docs: the repo claims Accept-Language detection that does not exist #177

Nytt ärende

Öppen

öppnade 2026-08-02 16:06:37 +00:00 av supernaut · 0 kommentarer

supernaut kommenterad

2026-08-02 16:06:37 +00:00

Ägare

Found while inventorying the end-to-end suite. The repo's own documentation says English is
"auto-detected via Accept-Language, with a switcher". No Accept-Language handling exists:
nothing in src/middleware.ts, nothing anywhere under src/, no navigator.language on the
client, and nothing in the Caddy configuration in bitborg-infra. Only the switcher exists. The
commit message for "compact the header below the medium breakpoint" confirms it from the other
direction — it accepts a trade-off explicitly "with no Accept-Language detection".

Either the sentence is wrong, or detection was intended and never landed.

Decide which. If the feature is wanted, file it separately — the mobile header work already noted
that landing in the wrong language is a real cost, since the switcher moved to the footer below
the medium breakpoint.
Either way, correct the documentation so it describes what the site does.

If detection is implemented later, note that Playwright can test it end to end via per-context
locale and extraHTTPHeaders — WebDriver could not, which is part of why this went unnoticed.

Related: the same false claim exists in the public architecture decision record and is tracked
separately as gitborg/gitborg-docs#72. That issue covers the decision record; this one covers this
repository's own documentation. Fix them consistently.

Found while inventorying the end-to-end suite. The repo's own documentation says English is "auto-detected via `Accept-Language`, with a switcher". No `Accept-Language` handling exists: nothing in `src/middleware.ts`, nothing anywhere under `src/`, no `navigator.language` on the client, and nothing in the Caddy configuration in bitborg-infra. Only the switcher exists. The commit message for "compact the header below the medium breakpoint" confirms it from the other direction — it accepts a trade-off explicitly "with no Accept-Language detection". Either the sentence is wrong, or detection was intended and never landed. - Decide which. If the feature is wanted, file it separately — the mobile header work already noted that landing in the wrong language is a real cost, since the switcher moved to the footer below the medium breakpoint. - Either way, correct the documentation so it describes what the site does. If detection is implemented later, note that Playwright can test it end to end via per-context `locale` and `extraHTTPHeaders` — WebDriver could not, which is part of why this went unnoticed. Related: the same false claim exists in the public architecture decision record and is tracked separately as `gitborg/gitborg-docs#72`. That issue covers the decision record; this one covers this repository's own documentation. Fix them consistently.

supernaut lade till etiketterna

2026-08-02 16:06:50 +00:00

supernaut lade till detta till projektet Bitborg Web

2026-08-02 16:06:57 +00:00

supernaut refererade till detta ärende

2026-10-03 00:11:24 +00:00

test(security): prove cross-origin form POSTs are rejected #280

supernaut refererade till detta ärende från en incheckning