[08:27:42]
<ljf> Hi, it seems that postgresql start working on supporting mysql queries... https://news.ycombinator.com/item?id=40094160
[08:28:56]
<ljf> This means we should be able in future to remove mariadb !
[09:12:33]
<Aleks (he/him/il/lui)> o.o
[09:15:01]
<Aleks (he/him/il/lui)> >Alpha will be released in July, Beta in September, and we expect to be GA by EOY.
o.o
[10:35:51]
<mrflos> https://aria.im/_matrix/media/v1/download/matrix.org/cegLLKztWVzVPKvPfSBhCGYE
[10:37:23]
<tituspijean> Packaging? Postgres.
Support? Postgres.
Hunger in the world? You got it, Postgres.
🙃
[10:38:58]
<orhtej2> Sorry in advance if I'm kicking hornets nest but have we considered splitting ynh 12 rollout into 'support for bookworm' and 'other improvements to be shipped when ready'?
[10:39:24]
<Bram> I wish I could use postgres for everything :/
[10:39:43]
<orhtej2> > <@titus:pijean.ovh> Packaging? Postgres.
> Support? Postgres.
> Hunger in the world? You got it, Postgres.
> 🙃
`Select response from http where request=x`
[10:40:18]
<Bram> > Sorry in advance if I'm kicking hornets nest but have we considered splitting ynh 12 rollout into 'support for bookworm' and 'other improvements to be shipped when ready'?
It was already a point of discussion a few months ago yes, it's more that people works on what motivate them and debugging a shit load of applications is just not fun
[10:41:27]
<Aleks (he/him/il/lui)> > Sorry in advance if I'm kicking hornets nest but have we considered splitting ynh 12 rollout into 'support for bookworm' and 'other improvements to be shipped when ready'?
even if it was possible, the blockers are no really the extra improvements
[10:43:12]
<tituspijean> *surprisingly*
[10:44:40]
<Aleks (he/him/il/lui)> having a look at https://github.com/YunoHost/issues/issues/2050#issuecomment-2059827536
- Fixing the major app regressions until we're left with a reasonable number like 10~5%
-> that's one of the major items, still ongoing
- Gotta reimplement DKIM signing with opendkim rather than rspamd (or bring rspamd back by default)
-> Super boring but i'm confident it can be done ... Worst case scenario we can re-add rspamd by default I suppose
- Clarify whatdo with the SSO basic HTTP auth header : with-password vs without-password
-> Kind of boring and not sure what to do, because we'd need to a have a clear idea of which app do need the password in the basic auth header, but we cant really know without switching it off and waiting for people to scream at the other end of the corridor
- Make sure we have a smooth easy way to install / remove metronome (from the global settings maybe, similar to rspamd?)
-> It's just a couple hours of work I think
- Further test/stabilize the bullseye->bookworm migration procedure
-> Second biggest chunck next to fixing the apps
[10:56:21]
<Yunohost Git/Infra notifications> [yunorunner] alexAubin pushed 2 commits to master ([688a1fdc8fd0...3c50ca2d91b4](https://github.com/YunoHost/yunorunner/compare/688a1fdc8fd0...3c50ca2d91b4))
[10:56:27]
<Yunohost Git/Infra notifications> [yunorunner/master] Revert "I have no idea why, but apparently we cant load the config inside main() now ..." This reverts commit 688a1fdc... - Alexandre Aubin
[10:56:38]
<Yunohost Git/Infra notifications> [yunorunner] dependabot[bot] edited [pull request #59](https://github.com/YunoHost/yunorunner/pull/59): Bump aiohttp from 3.9.2 to 3.9.4
[10:56:38]
<Yunohost Git/Infra notifications> [yunorunner] dependabot[bot] edited [pull request #58](https://github.com/YunoHost/yunorunner/pull/58): Bump idna from 2.8 to 3.7
[10:56:59]
<Yunohost Git/Infra notifications> [yunorunner] dependabot[bot] pushed 1 commit to dependabot/pip/idna-3.7: Bump idna from 2.8 to 3.7 Bumps [idna](https://github.com/kjd/idna) from 2.8 to 3.7. - [Release notes](https://github.c... ([5c6a0898](https://github.com/YunoHost/yunorunner/commit/5c6a0898b30512fb52bb0d7c25b9d6cf59f3d627))
[10:57:01]
<Yunohost Git/Infra notifications> [yunorunner] dependabot[bot] edited [pull request #58](https://github.com/YunoHost/yunorunner/pull/58): Bump idna from 2.8 to 3.7
[10:57:03]
<Yunohost Git/Infra notifications> [yunorunner] dependabot[bot] edited [pull request #58](https://github.com/YunoHost/yunorunner/pull/58): Bump idna from 2.8 to 3.7
[10:57:27]
<Yunohost Git/Infra notifications> [yunorunner] dependabot[bot] edited [pull request #58](https://github.com/YunoHost/yunorunner/pull/58): Bump idna from 2.8 to 3.7
[10:57:29]
<Yunohost Git/Infra notifications> [yunorunner] dependabot[bot] edited [pull request #59](https://github.com/YunoHost/yunorunner/pull/59): Bump aiohttp from 3.9.2 to 3.9.4
[10:57:30]
<Yunohost Git/Infra notifications> [yunorunner] dependabot[bot] pushed 1 commit to dependabot/pip/aiohttp-3.9.4: Bump aiohttp from 3.9.2 to 3.9.4 Bumps [aiohttp](https://github.com/aio-libs/aiohttp) from 3.9.2 to 3.9.4. - [Release n... ([e8cbcd62](https://github.com/YunoHost/yunorunner/commit/e8cbcd62794b70959765a36c2a3bc4ab3137ed24))
[10:57:30]
<Yunohost Git/Infra notifications> [yunorunner] dependabot[bot] edited [pull request #59](https://github.com/YunoHost/yunorunner/pull/59): Bump aiohttp from 3.9.2 to 3.9.4
[10:58:04]
<Yunohost Git/Infra notifications> [yunorunner] dependabot[bot] edited [pull request #59](https://github.com/YunoHost/yunorunner/pull/59): Bump aiohttp from 3.9.2 to 3.9.4
[11:00:30]
<Aleks (he/him/il/lui)> bruh
[13:15:30]
<kayou> about dkim, we're already using opendkim-tool to generate it: https://github.com/YunoHost/yunohost/blob/1f52c46c9b90054d9ae38eff0b721a0e397ca05d/hooks/conf_regen/31-rspamd#L37-L40
[13:18:18]
<kayou> (or i don't understand anything about this dkim stuff)
[13:22:08]
<Aleks (he/him/il/lui)> what i meant was validating dkim for email received, and adding dkim headers to emails sent, which happens via rspamd currently (technical word is "it's a milter") and ideally we should switch to the opendkim daemon for this (because rspamd just consumes a stupid amount of ram)
[13:22:16]
<Aleks (he/him/il/lui)> what i meant was validating dkim for email received, and adding dkim headers to emails sent, which happens via rspamd currently (technical word is "it's a milter") and ideally we should switch to the opendkim daemon for this (because rspamd just consumes a stupid amount of ram and is not really properly configured to act as antispam anyway currently)
[14:12:56]
<kayou> that's funny how I still don't know exactly how all this mail stack works
[14:13:48]
<kayou> https://tomthorp.me/blog/howto-verify-opendkim-authenticating-your-domain
https://wiki.debian.org/opendkim
but I found some good resources to configure it
[16:25:33]
<Guillaume Bouzige> email is deeptech (lol)
[20:17:42]
<Yunohost Git/Infra notifications> [issues] kutsaltotem labeled :space_invader: bug on [issue #2380](https://github.com/YunoHost/issues/issues/2380): Could not start the service dovecot
[20:17:42]
<Yunohost Git/Infra notifications> [issues] kutsaltotem opened [issue #2380](https://github.com/YunoHost/issues/issues/2380): Could not start the service dovecot