The configuration cheat sheet isn't the canonical documentation for the
CLI and this particular bit has no relation to `ENABLE_PPROF` setting
anymore.
Resolves forgejo/forgejo#5268
- it is related to installation instructions
- mention SQLite in the guide
- reduce the complexity of the guide (I think we don't need to cover all the details here, there are better guides available)
As of forgejo/forgejo#2869, console logging implements three adaptations
for running forgejo under systemd/journald:
- forgejo learned to emit printk-style log level prefixes that are
picked up by journald and saved as the severity level of this line
(e.g., `<3>some text\n` means that "some text" will be saved as an
error entry);
- forgejo learned a new pseudo-flag "journaldflags" which is supposed to
replace "stdflags" when running under journald to reduce log clutter
(specifically, it removes date/time from log entries since the
time information is automatically saved by journald and displayed by
journalctl, and replaces the textual log level by the above-described
machine-parsable log level prefix);
- finally, forgejo will try to automatically detect whether it is
running under journald by parsing `$JOURNAL_STREAM` environment
variable and use "journalflags" as default flags if that is the case).
Document those.
Signed-off-by: Ivan Shapovalov <intelfx@intelfx.name>
This changes terminal command blocks to always use the multiline formatting,
as the leading octothorpe looks very strange when displaying a command inline.
Two consecutive inline commands at the start was very difficult to read in the
generated html. See:
https://forgejo.org/docs/latest/admin/installation-binary/
`HTTP_PORT` has been highlighted as a code term. This also means we don't need
to worry about another underscore in the same line being added and interpreted
as syntax.
Simplifies and fixes sentence structure in a few places. I've seen arguments
for and against the use of "just" in technical writing, but I think here it's
better without?
Reworded some paragraphs so that the command comes at the end.
This PR
- adds more details regarding the nature of the secret used when registering runners offline using `forgejo-cli actions register`,
- explains how to change the secret for an existing runner ([Forgejo PR #4619](https://codeberg.org/forgejo/forgejo/pulls/4619)),
- explains how to keep the Forgejo-side command idempotent if the labels aren't known ([Forgejo PR #4610](https://codeberg.org/forgejo/forgejo/pulls/4610))
Reviewed-on: https://codeberg.org/forgejo/docs/pulls/769
Reviewed-by: twenty-panda <twenty-panda@noreply.codeberg.org>
Co-authored-by: Emmanuel BENOÎT <tseeker@nocternity.net>
Co-committed-by: Emmanuel BENOÎT <tseeker@nocternity.net>
The USER_UID/USER_GID are not interpretd in the rootless image, use
`user: "1024:100"` in the docker-compose example instead.
Fixes: https://codeberg.org/forgejo/docs/issues/704
- Eliminated confusing topic switch https->proxy->https.
- Moved all HTTPS mentions to end of intro.
- Added paragraph about when you'd want a reverse proxy.
- Use /code for subpath, to encourage purpose-based naming
Makes a minor correction to which is the next version number.
Also includes information about the `7.0` docker tag's versioning based on facts the author only can recall seeing on Matrix but currently cannot actually find direct proof of such messages existence for reference now.
Reviewed-on: https://codeberg.org/forgejo/docs/pulls/645
Reviewed-by: Earl Warren <earl-warren@noreply.codeberg.org>
Co-authored-by: adagio <h759bkyo4@mozmail.com>
Co-committed-by: adagio <h759bkyo4@mozmail.com>
