Records: main=4ffd777b, review-fix cycle lesson (fixture values must come from
backend serializers), health no-poll policy, baselines 368/493 + 31 web tests.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Adversarial review fixes for the #97 endpoint:
- nowPlaying now resolves like the XMLTV guide: when the wall-clock item
is pre/mid-roll filler, surface the guide group's program item (title +
full program bounds) instead of the filler; a guide entry with no
program item (e.g. fallback loop) reports null. Previously an ad break
would retitle the channel and reset progress every 30 seconds.
- covering-item lookup now runs one cheap projected query per distinct
playout offset instead of hydrating the full metadata include tree for
a window widened by the largest offset across ALL channels (up to 24h);
metadata is fetched in a second query scoped to the covering guide
groups, time-bounded to stay safe against GuideGroup recycling (mod
10000). Regression caught while testing: covering items are keyed by
(source channel, lookup time) so a source channel no longer inherits
its offset mirror's lookup result.
- GetChannelStatesForApi now carries Now explicitly (controller passes
DateTime.UtcNow), making boundary semantics testable: new tests pin
now == Start (playing) and now == Finish (not playing), plus mid-roll,
filler-only, and mirror-source-row cases.
- ChannelNowPlayingResponseModel title is non-nullable in the schema
(#nullable enable) — GetDisplayTitle always returns a string.
- GetState uses the typed-return controller convention; Playouts.Mapper
reverted to internal (handler lives in the same assembly).
- Spec doc notes the OnDemand drift caveat and filler semantics.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Living handoff document driving the sequential backend-gap work
(#97, #100-#111, merge pass, then UX screens). Each session works
one queue item and rewrites this file for the next.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Adds the Auto-tuning wizard as hero screen 5 (#69) and annotates the
Channels-list and Schedule-editor sections + prompts with the borrowed
enrichments: lineup health + resume badge (#72, #68), weighted
distribution / reroll / pad-to-boundary / seasonal (#70, #71, #77, #73).
Refs #59#69
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Replaces the inherited 4-editor channel-creation chain with a single
library-to-lineup builder: browse/search library with artwork, drag
shows into the lineup, friendly toggles (shuffle, always-on), one image
for logo+bug, technical settings via built-in/custom Channel Templates
with advanced override. Lists prerequisite backend gaps (composite
create endpoint, Channel Templates, library browse API, full image-type
support incl. transparent re-encode, unified logo/bug, always-on vs
on-demand) and a Claude Design prompt for the screen.
Refs #59
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Records the locked design direction for #59: React SPA on the REST API,
API-first (design pulls the API), Komodo-slick look, ChicoryTV dark +
chicory-blue brand. Includes the 4 hero screens with real domain
vocabulary and ready-to-paste Claude Design prompts.
Refs #59
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Should_Not_Crash_Mid_Roll_{Zero,One}_Chapters passed 'new PlayoutItem()'
to AddFiller, whose Start defaults to DateTime.MinValue. The mid-roll Pad
path (PlayoutModeSchedulerBase.cs:609) subtracts 'currentMinute' minutes
from StartOffset; since StartOffset does ToLocalTime(), a non-zero local
offset makes currentMinute non-zero, underflowing DateTimeOffset.MinValue
and throwing ArgumentOutOfRangeException. Under UTC the offset is 0 so it
stays in range (hence CI was green). Unreachable in production — real
playout items never start at MinValue — so this is a test-data artifact,
not a product bug.
Fix: seed the fixtures with a real Start (startState.CurrentTime.UtcDateTime),
matching the convention in the passing AddFiller tests. Verified across
UTC, Europe/Brussels, and Asia/Kolkata (+5:30); full Core.Tests now passes
under any timezone (484/0). Updated docs/contributing.md §8 per its §10
deviation policy.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Two-phase handoff for a parallel session: Phase 1 investigates the
existing API/CQRS/validation patterns and produces a design + increment
split (Channels-first); Phase 2 implements per vertical slice via
workflows/ultracode. refs #2
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Re-investigated ersatztv#1 against the live stack: ErsatzTV emits
request-host-driven URLs (44 cached guide fragments use {RequestBase},
zero baked localhost), and Jellyfin no longer tuners ErsatzTV directly —
it funnels through Dispatcharr. The historical localhost came from a
client fetching with Host: localhost:8409, a path that no longer exists.
Document the live topology + evidence + resolution; reframe the optional
base-URL setting as not-built; note the latent fragility. Remove the now
spent handoff doc.
refs #1
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- docker-build.yml: the image `build` job now `needs: [test, migrations]`, so a
broken migration blocks the image build (not just the test gate). main branch
protection also now requires the "EF migration integrity (SQLite + MySql)" check
on PR merges. (follow-through on #13)
- docs/handoffs/m3u-tvg-logo-fix.md: working brief for the next task (#1, M3U
tvg-logo localhost), incl. the finding that the issue's GenerateChannelLogoUrl
root-cause is a red herring (that localhost is the ffmpeg watermark path, fetched
in-container; investigate the request flow / forwarded headers / cached values
instead). References the #11 golden + #16 E2E regression nets.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
docs/contributing.md — a descriptive guide to the established patterns, derived from
the existing code: architecture/layering, CQRS handler conventions, LanguageExt
functional style, Blazor/MudBlazor UI, EF Core + dual-provider migrations, the FFmpeg
pipeline pattern, naming/formatting/analyzers, testing, and build/CI — each with
concrete file references — plus the deviation policy (match the established style;
diverge only with a concrete, stated reason). Several sections are now CI-enforced
(#12 layering, #15 analyzers, #13 migrations, #11 golden tests). CLAUDE.md points to
it so it's loaded as project guidance.
Survey across the sections done with parallel read-only agents; all file references
verified to exist.
Refs #10
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Extend the image smoke test from "serves HTTP" to a real E2E: after HTTP readiness,
assert the Jellyfin-facing surfaces on the freshly built image — /iptv/channels.m3u
returns 2xx containing #EXTM3U, and /iptv/xmltv.xml returns 2xx containing a <tv root.
xmltv.xml needs channels.xml (the scheduler writes it a few seconds after boot), so each
endpoint is polled with a deadline; container logs are dumped on failure. Catches
routing / base-URL (#1) / migration regressions that leave the app "up" but serving
broken playlists/guides.
Validated the exact checks against the real :latest image on a fresh container (both 200
with the expected shape) before committing.
Refs #16
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
New `migrations` job in docker-build.yml runs on every push/PR and, per provider:
- `dotnet ef migrations has-pending-model-changes` — fails on model drift (an entity
changed without a matching migration).
- `dotnet ef database update` against a fresh empty DB — applies all migrations,
catching broken/un-orderable ones.
SQLite (the prod provider) uses a throwaway file; MySql uses ServerVersion.AutoDetect
(connects at config time), so the job runs a `services: mysql:8.4` container (the
act_runner uses Docker execution on network downloadswarm, so the service is reachable
as `mysql:3306`). Independent gate for now (not a `needs:` of the image build) so the
MySql-service dependency can't block image builds until proven stable.
Validated both providers locally against real DBs: SQLite 787 migrations -> 139 tables;
MySql 8.4 305 migrations -> 137 tables; both model-drift checks clean.
Docs: ci-cd.md "Migration integrity" (dual-provider discipline, EF CLI pattern, the
non-transactional caveat) + CLAUDE.md convention.
Refs #13
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add Roslynator, SonarAnalyzer.CSharp, Meziantou.Analyzer, and AsyncFixer as central
analyzer references (Directory.Build.targets, CPM-versioned, guarded on
ManagePackageVersionsCentrally so the gitignored .mcp tool isn't affected).
.editorconfig defaults dotnet_analyzer_diagnostic.severity to `suggestion` so the packs
surface findings without failing the TreatWarningsAsErrors (TWAE) build; rules are
promoted to warning/error incrementally (promotion = enforcement via the existing TWAE
build, so no separate lint step is needed).
StyleCop.Analyzers is intentionally excluded: its latest stable (1.1.118) crashes
(AD0001) on C# records and overlaps the existing .editorconfig/Roslynator.
Blazor .razor: editorconfig severity overrides don't reach analyzer diagnostics in Razor
@code (source-generator limitation; dotnet format can't fix them either), so the
currently-firing SonarAnalyzer rules are temporarily NoWarn-ed in ErsatzTV.csproj and
burned down rule-by-rule in #25. The same rules run at suggestion on .cs.
Formatting normalization (mixed UTF-8 BOM + whitespace) is deferred to its own PR.
Full-solution Release build green (0 errors, 0 warnings).
Refs #15
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The previously-deferred dockerfile manager is live: it manages docker/Dockerfile
(mcr dotnet bases + internal ersatztv-ffmpeg) and reads the HTTP-only Gitea
registry via a RENOVATE_HOST_RULES host rule (insecureRegistry + read creds) set
in the workflow env. Vestigial arm/ffmpeg-tests Dockerfiles excluded; compose
manager unused (build-only).
refs server-management#484
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Update ci-cd.md: the scheduled vuln scan is green now that #8 cleared the
advisories (red = a new one); note the NU1901-1903 demotion is kept by design
(criticals still block; lower severities surface via the scan + Renovate).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Renovate is now live (PR #18/#19): self-hosted weekly Gitea Actions job that
opens dependency-update + OSV vuln-fix PRs and a Dependency Dashboard, with
patch-level auto-merge scoped to test/dev-only packages. Document it under
docs/ci-cd.md → Dependency management, plus a CLAUDE.md conventions pointer.
refs server-management#484
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- ci-cd.md: new "Dependency management" section (CPM, NuGet audit, scheduled
dependency-scan.yml); Dockerfile notes now list Directory.Packages.props and
why it's required before restore.
- CLAUDE.md: dependency convention (edit central Directory.Packages.props, never
re-add Version=); correct the test framework note (NUnit, not xUnit).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Document upstream's vYY.<release-seq>.<patch> scheme (year, sequential
release-in-year, patch) in docs/ci-cd.md + CLAUDE.md so we follow it going
forward: <release-seq> is NOT the calendar month (v25.2.0 shipped in June,
v26.3.0 in Feb), it's a per-year counter that resets each January. v26.3.1
= our infra-only rebuild of upstream 26.3.0; v26.4.0 reserved for the first
release with app changes. Also notes the [skip ci]-on-a-tagged-commit gotcha.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- Add concurrency group so the single jazz runner can't run the
push-main-then-push-tag release flow in parallel (shared :buildcache
+ smoke container would collide).
- Add pull_request trigger running the test job only (PRs had no gate);
skip the build job on PRs.
- Only push images from main or a v* tag (workflow_dispatch from other
refs now builds without publishing, instead of clobbering :latest/:prod).
- Replace the log-grep smoke check with a real HTTP readiness probe
(docker exec python3 -> http://localhost:8409/), unique container name,
and trap-based cleanup to avoid leaks on cancel.
- dotnet test now runs -c Release --no-build (was rebuilding in Debug).
- Directory.Build.props: WarningsAsErrors=NU1904 so critical NuGet
advisories block in every project, not just ones with
TreatWarningsAsErrors.
- Dockerfile copies global.json + .editorconfig too, so the image build
matches CI's SDK pin and analyzer severities.
- Remove dead .github/dependabot.yml + FUNDING.yml (upstream-pointed).
- Rewrite docs/ci-cd.md to the implemented pipeline.
Refs #4, #3, #8.
Live Host-header test against the running container proves tvg-logo follows
the request Host header — there is no hardcoded localhost in the M3U/XMLTV
path (the localhost in GenerateChannelLogoUrl only feeds the FFmpeg watermark
overlay). Record the confirmed cause, the operational vs. durable fixes, and
the key architectural fact for future work: ErsatzTV has no advertised/base-URL
setting, so every absolute URL derives from the request Host; ConfigElement is
key/value so adding a setting needs no EF migration. Code fix for #1 deferred
until CI/CD test/prod envs (#3, #4) exist.
Refs #1
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Moves ci-cd.md (59 lines) from Claude memory into docs/ alongside
existing architecture docs. Slims MEMORY.md from 42 to 18 lines by
removing sections duplicated in CLAUDE.md (Tech Stack, Key Patterns,
Architecture Docs index).
Total memory load per session: 101 → 18 lines.
Fixes#7
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Document channel architecture, M3U/XMLTV integration with Jellyfin,
and fork maintenance strategy for the archived upstream. Also includes
CLAUDE.md updates for implementer workflow and project boundaries.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
timothyandClaude Fable 5