Owner review 3 (docs/requests/2026-11-04-owner-review-3.md), phase 1: R1 the Uploader–server connection, R2 the screenshot baselines.
Why the Uploader could not connect
On a container whose ports are published as other host ports (the owner's Unraid: 8087→8080, 8444→8443), the panel showed the container's ports (8080/8443) as the ones to use, the template had no Public HTTP port, and both clients turned a bare host into https://host:443. On the owner's host, 8443 is another service, which redirects to its sign-in page.
What changes
Public ports are settings. Settings › Network and TLS has Public HTTP port and Public HTTPS port (locked by CHRONOS_PUBLIC_HTTP(S)_PORT), applied at once to discovery and the HTTPS redirect. In a container the listening ports are read-only ("Container HTTP/HTTPS port").
Client address. The panel shows what to type in Chronos and the Uploader, with Copy. Over HTTPS it uses the port the page came through, and a "Use " button fills it in when the saved one differs. Saving a Public HTTPS port other than that asks first.
One address rule for both clients (ChronosAddress): no scheme and no port means https on 8443, https://host keeps 443, a typed port is kept, and IPv6 gets its brackets (a zone index is refused).
"Not a Chronos server". An address that answers as another service (a redirect, an HTML or text page, JSON of another kind, an error without Chronos's header, an oversized answer) gives No Chronos server answers at …; check the port. Its certificate is never offered for trust, and no credentials are sent. The HTTP port typed without a scheme gives the same message instead of "did not answer".
Template and docs. The Unraid template gains Public HTTP port; compose, guide, README and BRIEF §12.7, §14.8, §14.9, §19.6 now explain Host vs Bridge networking (discovery leaves a container only on Host).
Screenshot baselines. 156 screens captured on the development PC are committed to tests/visual/baselines, and CI no longer pushes the ci-screenshots branch. While --capture freezes the clock, free space, latency and the last sync are shown as fixed sample values. Without that, the captures differed by up to 0.15 % between runs.
Strings for the 22 new or changed Admin keys are in all 38 languages.
Checks
Local: build with 0 warnings. Tests: Server 289, Core 1368, Data 832 (PostgreSQL in CI), Tools 60, App 767 (without the five classes that hang on this PC) all passed. Format, i18n check, xaml-lint and package-guard are clean.
Two captures in a row are pixel-identical, and screenshots diff against the committed baselines passes.
Reviewed by two reviewers (server and clients), fixed, then re-checked. The checker found one regression, now fixed in 2585b5d4: a Chronos error from server/info read its body twice.
Upgrade note for the owner's Unraid
Set Repository to …/chronos-server:latest and Public HTTP port 8087 and Public HTTPS port 8444, or enter them in the panel after the update. Then connect to the Client address the panel shows (https://192.168.30.23:8444).
Owner review 3 (`docs/requests/2026-11-04-owner-review-3.md`), phase 1: R1 the Uploader–server connection, R2 the screenshot baselines.
## Why the Uploader could not connect
On a container whose ports are published as other host ports (the owner's Unraid: 8087→8080, 8444→8443), the panel showed the container's ports (8080/8443) as the ones to use, the template had no Public HTTP port, and both clients turned a bare host into `https://host:443`. On the owner's host, 8443 is another service, which redirects to its sign-in page.
## What changes
- **Public ports are settings.** Settings › Network and TLS has Public HTTP port and Public HTTPS port (locked by `CHRONOS_PUBLIC_HTTP(S)_PORT`), applied at once to discovery and the HTTPS redirect. In a container the listening ports are read-only ("Container HTTP/HTTPS port").
- **Client address.** The panel shows what to type in Chronos and the Uploader, with Copy. Over HTTPS it uses the port the page came through, and a "Use <port>" button fills it in when the saved one differs. Saving a Public HTTPS port other than that asks first.
- **One address rule for both clients** (`ChronosAddress`): no scheme and no port means https on 8443, `https://host` keeps 443, a typed port is kept, and IPv6 gets its brackets (a zone index is refused).
- **"Not a Chronos server".** An address that answers as another service (a redirect, an HTML or text page, JSON of another kind, an error without Chronos's header, an oversized answer) gives *No Chronos server answers at …; check the port*. Its certificate is never offered for trust, and no credentials are sent. The HTTP port typed without a scheme gives the same message instead of "did not answer".
- **Template and docs.** The Unraid template gains Public HTTP port; compose, guide, README and BRIEF §12.7, §14.8, §14.9, §19.6 now explain Host vs Bridge networking (discovery leaves a container only on Host).
- **Screenshot baselines.** 156 screens captured on the development PC are committed to `tests/visual/baselines`, and CI no longer pushes the `ci-screenshots` branch. While `--capture` freezes the clock, free space, latency and the last sync are shown as fixed sample values. Without that, the captures differed by up to 0.15 % between runs.
- Strings for the 22 new or changed Admin keys are in all 38 languages.
## Checks
- Local: build with 0 warnings. Tests: Server 289, Core 1368, Data 832 (PostgreSQL in CI), Tools 60, App 767 (without the five classes that hang on this PC) all passed. Format, i18n check, xaml-lint and package-guard are clean.
- Two captures in a row are pixel-identical, and `screenshots diff` against the committed baselines passes.
- Reviewed by two reviewers (server and clients), fixed, then re-checked. The checker found one regression, now fixed in `2585b5d4`: a Chronos error from `server/info` read its body twice.
## Upgrade note for the owner's Unraid
Set *Repository* to `…/chronos-server:latest` and *Public HTTP port* 8087 and *Public HTTPS port* 8444, or enter them in the panel after the update. Then connect to the *Client address* the panel shows (`https://192.168.30.23:8444`).
🤖 Generated with [Claude Code](https://claude.com/claude-code)
The owner's Uploader could not connect: the panel showed the container's
listening ports (8080/8443) as if clients used them, the public ports were
only environment variables (CHRONOS_PUBLIC_HTTP_PORT not even in the Unraid
template), and the clients mapped a bare host to 443, where another service
answered.
- Public HTTP/HTTPS ports are settings (server.json, or locked by
CHRONOS_PUBLIC_*) and apply at once; ServerSettingsDto gains appended
PublicHttpPort, PublicHttpsPort and InContainer members.
- In a container the listening ports are fixed by the template: locked in the
panel, and saved values from earlier versions are ignored with a warning.
- Settings > Network and TLS shows the client address with a Copy button; the
discovery tooltip says a container needs network type Host.
- ChronosAddress (shared by Chronos and the Uploader): no scheme and no port is
https on 8443, https:// keeps 443, a typed port is kept, IPv6 is bracketed.
- GetInfoAsync throws NotChronosServerException for answers that are not a
Chronos server; another service's certificate is never offered for trust
and no credentials are sent to it. Both clients name the address asked.
- Unraid template, compose file, user guide, README, BRIEF, DECISIONS,
CHANGELOG (with an Unraid upgrade note) and ARCHITECTURE updated; en/it
strings added.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Review fixes on top of the public ports change (owner review 3):
- Settings > Network and TLS: opened over HTTPS, the client address uses the port
the page came through (a container's 8443 published as 8444 with Public HTTPS
port still empty showed the wrong 8443), and a "Use <port>" button fills that
port in. Saving a Public HTTPS port other than the page's asks first, since the
HTTPS redirect follows it at once and a typo would lock the panel's link out.
- GetInfoAsync tells Chronos from other services by stricter rules on both the
trusted and the untrusted-certificate path: any redirect, any answer that is not
JSON, more than 1 MB, or an error without the X-Chronos-Api header is not
Chronos. Another service's certificate is no longer offered for trust after a
cross-scheme redirect, an error page or a JSON 401/403/404, and a JSON 401 over
HTTP no longer reads as a wrong password.
- The HTTP port typed without a scheme (TLS to a plain HTTP port) now gives the
not-Chronos error instead of "the server did not answer".
- The Uploader's certificate question names the address with its port.
- A link-local IPv6 address with a zone index is refused instead of losing it.
- In a container, the first save of the settings removes HttpPort/HttpsPort from
server.json, so the start-up warning stops; the warning says so.
- Guide, README, compose and template: with network type Host the mappings do
nothing and CHRONOS_HTTP_PORT / CHRONOS_HTTPS_PORT move the listening ports off
ones another service uses; the template overview is back to two sentences.
- BRIEF, DECISIONS and CHANGELOG updated; en/it strings added.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The identity check reads Chronos's error body to tell it from another
service; GetInfoAsync then read it again and threw ObjectDisposedException.
The identity now carries the parsed error and GetInfoAsync throws it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The owner chose to capture the baselines on the development PC (owner
review 3, R2): 156 screens go to tests/visual/baselines and CI no longer
pushes a ci-screenshots branch. While --capture freezes AppClock, free
space, a server's latency and the last sync are fixed sample values; they
moved the Server and Import screens past the diff tolerance at random.
The 22 new or changed Admin strings are translated into the 36 other
languages, and the request file records the owner's text and answers.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Owner review 3 (
docs/requests/2026-11-04-owner-review-3.md), phase 1: R1 the Uploader–server connection, R2 the screenshot baselines.Why the Uploader could not connect
On a container whose ports are published as other host ports (the owner's Unraid: 8087→8080, 8444→8443), the panel showed the container's ports (8080/8443) as the ones to use, the template had no Public HTTP port, and both clients turned a bare host into
https://host:443. On the owner's host, 8443 is another service, which redirects to its sign-in page.What changes
CHRONOS_PUBLIC_HTTP(S)_PORT), applied at once to discovery and the HTTPS redirect. In a container the listening ports are read-only ("Container HTTP/HTTPS port").ChronosAddress): no scheme and no port means https on 8443,https://hostkeeps 443, a typed port is kept, and IPv6 gets its brackets (a zone index is refused).tests/visual/baselines, and CI no longer pushes theci-screenshotsbranch. While--capturefreezes the clock, free space, latency and the last sync are shown as fixed sample values. Without that, the captures differed by up to 0.15 % between runs.Checks
screenshots diffagainst the committed baselines passes.2585b5d4: a Chronos error fromserver/inforead its body twice.Upgrade note for the owner's Unraid
Set Repository to
…/chronos-server:latestand Public HTTP port 8087 and Public HTTPS port 8444, or enter them in the panel after the update. Then connect to the Client address the panel shows (https://192.168.30.23:8444).🤖 Generated with Claude Code