[00:00:15]
<kayou> well, it's maybe not a bookworm regression
[00:05:03]
<Aleks (he/him/il/lui)> uuuh maybe the borg client tries multiple keys ?
[00:05:34]
<Aleks (he/him/il/lui)> sometimes people get ban because of this, they have a gazillion keys registered and ssh tries each one of them but each one registers as an auth failure
[00:06:05]
<Aleks (he/him/il/lui)> could it be that the borg client has many keys, ie from /root/.ssh ?
[00:06:24]
<kayou> nah, only one key is used
[00:06:55]
<Aleks (he/him/il/lui)> hmkey and `BORG_RSH` specifies the key explicitly
[00:07:12]
<Aleks (he/him/il/lui)> idk, did you try to grep the auth failures from /var/log/auth.log for that ip ?
[00:07:54]
<kayou> Anyway, I think I'll try it tomorrow when my two servers aren't on the same network, and I'll whitelist the IP if I still have this problem.
[00:08:51]
<kayou> Yes, I tried, but I couldn't find anything interesting
[03:58:17]
<Yunohost Git/Infra notifications> [issues] OniriCorpe labeled :mega: opinion needed on [issue #2412](https://github.com/YunoHost/issues/issues/2412): find a new name for antifeatures?
[03:58:17]
<Yunohost Git/Infra notifications> [issues] OniriCorpe opened [issue #2412](https://github.com/YunoHost/issues/issues/2412): find a new name for antifeatures?
[03:58:28]
<Yunohost Git/Infra notifications> [issues] OniriCorpe assigned OniriCorpe on [issue #2412](https://github.com/YunoHost/issues/issues/2412): find a new name for antifeatures?
[09:41:36]
<kayou> some use use-git for the webadmin? I can't make it work on both bullseye and bookworm
[10:12:14]
<Aleks (he/him/il/lui)> kayou: i do use it sometimes but it's been a while
[10:12:16]
<Aleks (he/him/il/lui)> what's the issue ?
[10:21:45]
<Aleks (he/him/il/lui)> hmmmm random shower thought but `ynh_die` could return the error message via `$YNH_STDRETURN` such that we could obtain the error message from the python core and display it instead of the boring "An error happened inside the app script"
[10:31:26]
<kayou> i'm not sure if i must use `./ynh-dev use_git yunohost-admin` or `yunohost-admin-build`
[10:33:38]
<kayou> > VITE v4.5.3 ready in 313 ms
➜ Local: https://localhost:8080/
➜ Network: https://10.49.68.173:8080/
➜ press h to show help
[10:34:05]
<kayou> when i go to https://10.49.68.173:8080/, it infinite load
[10:36:00]
<kayou> `yunohost firewall allow TCP 8080` fix the issue
[10:37:28]
<kayou> https://github.com/YunoHost/ynh-dev/pull/67
[10:38:09]
<kayou> i was not aware of this, i was confused too because I expected to use the domain.tld with the nginx tweak
[10:38:25]
<kayou> alright, I'll open the port
[10:40:01]
<Aleks (he/him/il/lui)> hmyeah it's not super clear to me, everytime i need to use-git the webadmin i always try a few different thing until it tombe en marche, between using the container IP vs. the domain name, with/without :8080, with/without httpS
[13:19:05]
<Yunohost Git/Infra notifications> gcollin forked yunohost to [gcollin/yunohost](https://github.com/gcollin/yunohost)
[13:41:06]
<Aleks (he/him/il/lui)> ljf: do you have some understand of the things happening in _get_scalable_php_fpm thing where we compute the max children stuff etc ? x_x
[13:41:22]
<Aleks (he/him/il/lui)> i'm so tired of this
[13:43:17]
<ljf> a bit, in the past, this helpers was not totally accurate, and could compute very weird things for example with a server with a lot of ram
[13:43:36]
<Aleks (he/him/il/lui)> yeah when set to "usage" = "high" etc...
[13:43:43]
<ljf> exact
[13:44:29]
<ljf> sometimes i tweak manually some nextcloud configs for best perf
[13:45:01]
<Aleks (he/him/il/lui)> can you share the snippet ?
[13:46:10]
<Aleks (he/him/il/lui)> at the end of the day it looks like there's "only" 5 values handled by this, `php_pm`, `php_max_children`, `min_spare_servers`, `max_spare_servers`, `start_servers` ... and in 95% of apps, the last 3 are just set to 0 by default (because usage/footprint=low or whatever)
[13:46:31]
<ljf> for example i have that in one of my file:
```
pm = static
pm.max_children = 15
request_terminate_timeout = 1d
[...]
php_value[opcache.revalidate_freq]=60
```
[13:47:50]
<Aleks (he/him/il/lui)> i feel like trashing everything except the max_children computation, and then we implement a standard autoamtic config panel for apps to manually edit all the values, and voila >_>
[13:48:25]
<ljf> My point of view on this: it should consider other apps to configure this. Because, what we want , is to use server resources efficiently, and share them between all apps that need it.
[13:49:28]
<ljf> i am searching for an article that helped me to understand this mess
[13:52:22]
<ljf> it's a way, but it's not an ergonomic way.
[13:53:01]
<Aleks (he/him/il/lui)> yeah but right now nobody has any clue how to use this and packages are full of useless "fpm_usage=derp" all over the place ~_~
[13:54:07]
<Aleks (he/him/il/lui)> the raw settings are obscure enough without adding an additional layer which doesn't really help understand what's happening x_x
[13:54:22]
<Aleks (he/him/il/lui)> and in the vast majority of cases having the default "usage = low / footprint = low" is just fine
[13:56:34]
<Aleks (he/him/il/lui)> https://github.com/YunoHost/yunohost/blob/dev/helpers/helpers.v1.d/php#L460
china vibes
[13:57:35]
<ljf> > <@Alekswag:matrix.org> and in the vast majority of cases having the default "usage = low / footprint = low" is just fine
it's not my point of view. Nextcloud with 1 or 2Go is pretty slow due to this, even for 1 user :/
[13:58:45]
<Aleks (he/him/il/lui)> in that case the app should set appropriate default values for the parameter but right now nobody has any idea so zzzz
[13:58:49]
<ljf> people let low/low because they don't understand those questions. So the questions are bad, but configure php fpm could improve quality
[14:01:44]
<ljf> We could imagine to detect in logs "server reached pm.max_children setting" in order to advice to increase this value
[14:07:26]
<ljf> Or may be ask how many simultaneous user the app should have, in order to provision the resources and alert if the instance admin under evaluate the ram and cpu.
[14:09:16]
<ljf> sorry i need to go
[14:10:38]
<ljf> I remember an article who explain how determine the ram per processus. I guess we could compute things in a smart way. If not, your solution to expose the low level settings, will be fine.
[14:13:49]
<Salamandar> Anyone against changing the cron of the dashboard from 6 hours to 2 hours ?
[14:17:03]
<Aleks (he/him/il/lui)> the one fetching data ? the big question are we gonna hit the ratelimit of github because it needs to query quite a bunch of things
[14:17:09]
<Aleks (he/him/il/lui)> and didnt use graphql because ugh
[14:20:26]
<Salamandar> ah ah
[14:20:33]
<Salamandar> yeah ok I put 2 hours, let's see
[14:27:17]
<Yunohost Git/Infra notifications> [yunohost] alexAubin pushed 3 commits to dev ([2f933b5bf25c...dd8db1883ad7](https://github.com/YunoHost/yunohost/compare/2f933b5bf25c...dd8db1883ad7))
[14:27:24]
<Yunohost Git/Infra notifications> [yunohost/dev] apps/logs: fix some information not being redacted because of the packaging v2 flows - Alexandre Aubin
[14:27:40]
<Yunohost Git/Infra notifications> [yunohost] github-actions[bot] created new branch actions/black
[14:27:41]
<Yunohost Git/Infra notifications> [yunohost] github-actions[bot] pushed 1 commit to actions/black: :art: Format Python code with Black ([e66126ee](https://github.com/YunoHost/yunohost/commit/e66126eeaea97bab3ee92ca0c7ee563ea89ff060))
[14:27:42]
<Yunohost Git/Infra notifications> [yunohost] github-actions[bot] opened [pull request #1872](https://github.com/YunoHost/yunohost/pull/1872): Format Python code with Black
[14:30:31]
<Yunohost Git/Infra notifications> 🏗️ Starting build for yunohost/11.2.15+202406211430 for bullseye/unstable/all ...
[14:32:23]
<Yunohost Git/Infra notifications> ✔️ Completed build for yunohost/11.2.15+202406211430 for bullseye/unstable/all.
[15:46:49]
<Yunohost Git/Infra notifications> [yunohost] 🔴 Pipeline [#1342319820](https://gitlab.com/YunoHost/yunohost/-/pipelines/1342319820) failed on branch actions/black
[15:51:42]
<Émy - OniriCorpe> > <@yunohostinfra:matrix.org> [issues] OniriCorpe opened [issue #2412](https://github.com/YunoHost/issues/issues/2412): find a new name for antifeatures?
Pssst 👀
[15:54:24]
<Yunohost Git/Infra notifications> [issues] mrflos [commented](https://github.com/YunoHost/issues/issues/2412#issuecomment-2183015033) on [issue #2412](https://github.com/YunoHost/issues/issues/2412) find a new name for antifeatures?: "Potential issues" ?
[15:55:01]
<Yunohost Git/Infra notifications> [yunohost-portal] yunohost-bot opened [pull request #11](https://github.com/YunoHost/yunohost-portal/pull/11): Translations update from Weblate
[15:56:44]
<Aleks (he/him/il/lui)> > <@oniricorpe:im.emelyne.eu> Pssst 👀
yeah i'm as puzzled as you are, some of them are "controversial features" (rather than anti features), but some others are just ugh "bothersome facts" like alpha/deprecated
[16:02:19]
<tituspijean> Proposal to call it "the more you know" key
[16:02:24]
<mrflos> > <@Alekswag:matrix.org> yeah i'm as puzzled as you are, some of them are "controversial features" (rather than anti features), but some others are just ugh "bothersome facts" like alpha/deprecated
i proposed "Potential issues"
[16:05:11]
<Salamandar> meh
[16:05:13]
<Salamandar> I mean antifeatures is a commonly used term at this point
[16:06:41]
<Yunohost Git/Infra notifications> [yunohost] 🔴 Pipeline [#1342319871](https://gitlab.com/YunoHost/yunohost/-/pipelines/1342319871) failed on branch dev, actions/black
[16:13:49]
<Aleks (he/him/il/lui)> i guess yeah, and in UI we can always have "Stuff you may not like" as a display name
[16:24:02]
<Émy - OniriCorpe> > <@Salamandar:matrix.org> I mean antifeatures is a commonly used term at this point
maybe not, since it's translated to "Undesirable features" on some languages
[16:25:54]
<mrflos> > <@oniricorpe:im.emelyne.eu> maybe not, since it's translated to "Undesirable features" on some languages
there is at least 2 definitions of "anti-feature" https://directory.fsf.org/wiki/Free_Software_Directory:Antifeatures ( behavior that benefits the developer, but that the end user of the software would prefer not to be there ) and https://www.wordsense.eu/anti-feature/ (Functionality that the seller of a product (typically technology) will charge users to not include)
[16:27:47]
<mrflos> the two are not getting the full scope of the tags used for yunohost "anti-features" atm
[16:30:10]
<mrflos> but I suppose it's good enough to keep it as it is, some other subjects may be more urgent
[16:31:56]
<Yunohost Git/Infra notifications> [yunohost] alexAubin pushed 1 commit to ynh_app_setting_set_default: helpers: add a new ynh_app_setting_set_default to replace the unecessarily complex if [ {foo:-} ] trick ([1b5074d8](https://github.com/YunoHost/yunohost/commit/1b5074d857ac22ca0dff1b35702305181eff93d6))
[16:31:57]
<Yunohost Git/Infra notifications> [yunohost] alexAubin created new branch ynh_app_setting_set_default
[16:31:58]
<Yunohost Git/Infra notifications> [yunohost] alexAubin opened [pull request #1873](https://github.com/YunoHost/yunohost/pull/1873): helpers: add a new ynh_app_setting_set_default to replace the unecessarily complex if [ -z {foo:-} ] trick
[16:31:59]
<Émy - OniriCorpe> > <@mrfloss:matrix.org> but I suppose it's good enough to keep it as it is, some other subjects may be more urgent
there always will be anything more urgent than such details
be good details are the exact thing that made something good
and generally, I think I'm always taking care of the details in ynh
[16:33:04]
<Yunohost Git/Infra notifications> [yunohost] alexAubin created new branch rework_fpm_settings_madness
[16:33:08]
<Yunohost Git/Infra notifications> [yunohost] alexAubin pushed 1 commit to rework_fpm_settings_madness: helpers2.1: rework the fpm usage/footprint madness ([9944461f](https://github.com/YunoHost/yunohost/commit/9944461fc59798ee5a7480ca00d5e2492345458d))
[16:39:32]
<mrflos> > <@oniricorpe:im.emelyne.eu> there always will be anything more urgent than such details
> be good details are the exact thing that made something good
> and generally, I think I'm always taking care of the details in ynh
sure, thanks for taking care! I just tried to search for a better term, but after some research, my opinion is that there is no easy way to address this, but maybe someone will find a good way to solve this issue.
[16:39:53]
<Yunohost Git/Infra notifications> [yunohost] alexAubin pushed 1 commit to rework_fpm_settings_madness: helpers2.1: rework the fpm usage/footprint madness ([765d5e34](https://github.com/YunoHost/yunohost/commit/765d5e340d4fbcc94fd0c00373e53c57d0f324b1))
[16:42:52]
<Émy - OniriCorpe> > <@Alekswag:matrix.org> i guess yeah, and in UI we can always have "Stuff you may not like" as a display name
i think this is the way
[16:46:28]
<Yunohost Git/Infra notifications> [yunohost] alexAubin opened [pull request #1874](https://github.com/YunoHost/yunohost/pull/1874): helpers2.1: rework the fpm usage/footprint madness
[16:47:00]
<Aleks (he/him/il/lui)> > <@Alekswag:matrix.org> ljf: do you have some understand of the things happening in _get_scalable_php_fpm thing where we compute the max children stuff etc ? x_x
followup: https://github.com/YunoHost/yunohost/pull/1874
[16:51:32]
<Yunohost Git/Infra notifications> [yunohost] alexAubin pushed 1 commit to rework_fpm_settings_madness: helpers2.1: rework the fpm usage/footprint madness ([8f80d7da](https://github.com/YunoHost/yunohost/commit/8f80d7daabe1fbdb57a88181286a3d852ae2046a))
[16:51:32]
<Yunohost Git/Infra notifications> [yunohost] gcollin opened [pull request #1875](https://github.com/YunoHost/yunohost/pull/1875): Mongo support for Bookworm & avoid reinstall
[16:53:32]
<Yunohost Git/Infra notifications> [yunohost] gcollin edited [pull request #1875](https://github.com/YunoHost/yunohost/pull/1875): Mongo support for Bookworm & avoid reinstall
[16:58:56]
<Yunohost Git/Infra notifications> [yunohost] 🔴 Pipeline [#1342538396](https://gitlab.com/YunoHost/yunohost/-/pipelines/1342538396) failed on branch
[17:01:13]
<kayou> https://www.shellcheck.net/wiki/SC2069
[17:01:19]
<kayou> i was not aware of that
[17:01:55]
<kayou> `cmd 2>&1 > /dev/null` != `cmd > /dev/null 2>&1`
[17:38:29]
<Aleks (he/him/il/lui)> ah yeah
[17:38:31]
<Aleks (he/him/il/lui)> classic bash
[17:45:14]
<Salamandar> ah ah
[17:45:17]
<Salamandar> shellcheck <3
[17:48:27]
<kayou> ljf (pls tell me if I bothering you with my questions) I'm trying to improve the borg apps, and I'd like to know when/where in the core [this line is called](https://github.com/YunoHost-Apps/borg_ynh/blob/159bd111c63fcce0a1b8b3d589b517aa15007a21/conf/backup_method#L89)
[17:50:26]
<kayou> to me it's called only during the restore part, which is not handle yet by the core (for borg)
[17:54:44]
<Yunohost Git/Infra notifications> [yunohost] 🔴 Pipeline [#1342520690](https://gitlab.com/YunoHost/yunohost/-/pipelines/1342520690) failed on branch ynh_app_setting_set_default
[18:23:17]
<Yunohost Git/Infra notifications> [yunohost] 🔴 Pipeline [#1342545475](https://gitlab.com/YunoHost/yunohost/-/pipelines/1342545475) failed on branch rework_fpm_settings_madness
[19:19:16]
<Aleks (he/him/il/lui)> > <@kayou:matrix.org> to me it's called only during the restore part, which is not handle yet by the core (for borg)
+1, wouldn't be surprised that it was "for later™" as it's just the symmetric of other lines
[19:20:01]
<Aleks (he/him/il/lui)> thooo, what makes you say that "the restore part is not yet handled" ? 🤔
[19:24:47]
<Yunohost Git/Infra notifications> [yunohost] centralscrutinizer21 [commented](https://github.com/YunoHost/yunohost/pull/1759#issuecomment-2183329281) on [issue #1759](https://github.com/YunoHost/yunohost/pull/1759) Draft: Bullseye->Bookworm migration: ok since I broke my matrix server for the n time I am going to be bad with my server and try this upgrade with several a...
[19:32:46]
<kayou> > <@Alekswag:matrix.org> thooo, what makes you say that "the restore part is not yet handled" ? 🤔
we have to handle the restore part manually, and not with `yunohost backup restore borg::my_superbackup`
[19:33:47]
<Yunohost Git/Infra notifications> [yunohost] alexAubin [commented](https://github.com/YunoHost/yunohost/pull/1875#issuecomment-2183340565) on [issue #1875](https://github.com/YunoHost/yunohost/pull/1875) Mongo support for Bookworm & avoid reinstall: Related notes from @orhtej2 : https://github.com/YunoHost-Apps/wekan_ynh/pull/284#issuecomment-2106337856 but its not ...
[20:02:20]
<Yunohost Git/Infra notifications> [yunohost] 🔴 Pipeline [#1341021958](https://gitlab.com/YunoHost/yunohost/-/pipelines/1341021958) failed on branch dev, actions/black, rework_fpm_settings_madness, ynh_app_setting_set_default
[20:46:33]
<Yunohost Git/Infra notifications> [yunohost] OniriCorpe approved [pull request #1872](https://github.com/YunoHost/yunohost/pull/1872#pullrequestreview-2133398067) Format Python code with Black
[20:47:12]
<Yunohost Git/Infra notifications> [yunohost] OniriCorpe merged [pull request #1872](https://github.com/YunoHost/yunohost/pull/1872): Format Python code with Black
[20:47:12]
<Yunohost Git/Infra notifications> [yunohost] OniriCorpe pushed 1 commit to dev: :art: Format Python code with Black ([0bd69e56](https://github.com/YunoHost/yunohost/commit/0bd69e56228bf4a53436bec940990c4f66aa029f))
[20:47:36]
<Yunohost Git/Infra notifications> [yunohost] OniriCorpe deleted branch actions/black
[21:00:21]
<Yunohost Git/Infra notifications> 🏗️ Starting build for yunohost/11.2.15+202406212100 for bullseye/unstable/all ...
[21:01:48]
<Yunohost Git/Infra notifications> ✔️ Completed build for yunohost/11.2.15+202406212100 for bullseye/unstable/all.
[22:11:56]
<Yunohost Git/Infra notifications> [yunohost] 🔴 Pipeline [#1342751122](https://gitlab.com/YunoHost/yunohost/-/pipelines/1342751122) failed on branch dev
[23:59:37]
<Yunohost Git/Infra notifications> [yunohost] alexAubin pushed 1 commit to rework_fpm_settings_madness: helpers2.1: rework the fpm usage/footprint madness ([3d728d90](https://github.com/YunoHost/yunohost/commit/3d728d90ceba004968947fc3b8bcff71c7609650))