I mean; kinda. I m probably going to make some fun 9p arigato-based filesystems for use around my LAN, but I don t think I ll be moving to use debugfs until I can figure out how to ensure the connection is more resilient to changing networks, server restarts and fixes on i/o performance. I think it was a useful exercise and is a pretty great hack, but I don t think this ll be shipping anywhere anytime soon. Along with me publishing this post, I ve pushed up all my repos; so you should be able to play along at home! There s a lot more work to be done on arigato; but it does handshake and successfully export a working 9P2000.u filesystem. Check it out on on my github at arigato, debugfs and also on crates.io and docs.rs. At least I can say I was here and I got it working after all these years.

File /usr/lib/python3/dist-packages/django/contrib/auth/mixins.py , line 5, in
from django.contrib.auth.views import redirect_to_login
File /usr/lib/python3/dist-packages/django/contrib/auth/views.py , line 20, in
from django.utils.http import (
ImportError: cannot import name url_has_allowed_host_and_scheme from django.utils.http (/usr/lib/python3/dist-packages/django/utils/http.py) During handling of the above exception, another exception occurred:

• apt remove mailman3-web
  • keep db and config files (do not purge)
• apt autoremove
  • remove django related packages
• apt install mailman3-web mailman3-full

Prelude to an Adventure Our story begins back in March 2008. I was working at Engine Yard (EY), a now largely-forgotten Rails-focused hosting company, which pioneered several advances in Rails application deployment. Probably EY s greatest claim to lasting fame is that they helped launch a little code hosting platform you might have heard of, by providing them free infrastructure when they were little more than a glimmer in the Internet s eye. I am, of course, talking about everyone s favourite Microsoft product: GitHub. Since GitHub was in the right place, at the right time, with a compelling product offering, they quickly started to gain traction, and grow their userbase. With growth comes challenges, amongst them the one we re focusing on today: SSH login times. Then, as now, GitHub provided SSH access to the git repos they hosted, by SSHing to git@github.com with publickey authentication. They were using the standard way that everyone manages SSH keys: the ~/.ssh/authorized_keys file, and that became a problem as the number of keys started to grow. The way that SSH uses this file is that, when a user connects and asks for publickey authentication, SSH opens the ~/.ssh/authorized_keys file and scans all of the keys listed in it, looking for a key which matches the key that the user presented. This linear search is normally not a huge problem, because nobody in their right mind puts more than a few keys in their ~/.ssh/authorized_keys, right?

Of course, as a popular, rapidly-growing service, GitHub was gaining users at a fair clip, to the point that the one big file that stored all the SSH keys was starting to visibly impact SSH login times. This problem was also not going to get any better by itself. Something Had To Be Done. EY management was keen on making sure GitHub ran well, and so despite it not really being a hosting problem, they were willing to help fix this problem. For some reason, the late, great, Ezra Zygmuntowitz pointed GitHub in my direction, and let me take the time to really get into the problem with the GitHub team. After examining a variety of different possible solutions, we came to the conclusion that the least-worst option was to patch OpenSSH to lookup keys in a MySQL database, indexed on the key fingerprint. We didn t take this decision on a whim it wasn t a case of yeah, sure, let s just hack around with OpenSSH, what could possibly go wrong? . We knew it was potentially catastrophic if things went sideways, so you can imagine how much worse the other options available were. Ensuring that this wouldn t compromise security was a lot of the effort that went into the change. In the end, though, we rolled it out in early April, and lo! SSH logins were fast, and we were pretty sure we wouldn t have to worry about this problem for a long time to come. Normally, you d think patching OpenSSH to make mass SSH logins super fast would be a good story on its own. But no, this is just the opening scene.

Chekov s Gun Makes its Appearance Fast forward a little under a month, to the first few days of May 2008. I get a message from one of the GitHub team, saying that somehow users were able to access other users repos over SSH. Naturally, as we d recently rolled out the OpenSSH patch, which touched this very thing, the code I d written was suspect number one, so I was called in to help.

They're called The Usual Suspects for a reason, but sometimes, it really is Keyser S ze

Eventually, after more than a little debugging, we discovered that, somehow, there were two users with keys that had the same key fingerprint. This absolutely shouldn t happen it s a bit like winning the lottery twice in a row¹ unless the users had somehow shared their keys with each other, of course. Still, it was worth investigating, just in case it was a web application bug, so the GitHub team reached out to the users impacted, to try and figure out what was going on. The users professed no knowledge of each other, neither admitted to publicising their key, and couldn t offer any explanation as to how the other person could possibly have gotten their key. Then things went from weird to what the ? . Because another pair of users showed up, sharing a key fingerprint but it was a different shared key fingerprint. The odds now have gone from winning the lottery multiple times in a row to as close to this literally cannot happen as makes no difference.

Once we were really, really confident that the OpenSSH patch wasn t the cause of the problem, my involvement in the problem basically ended. I wasn t a GitHub employee, and EY had plenty of other customers who needed my help, so I wasn t able to stay deeply involved in the on-going investigation of The Mystery of the Duplicate Keys. However, the GitHub team did keep talking to the users involved, and managed to determine the only apparent common factor was that all the users claimed to be using Debian or Ubuntu systems, which was where their SSH keys would have been generated. That was as far as the investigation had really gotten, when along came May 13, 2008.

Chekov s Gun Goes Off With the publication of DSA-1571-1, everything suddenly became clear. Through a well-meaning but ultimately disasterous cleanup of OpenSSL s randomness generation code, the Debian maintainer had inadvertently reduced the number of possible keys that could be generated by a given user from bazillions to a little over 32,000. With so many people signing up to GitHub some of them no doubt following best practice and freshly generating a separate key it s unsurprising that some collisions occurred. You can imagine the sense of oooooooh, so that s what s going on! that rippled out once the issue was understood. I was mostly glad that we had conclusive evidence that my OpenSSH patch wasn t at fault, little knowing how much more contact I was to have with Debian weak keys in the future, running a huge store of known-compromised keys and using them to find misbehaving Certificate Authorities, amongst other things.

Lessons Learned While I ve not found a description of exactly when and how Luciano Bello discovered the vulnerability that became CVE-2008-0166, I presume he first came across it some time before it was disclosed likely before GitHub tripped over it. The stable Debian release that included the vulnerable code had been released a year earlier, so there was plenty of time for Luciano to have discovered key collisions and go hmm, I wonder what s going on here? , then keep digging until the solution presented itself. The thought hmm, that s odd , followed by intense investigation, leading to the discovery of a major flaw is also what ultimately brought down the recent XZ backdoor. The critical part of that sequence is the ability to do that intense investigation, though. When I reflect on my brush with the Debian weak keys vulnerability, what sticks out to me is the fact that I didn t do the deep investigation. I wonder if Luciano hadn t found it, how long it might have been before it was found. The GitHub team would have continued investigating, presumably, and perhaps they (or I) would have eventually dug deep enough to find it. But we were all super busy myself, working support tickets at EY, and GitHub feverishly building features and fighting the fires in their rapidly-growing service. As it was, Luciano was able to take the time to dig in and find out what was happening, but just like the XZ backdoor, I feel like we, as an industry, got a bit lucky that someone with the skills, time, and energy was on hand at the right time to make a huge difference. It s a luxury to be able to take the time to really dig into a problem, and it s a luxury that most of us rarely have. Perhaps an understated takeaway is that somehow we all need to wrestle back some time to follow our hunches and really dig into the things that make us go hmm .

Support My Hunches If you d like to help me be able to do intense investigations of mysterious software phenomena, you can shout me a refreshing beverage on ko-fi.

---

1. the odds are actually probably more like winning the lottery about twenty times in a row. The numbers involved are staggeringly huge, so it s easiest to just approximate it as really, really unlikely .

from __future__ import annotations
from dataclasses import dataclass
from dataclasses_deb822 import read_deb822, field_deb822

@dataclass
class Package:

    package: str = field_deb822('Package')
    version: str = field_deb822('Version')
    arch: str = field_deb822('Architecture')

    multi_arch: Optional[str] = field_deb822(
        'Multi-Arch',
        default=None,
    )

for p in read_deb822(Package, sys.stdin, ignore_unknown=True):
    print(p)

from __future__ import annotations
from dataclasses import dataclass
from debian_cloud_images.utils.dataclasses_deb822 import read_deb822, field_deb822
from typing import Optional
import sys
@dataclass
class Package:
    package: str = field_deb822('Package')
    version: str = field_deb822('Version')
    arch: str = field_deb822('Architecture')
    multi_arch: Optional[str] = field_deb822(
        'Multi-Arch',
        default=None,
    )
for p in read_deb822(Package, sys.stdin, ignore_unknown=True):
    print(p)

• Needs postponed evaluation of annotations.
• Does not allow for different case of keys.
• Fields are stored with continuation character embedded (aka: text\n text').

$ sudo apt update
$ sudo apt install flatpak

$ flatpak remote-add --if-not-exists flathub https://dl.flathub.org/repo/flathub.flatpakrepo

$ which -s gnome-software  && sudo apt install gnome-software-plugin-flatpak
$ which -s plasma-discover && sudo apt install plasma-discover-backend-flatpak

$ dpkg -l   grep xdg-desktop-portal   awk ' print $2 '
xdg-desktop-portal
xdg-desktop-portal-gnome
xdg-desktop-portal-gtk

Flatpak commands are run system-wide by default. If you are installing applications for day-to-day usage, it is recommended to stick with this default behavior.

$ flatpak install flathub org.mozilla.firefox

$ flatpak run org.mozilla.firefox

• from ~/.mozilla/ -- where the Firefox Debian package stores its data
• into ~/.var/app/org.mozilla.firefox/.mozilla/ -- where the Firefox Flatpak app stores its data

# BEWARE! Below I'm erasing data!
$ rm -fr ~/.var/app/org.mozilla.firefox/.mozilla/firefox/
$ cp -a ~/.mozilla/firefox/ ~/.var/app/org.mozilla.firefox/.mozilla/

$ mv ~/.mozilla/firefox ~/.mozilla/firefox.old.$(date --iso-8601=date)

$ cat /usr/local/bin/firefox 
#!/bin/sh
exec flatpak run org.mozilla.firefox "$@"

$ sudo apt update
$ sudo apt full-upgrade

flatpak update [OPTION...] [REF...] Updates applications and runtimes. [...] If no REF is given, everything is updated, as well as appstream info for all remotes.

$ flatpak update

1. Teach myself to run flatpak update additionally to apt update, manually, everytime I update my system.
2. Go crazy: let something automatically update my Flatpak apps, in my back and without my consent.

1. Reproducible tarballs. The idea is that published source tarballs should be possible to reproduce independently somehow, and that this should be continuously tested and verified preferrably as part of the upstream project continuous integration system (e.g., GitHub action or GitLab pipeline). While nominally this looks easy to achieve, there are some complex matters in this, for example: what timestamps to use for files in the tarball? I ve brought up this aspect before.
2. Minimal source tarballs without generated vendor files. Most GNU Autoconf/Automake-based tarballs pre-generated files which are important for bootstrapping on exotic systems that does not have the required dependencies. For the bootstrapping story to succeed, this approach is important to support. However it has become clear that this practice raise significant costs and risks. Most modern GNU/Linux distributions have all the required dependencies and actually prefers to re-build everything from source code. These pre-generated extra files introduce uncertainty to that process.

1. The tarball should allow users to build the project, which is the entire purpose of all this. This means that at least all source code for the project has to be included.
2. The tarballs should be signed, for example with PGP or minisign.
3. The tarball should be possible to reproduce bit-by-bit by a third party using upstream s version controlled sources and a pointer to which revision was used (e.g., git tag or git commit).
4. The tarball should not require an Internet connection to download things.
   • Corollary: every external dependency either has to be explicitly documented as such (e.g., gcc and GnuTLS), or included in the tarball.
   • Observation: This means including all *.po gettext translations which are normally downloaded when building from version controlled sources.
5. The tarball should contain everything required to build the project from source using as much externally released versioned tooling as possible. This is the minimal property lacking today.
   • Corollary: This means including a vendored copy of OpenSSL or libz is not acceptable: link to them as external projects.
   • Open question: How about non-released external tooling such as gnulib or autoconf archive macros? This is a bit more delicate: most distributions either just package one current version of gnulib or autoconf archive, not previous versions. While this could change, and distributions could package the gnulib git repository (up to some current version) and the autoconf archive git repository and packages were set up to extract the version they need (gnulib s ./bootstrap already supports this via the gnulib-refdir parameter), this is not normally in place.
   • Suggested Corollary: The tarball should contain content from git submodule s such as gnulib and the necessary Autoconf archive M4 macros required by the project.
6. Similar to how the GNU project specify the ./configure interface we need a documented interface for how to bootstrap the project. I suggest to use the already well established idiom of running ./bootstrap to set up the package to later be able to be built via ./configure. Of course, some projects are not using the autotool ./configure interface and will not follow this aspect either, but like most build systems that compete with autotools have instructions on how to build the project, they should document similar interfaces for bootstrapping the source tarball to allow building.

Work report Kubuntu: Bug bashing! I am triaging allthebugs for Plasma which can be seen here: https://bugs.launchpad.net/plasma-5.27/+bug/2053125 I am happy to report many of the remaining bugs have been fixed in the latest bug fix release 5.27.11. I prepared https://kde.org/announcements/plasma/5/5.27.11/ and Rik uploaded to archive, thank you. Unfortunately, this and several other key fixes are stuck in transition do to the time_t64 transition, which you can read about here: https://wiki.debian.org/ReleaseGoals/64bit-time . It is the biggest transition in Debian/Ubuntu history and it couldn t come at a worst time. We are aware our ISO installer is currently broken, calamares is one of those things stuck in this transition. There is a workaround in the comments of the bug report: https://bugs.launchpad.net/ubuntu/+source/calamares/+bug/2054795 Fixed an issue with plasma-welcome. Found the fix for emojis and Aaron has kindly moved this forward with the fontconfig maintainer. Thanks! I have received an https://kfocus.org/spec/spec-ir14.html laptop and it is truly a great machine and is now my daily driver. A big thank you to the Kfocus team! I can t wait to show it off at https://linuxfestnorthwest.org/. KDE Snaps: You will see the activity in this ramp back up as the KDEneon Core project is finally a go! I will participate in the project with part time status and get everyone in the Enokia team up to speed with my snap knowledge, help prepare the qt6/kf6 transition, package plasma, and most importantly I will focus on documentation for future contributors. I have created the ( now split ) qt6 with KDE patchset support and KDE frameworks 6 SDK and runtime snaps. I have made the kde-neon-6 extension and the PR is in: https://github.com/canonical/snapcraft/pull/4698 . Future work on the extension will include multiple versions track support and core24 support.

A welcome sign at Bangkok's Suvarnabhumi airport.

Bus from Suvarnabhumi Airport to Jomtien Beach in Pattaya.

Road near Jomtien beach in Pattaya

Photo of a songthaew in Pattaya. There are shared songthaews which run along Jomtien Second road and takes 10 bath to anywhere on the route.

Jomtien Beach in Pattaya.

A welcome sign at Pattaya Floating market.

This Korean Vegetasty noodles pack was yummy and was available at many 7-Eleven stores.

Wat Arun temple stamps your hand upon entry

Wat Arun temple

Khao San Road

A food stall at Khao San Road

Chao Phraya Express Boat

Banana with yellow flesh

Fruits at a stall in Bangkok

Trimmed pineapples from Thailand.

Corn in Bangkok.

A board showing coffee menu at a 7-Eleven store along with rates in Pattaya.

In this section of 7-Eleven, you can buy a premix coffee and mix it with hot water provided at the store to prepare.

Red wine being served in Air India

Notes

• In this whole trip spanning two weeks, I did not pay for drinking water (except for once in Pattaya which was 9 baht) and toilets. Bangkok and Kuala Lumpur have plenty of malls where you should find a free-of-cost toilet nearby. For drinking water, I relied mainly on my accommodation providing refillable water for my bottle.
• Thailand seemed more expensive than Malaysia on average. Malaysia had discounted price due to the Chinese New year.
• I liked Pattaya more than Bangkok. Maybe because Pattaya has beach and Bangkok doesn t. Pattaya seemed more lively, and I could meet and talk to a few people as opposed to Bangkok.
• Chao Phraya River express boat costs 150 baht for one day where you can hop on and off to any boat.