Commit b0844109ca “settings: Show UTC
offsets in timezone dropdown” (#33609) baked a fixed UTC offset for
each time zone into the generated timezones.json at installation time.
UTC offsets are not constants: they change with daylight saving
time (and tzdata updates), so the baked values were wrong for much of
the year. The Python formatting was also incorrect for zones with
negative fractional-hour offsets (divmod rounds toward negative
infinity, so America/St_Johns at UTC-03:30 rendered as UTC-4:30).
Revert the generated file to a plain list of names, and compute each
offset in the web app when the picker is rendered, reusing the
UTC±HH:MM formatting that message timestamp tooltips use. Computing
all 446 offsets takes ~50ms the first time and ~1ms once @date-fns/tz
has cached its formatters.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Signed-off-by: Anders Kaseorg <anders@zulip.com>
This automates backend provisioning, fixes RabbitMQ hostname resolution,
grants PostgreSQL passwordless sudo, and explicitly forwards only the
required web port (9991).
Zulip production suite / ${{ matrix.name }} (zulip/ci:bookworm, --test-custom-db, Debian 12 production install with custom db name and user, bookworm) (push) Has been cancelled
Zulip production suite / ${{ matrix.name }} (zulip/ci:jammy, , Ubuntu 22.04 production install and PostgreSQL upgrade with pgroonga, jammy) (push) Has been cancelled
Provisioning fails when the system username is 'zulip' because the
script attempts to run 'GRANT zulip TO "zulip"', which PostgreSQL
rejects as a circular dependency.
This change adds a conditional check to ensure the GRANT is only
attempted if the database username and system username differ.
We are using browserslist-useragent-regexp to build a regex to compare
our user agents against. This regex is generated and stored in
web/generated and is generated on every provision. This will remain
mostly unchanged since .browserlistrc is not a frequently modified file.
Still, we run it on every provision since we don't have a good mechanism
to detect changes in that list. We can look into that in a followup PR.
In terms of chosing the library, the regexp library is the current
recommendation browserslist github. Another library called
browserslist-useragent exists but it is not actively being worked upon
and is slower as well:
https://gist.github.com/dangreen/55c41072d8891efd3a772a4739d6cd9d
In terms of which browsers to flag as unsupported, we don't flag any
browsers that are not part of >=0 % query of browserslist. If they are
part of >= 0 %, but not part of `baseline widely available with
downstream`, we flag the browser as unsupported.
We also allow higher versions while generating the regex, so if our
regex is outdated a little bit, users with newer browsers don't get the
unsupported browser warning.
- Make the previously optional `logo_url` a required field.
- Add a default logo for the EmbeddedBots, none of which have a logo, by
adding a `fallback_logo_path` field that is used only by the
BotIntegrations and EmbeddedBots.
- Add tests for `get_logo_path`.
Fixes#22582.
This commit
- Replace the blank option with an italicized option that's the current
default language, if there is one selected with "default" label.
- Make the "text" option more informative by adding (no highlighting)
to the label.
- Remove the hint for "text".
- Prioritize as left to right, before start typing:
blank/default language, text, quote, spoiler, math, everything else...
fixes: #33682
To zulip/python-zulip-api, to keep them closer to their source code.
- Renamed the generate_zulip_bots_static_files to
generate_bots_integrations_static_files to accomodate the new function.
- Added a new function to
tools/setup/generate_bots_integrations_static_files to copy the
integration docs into static/generated/integrations.
- Updated integrations.py and computed_settings.py to use the new doc
paths.
- Deleted the affected integration docs.
- Updated the dependency URL.
