On 22 Oct a cancelled job left its analysis server running in the
runner, 4 GB, until someone noticed. forgejo-runner v13.1.0 starts every
host step as a process group and sends the group SIGKILL on a cancel or
at timeout-minutes (act/common/cmdgroup.go), and after that no later step
runs, `if: always()` ones neither (they get the cancelled context; only
actions' post steps get a fresh one), so a cleanup at the job's end
can't be relied on. The check job's second step, after the sandbox's
proof, stops every process of a job's working folder (`$HOME/work/`,
also one removed since) that is neither the runner nor the step itself,
TERM then KILL after 5 s, and names each in the log. The runner takes
one job at a time (capacity: 1), so nothing of a running job is touched.
No change on the server.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
On 23 Oct a push went red in CI because an endpoint's changed comment
hadn't been generated into the module's client: only CI compared the
committed generated code with the generator's output. tool/check.dart now
runs every server's generation (Melos' `generate`, about 15 s) after
fetching the packages and fails, naming the files, where the working
tree's generated code changed by it; the files stay written anew, to be
committed. It compares the files that differ from the committed ones
before and after by their content, so uncommitted changes count as they
do everywhere else in the checks.
It refuses a Serverpod CLI at another version than the lock holds
Serverpod at (`dart install serverpod_cli@<version>`), and the root's
tests hold CI's SERVERPOD_VERSION to the lock. CI keeps installing the
CLI at that version and puts it on the job's PATH; its own comparison
step goes, as the checks run it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
Caddy's network had IPv4 alone, so Docker's userland proxy forwarded
every IPv6 visitor to Caddy from the network's gateway, and the server
counted all of them as one network (the second reading's M5, reasoned).
The edge network gets a unique local /64 (fd4a:6c6b:7e80::/64), Caddy a
fixed address in it beside its IPv4 one, both trusted as the proxy; the
CI runner's unit denies fd00::/8 and its sandbox check tries Caddy's IPv6
address too. A commit of its own: it needs Docker 27 or later with
ip6tables on the server and the network made anew (compose down, never
-v, then up), checked by hand first (NEXT.md).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
The hackathon's judging is over, so learn.halkyon.eu follows main once
more. Pushed together with build 21's release, as released apps up to
build 20 can't sign in against the new server.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
TOTP_KEY, RECOVERY_CODE_KEY and SIGN_IN_DECOY_KEY switch password
sign-in on (each at least 32 characters, or the server stops), and
GITHUB_SIGN_IN_ENDS names the day GitHub stops signing anyone in. CI
gives the test runs throwaway keys as it does the other passwords.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
- Melos 8.9 as the root's dev dependency, with scripts across the
workspace (analyze, test:dart, test:flutter, generate, check) and its
filters (--diff, --include-dependents); cli_util overridden to 0.5.2
for it, since Learn's flutter_launcher_icons asks for 0.4 and uses only
its logger, unchanged in 0.5.
- tool/check.dart in place of tool/check.sh: packages, formatting of our
own files, analysis, the tools' tests, then every package's tests
through Melos (--changed-since=<commit> for only what changed and what
depends on it). tool/pre-push is the hook: the signature check, then
the checks. CI generates every server's code and runs the same check.
- The images build from the root with a workspace narrowed to what they
need (tool/workspace_only.sh); compose, the deploy's migration test,
the snap, the release script, the editor's launch files and Learn's
asset tools (run from apps/learn) follow the new layout.
- The project's source is named as project-halkyon/halkyon (source link,
snap, F-Droid listing, release uploads); the AGPL packages carry the
licence's own text.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
People without GitHub can now keep an anonymous account, and sign in on
another device, with an email address and a code mailed to it each time;
no password. The instance keeps only HMAC-SHA256 of the address, keyed
with its emailSignInPepper (ADR 0010): from what someone types it finds
the person and tells whether an address is free, it mails only an address
someone has just typed, and it can't write to anybody otherwise or show
anybody their address. The device remembers the address it signed in
with, in the person's folder, to show it.
It works as GitHub does. A code for a new address makes it the signed-in
person's, who becomes identified and keeps everything. A code for another
person's address gives a switch token, which signs in as them without a
session; a device holding studysets asks first. Asking for a code answers
alike for every address. Codes are eight Crockford characters, valid 15
minutes and three tries (Serverpod's code challenges); three codes an
hour per address, ten per network. In development the code goes to the
server's log; elsewhere there is no email sign-in until a mail service
is set up (instance.info().emailSignIn).
Someone identified may add the other way in: the Profile tab lists how
they sign in and adds GitHub or an address. They are never switched to
another account: an address or GitHub account that is another's is said
to be so. Switching now deletes the account left behind only when it was
anonymous; it deleted any, which only anonymous people could reach so far.
The migration adds email_identity and email_code_request with their
indexes and foreign keys to Serverpod's challenges; nothing else changes.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
The judges' v0.17.0 uses learn.halkyon.eu, which stays at v0.17.0
meanwhile, while main moves on and is tested locally. The deploy job is
switched off until then; afterwards it runs for main again.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
A release pushes main and then its tag, both on the same commit, and
on: push started a second run for the tag, checking that commit again
while main's run still did: "Checked before?" knows only runs that
finished. A tag deploys nothing, and the release script asks CI about
the commit, not the tag, so tags now start no run.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
The servers' names are no secret, so the settings that carry them stay
as they were, and nothing on a server or a laptop needs changing: CI
runs on `sahaquiel` again, the release uploads to `ramiel-fdroid`, and
deploys by hand go through `sahaquiel-deploy`. Prose still names a
server by the domain it serves, which says what a reader needs.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
Files called the servers by their own host names, which are the
operator's and say nothing a reader needs. They now say which domain a
server serves: learn.halkyon.eu's server (the instance and its CI
runner), the halkyon.app server (Forgejo and the F-Droid repository).
Two names are settings rather than words. CI's runner label becomes
`instance` (`runs-on: instance`, `labels: ["instance:host"]`): the
runner reads its installed copy of deploy/runner/config.yml, so jobs
wait in the queue until that copy is updated and the runner restarted.
The SSH aliases the scripts and docs use become `learn-deploy` and
`fdroid-upload`. AGENTS.md says to keep it so.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
serverpod_offline_sync v0.0.10 collects outbound changes in one
consistent snapshot, so a sync no longer fails on writes made while it
runs (upstream #84, #163), keeps checkpoints per author, so a device
clock moving backwards no longer stalls sync or resends data, and keeps
unique conflicts and foreign key defaults alike on every device. Its
tables and its messages are unchanged: no migration, and apps on v0.0.8
keep syncing. It needs Serverpod >=4.0.4 <4.1.0 (v4.0.4 is fixes only),
so the root overrides that held the workspace on one Serverpod go: each
package pins v4.0.4 itself. The generated code is identical under the
v4.0.4 CLI, which CI now uses.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MMm7usAo6q9ZZRoPkWjoxi
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
A fast-forward merge brings the commit a branch's push checked to main
unchanged, and CI ran every check a second time before deploying it. The
check job now marks every commit that passed in the runner's home
(checked/, kept 90 days), and a commit marked already skips straight to
its deploy; the sandbox check still runs every time. Only pushes run CI,
so only those who may push can leave a mark.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
`flutter pub get --enforce-lockfile` installs what pubspec.lock says or
fails with pub's own message, so CI never resolves something else on
its own.
The second try works around https://github.com/dart-lang/pub/issues/4912
(Dart v3.13): when the runner's cached copy of a git dependency's
repository lacks the locked commit, pub resolves that package without
its dependencies. That failed the CI run of 079179d with Flutter's
"package_graph.json: dependencies for `universal_platform` missing".
With --enforce-lockfile the first run fails at pub instead, having
fetched the commit, and the second goes through; checked locally on a
copy of the pub cache with the editor fork's commit removed. pub's fix
(ebba0dc) is in no Dart release yet.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
Two of Serverpod's own conventions, as its v4 project template and roles have them.
CI now runs `serverpod generate` (CLI pinned to the server's Serverpod version, like
the template's workflow) and fails if the committed generated code differs from its
output, so a model or endpoint changed without regenerating can't reach main. Checked in
a clean clone: today's code passes, a committed stale generated file fails.
redeploy.sh now applies the migrations in a run of their own, with Serverpod's
maintenance role (--role=maintenance --apply-migrations), after the dump and before
switching to the new server. That run exits with 0 only if the migrations worked, and
each migration is one transaction, so a failing one stops the deploy while the old
server keeps serving; before, the new server applied them at start and would have
crash-looped. Tested on the local stack: a deploy through the new step, and a failing
run (wrong database password) that exits with 1 after six attempts. AGENTS.md says
migrations must keep the old server working until the switch.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
The first CI run failed at the server tests: they need the test run mode's passwords,
which passwords.yaml holds on a developer's machine and which is never committed. Without
them Serverpod's test tools exit with code 1 and print nothing. Reproduced in a clean
clone; with the three passwords in the environment (SERVERPOD_DATABASE_PASSWORD,
SERVERPOD_PASSWORD_serverSideSessionKeyHashPepper, SERVERPOD_PASSWORD_serviceSecret) all
checks pass. The job now draws new random ones every time: they only guard the test
database it creates in its own checkout.
Also switch off Flutter's and Dart's usage reports on every job, not only when installing
the SDK: the cancelled first run left an SDK behind before getting that far.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
Installing the runner on sahaquiel, `useradd --system` gave `runner` user ID 999: the ID
under which PostgreSQL and the server run inside their containers, which Docker doesn't
remap. To the kernel, those processes were the runner's own, so a job could have read the
server's environment (database password, session pepper, GitHub secret) and, through
/proc, PostgreSQL's files and its local socket, which the official image trusts. The
sandbox's other walls don't help there: ProtectProc hides only other users' processes,
IPAddressDeny doesn't cover Unix sockets. Only our own check ran as 999; `runner` now has
user and group ID 61000.
The check job now starts with a Sandbox step that fails unless only the runner's own
processes are visible, all in one process namespace, /home/deploy is invisible, the
Docker socket is unusable and Docker's bridge network unreachable. It would have caught
this; outside a sandbox it fails as it should. The unit and AGENTS.md say how to create
the user.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
Pushes to main were deployed by hand. Now Forgejo Actions runs tool/check.sh on every
push, and a green main deploys learn.halkyon.eu by itself.
The runner (forgejo-runner v13.1.0) runs on sahaquiel as the unprivileged user `runner`,
with jobs directly on the host: containers would take Docker access, which is root on the
production server, and Forgejo warns that host jobs have no isolation from the runner's
user. So its systemd unit (deploy/runner/forgejo-runner.service) provides it: /home and
the Docker socket invisible, Docker's container networks unreachable, no capabilities,
two cores and 3 GB at most, lowest priority. It is registered to this repository only,
with the UUID and token of Forgejo's "Create new runner" (the old `register` is
deprecated); the token reaches it as a systemd credential. The configuration
(deploy/runner/config.yml) was dry-run against v13.1.0, the unit rates 1.8 in
`systemd-analyze security`. CI runs on pushes only: a pull request from a fork would run
its code on the production server.
The deploy job calls deploy/redeploy.sh as `deploy`, through a key that authorized_keys
pins to that script and to connections from sahaquiel itself. The script deploys the
commit CI checked while it is still main's tip, as a fast-forward, only if every new
commit is signed by a key in deploy/allowed-signers (as the deployed checkout lists
them), after dumping the database (the last ten are kept), and fails unless the server
answers afterwards; a failure puts the checkout back. Tested against the local stack:
refusals of a malformed argument, a stale commit, a forged unsigned commit and a
non-fast-forward, a real deploy with its dump, a repeat, and a failed build.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>
