Wednesday, December 27, 2023
apps@conference.yunohost.org
December
Mon Tue Wed Thu Fri Sat Sun
        1
2
3
4
5
6
7
8
9
10
11 12 13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
             

[00:04:21] <Aleks (he/him/il/lui)> in theory yes, in practice given the state of the code it involves some bit of refactoring
[00:04:59] <Yunohost Git/Infra notifications> App archivist failed all tests in job [#21696](https://ci-apps.yunohost.org/ci/job/21696) :(
[00:11:13] <Émy - OniriCorpe> > <@Alekswag:matrix.org> in theory yes, in practice given the state of the code it involves some bit of refactoring

okay
[00:12:01] <Émy - OniriCorpe> hm, I think my "small" (+319 −119) PR is finished *relieved*
[00:12:31] <Émy - OniriCorpe> now, time to test
[00:22:22] <Aleks (he/him/il/lui)> https://github.com/YunoHost-Apps/adguardhome_ynh/commit/fced44bdb10e3d3d9e34db02c1cf394ddd49cd42 > eeeh i'm thinking the unbound var thingy is rather because you need to use `${foobar:-}` instead of `$foobar` to handle the case where $foobar was not defined ?
[00:22:42] <Aleks (he/him/il/lui)> (in the if statements)
[00:23:01] <Aleks (he/him/il/lui)> #bash
[00:23:39] <Émy - OniriCorpe> "You can also create doc/PRE_UPGRADE.d/{version}.md to have a note displayed before upgrading to a version equal or higher than {version}"
what if I want to display a message but not for eternity? x)
[00:24:15] <Émy - OniriCorpe> > <@Alekswag:matrix.org> https://github.com/YunoHost-Apps/adguardhome_ynh/commit/fced44bdb10e3d3d9e34db02c1cf394ddd49cd42 > eeeh i'm thinking the unbound var thingy is rather because you need to use `${foobar:-}` instead of `$foobar` to handle the case where $foobar was not defined ?

ah yes, is it better that ynh_app_setting_get, tho?
[00:24:17] <Aleks (he/him/il/lui)> eeh yeah this aint for eternity, this is just when going through the specified version ? 😅
[00:24:42] <Émy - OniriCorpe> > <@Alekswag:matrix.org> eeh yeah this aint for eternity, this is just when going through the specified version ? 😅

"or higher than {version}"
[00:25:12] <Aleks (he/him/il/lui)> hmmm i guess ynh_app_setting_get does define the variable as empty string even if no setting exists yet but meh if just feels weird to use "ynh_app_setting_get" in packaging v2 because this is just for legacy stuff é_è
[00:25:39] <Émy - OniriCorpe> > <@Alekswag:matrix.org> hmmm i guess ynh_app_setting_get does define the variable as empty string even if no setting exists yet but meh if just feels weird to use "ynh_app_setting_get" in packaging v2 because this is just for legacy stuff é_è

okay, fine !!
[00:26:47] <Aleks (he/him/il/lui)> zblerg yeah the wording ain't great but what is meant is when "upgrading from a version lower than X to a version greater-or-equal to X"
[00:27:13] <Aleks (he/him/il/lui)> `previous_version < X and new_version >= X`
[00:28:28] <Yunohost Git/Infra notifications> App ethercalc failed all tests in job [#21699](https://ci-apps.yunohost.org/ci/job/21699) :(
[00:29:32] <Émy - OniriCorpe> hum, the thing I'd like to do is warn people who are going to install the version my PR will be on, that the package behavior has changed, and that they need to go to the config panel to correct the settings if necessary
[00:31:02] <Aleks (he/him/il/lui)> yeah that's exactly for that kind of use case
[00:37:49] <Émy - OniriCorpe> just updated my AGH with my PR and I got this at the end
[00:37:57] <Émy - OniriCorpe> Succès ! adguardhome mis à jour
Succès ! Mise à jour terminée
Attention : Uhoh, somehow the lock /var/run/moulinette_yunohost.lock did not exist ...
[00:38:24] <Émy - OniriCorpe> first time I've seen that
[00:43:25] <Aleks (he/him/il/lui)> hmmm
[00:43:39] <Aleks (he/him/il/lui)> yeah i would just ignore it if you're not able to reproduce it 😅
[01:34:27] <Émy - OniriCorpe> omg what a horror I've made 😭 https://github.com/YunoHost-Apps/adguardhome_ynh/pull/154/commits/556b2458ef39fdb2014a53a666da07a97b635908
[01:35:10] <Aleks (he/him/il/lui)> o.o
[01:36:38] <Émy - OniriCorpe> I feel dirty tbh :")
[01:37:16] <Aleks (he/him/il/lui)> so like there's some reason such that the conf can't just be regenerated using a classing `ynh_add_config` call ?
[01:37:59] <Émy - OniriCorpe> yes, AGH modifies itself the config file, so each regeneration would break the user's changes
[01:38:25] <Aleks (he/him/il/lui)> é_è
[01:39:07] <Émy - OniriCorpe> yeah..........
[01:39:19] <Émy - OniriCorpe> I feel dirty but for a good reason
[01:42:38] <Émy - OniriCorpe> the CI: WARNING Failed to provision ports : Port 53 is already used by another process or app. 😓
[01:42:59] <Aleks (he/him/il/lui)> yeah :/
[01:43:40] <Aleks (he/him/il/lui)> i'm a bit confused, is port 53 meant to be used by adguardhome ?
[01:43:43] <Émy - OniriCorpe> but if AGH can't have port 53 it would break it x)
[01:43:49] <Émy - OniriCorpe> yes
[01:44:36] <Émy - OniriCorpe> i just modified that https://github.com/YunoHost-Apps/adguardhome_ynh/pull/154/commits/f47ae330919189c3dfaeaaf68c10d346df0ccd92
[01:44:36] <Aleks (he/him/il/lui)> so like dnsmasq must be stopped and disabled first ?
[01:45:40] <Aleks (he/him/il/lui)> ah now i realize that i didnt click the "submit" button with my comment warning that this would probably happen xD
[01:47:27] <Émy - OniriCorpe> it's used here https://github.com/YunoHost-Apps/adguardhome_ynh/pull/154/files#diff-c3d7d1d931adcec790544ebfc5c14efcadeada4151c3c792cf3338b4a67eaaa4R19
[01:47:31] <Émy - OniriCorpe> line 19
[01:48:58] <Émy - OniriCorpe> it's used to provide a DNS server to the local network and it can be exposed on Internet if the users desires
[01:49:14] <Émy - OniriCorpe> without it, we can't provide DNS and AGH becomes useless :/
[01:49:54] <Aleks (he/him/il/lui)> yeah i understand that part, but does this mean that you want to stop/disable dnsmasq and have adguardhome "take over" the DNS resolution ?
[01:50:03] <Aleks (he/him/il/lui)> (maybe this is obvious idk i'm trying to clarify)
[01:51:13] <Émy - OniriCorpe> it's unclear to me too
[01:51:14] <Émy - OniriCorpe> dnsmask is active on my yhn where AGH is installed
[01:51:45] <Aleks (he/him/il/lui)> one workaround can be to add a PRE_INSTALL.md disclaimer stating that people must stop/disable dnsmasq (and maybe put some temporary dns resolver in /etc/resolv.conf otherwise dns resolution is going to be broken for the install ?) before installing adguardhome
[01:51:48] <Émy - OniriCorpe> there's an issue about that question and I can't replay because I don't know https://github.com/YunoHost-Apps/adguardhome_ynh/issues/113
[01:52:36] <Émy - OniriCorpe> I'm not a networking nerd :")
[01:53:23] <Aleks (he/him/il/lui)> i try not to become one either xD
[01:54:43] <Émy - OniriCorpe> i'm just a tinkerer who had to learn how to program :"3
[01:56:03] <Aleks (he/him/il/lui)> i would do this:

- add a PRE\_INSTALL.md disclaimer warning that one must stop/disable dnsmasq before installing the app + possibly run `echo 'nameserver 1.1.1.1 > /etc/resolv.conf` as a dns resolver replacement, and also that the app is likely to interfere with other "system apps" like vpnclient and hotspot
- add a `preinstall` section in the app's `tests.toml` as explained in https://github.com/YunoHost/package_check#teststoml-syntax containing `systemctl disable dnsmasq --now` and `echo 'nameserver 1.1.1.1' > /etc/resolv.conf`
[01:56:29] <Émy - OniriCorpe> hum yesy but dnsmask is active and running on my YNH AND I'm using AGH
[01:56:32] <Émy - OniriCorpe> so: confused
[01:56:59] <Aleks (he/him/il/lui)> nani.mp4
[01:57:04] <Émy - OniriCorpe> yep yep
[01:57:29] <Aleks (he/him/il/lui)> but which program is using port 53 if you check `sudo netstat -tulpn | grep '53 '` ?
[02:00:17] <Émy - OniriCorpe> depends
[02:00:37] <Émy - OniriCorpe> netstat -tulpn | grep ":53"
tcp 0 0 192.168.0.80:53 0.0.0.0:* LISTEN 413053/AdGuardHome
tcp 0 0 127.0.0.1:53 0.0.0.0:* LISTEN 412297/dnsmasq
[02:01:09] <Émy - OniriCorpe> same on IPv6
[02:11:08] <Émy - OniriCorpe> so if I understand correctly, AGH asks to localhost dnsmasq and reply with the local IP or public IP
[02:11:19] <Émy - OniriCorpe> so AGH need port 53 but for localhost
[02:11:27] <Aleks (he/him/il/lui)> zblerg yeah so AGH listens on the local network while dnsmasq listens on the local loop
[02:11:34] <Émy - OniriCorpe> yeah more shit to deal with
[02:11:56] <Aleks (he/him/il/lui)> idk if that makes sense but AGH could also be the local loop resolver too ... idk if that makes sense, maybe AGH doesnt really handle the DNS resolution but only the DOH/DOT interface with the client ?
[02:12:16] <Émy - OniriCorpe> > <@Alekswag:matrix.org> idk if that makes sense but AGH could also be the local loop resolver too ... idk if that makes sense, maybe AGH doesnt really handle the DNS resolution but only the DOH/DOT interface with the client ?

is that really relevant?
[02:12:17] <Aleks (he/him/il/lui)> anyway the workaround about stopping dnsmasq still hold, but the topology revealed by netstat suggest that it should be fine to restart dnsmasq once AGH is installed
[02:27:33] <Aleks (he/him/il/lui)> or we add something in the core like "dont_check_if_port_is_already_used" x_x
[02:27:37] <Émy - OniriCorpe> else i remove the adguard.fixed = true? that worked fine without it before
[02:27:41] <Émy - OniriCorpe> but there is a non-zero chance that agh will not receive its port 53
[02:27:50] <Aleks (he/him/il/lui)> > <@oniricorpe:im.emelyne.eu> is that really relevant?

i dunno, i don't really know the detail of DoH / DoT, like does the resolver "resolves" the stuff using DoH/DoT like some sort of DNSsec or does it delegates the resolution to another service ... i guess it doesn't really matter "in normal life" because you use an exernal service and don't want your ISP snooping on your DNS requests ... but in a context of selfhosting, if the *actual* DNS resolution still happens using regular DNS protocol, your ISP is still able to snoop on them
[02:27:51] <Émy - OniriCorpe> I don't use DoH / DoT, I use my AGH only locally on my domestic network to block ads, trackers and stuff
[02:27:56] <Aleks (he/him/il/lui)> ah yes
[02:27:57] <Émy - OniriCorpe> people would use it with DoH, ther's 3 issues about it, so I worked on it
[02:27:59] <Aleks (he/him/il/lui)> https://botsin.space/@scream
[02:27:59] <Aleks (he/him/il/lui)> soooo aaaaa
[02:28:23] <Émy - OniriCorpe> > <@Alekswag:matrix.org> soooo aaaaa

what a mood
[02:51:24] <Émy - OniriCorpe> yup :/
[02:51:24] <Aleks (he/him/il/lui)> oooor
[02:51:24] <Aleks (he/him/il/lui)> > <@oniricorpe:im.emelyne.eu> but there is a non-zero chance that agh will not receive its port 53

yeah, that would work too, and then maybe people can set it back to port 53 if it's exposed in the config panel ... still cumbersome but that would work and there's no perfect solution apart from handling the edge case in the core somehow :/
[02:51:29] <Émy - OniriCorpe> > <@Alekswag:matrix.org> yeah, that would work too, and then maybe people can set it back to port 53 if it's exposed in the config panel ... still cumbersome but that would work and there's no perfect solution apart from handling the edge case in the core somehow :/

I'll do that and that's all
[02:51:43] <Aleks (he/him/il/lui)> actually maybe you do not need to declare the port at all in the manifest ?
[02:51:50] <Aleks (he/him/il/lui)> and just hardcode port 53 in the conf file ?
[02:51:54] <Émy - OniriCorpe> yeah maybe
[02:51:56] <Émy - OniriCorpe> I think it's the better """"""solution"""""
[02:52:01] <Émy - OniriCorpe> so https://github.com/YunoHost-Apps/adguardhome_ynh/pull/154#issuecomment-1869889497
[10:57:54] <Yunohost Git/Infra notifications> [apps] @yunohost-bot pushed 1 commit to add-to-wishlist-gramps-web: Add Gramps Web to wishlist ([eabf3d50](https://github.com/YunoHost/apps/commit/eabf3d5077a4e052dfa674dc064d0adcea7fc80b))
[10:57:56] <Yunohost Git/Infra notifications> [apps] @yunohost-bot created new branch add-to-wishlist-gramps-web
[10:57:57] <Yunohost Git/Infra notifications> [apps] @yunohost-bot opened [pull request #1936](https://github.com/YunoHost/apps/pull/1936): Add Gramps Web to wishlist
[11:21:10] <Yunohost Git/Infra notifications> [apps] @dkoukoul opened [pull request #1937](https://github.com/YunoHost/apps/pull/1937): cjdns working status
[11:52:30] <Yunohost Git/Infra notifications> [apps] @yunohost-bot created new branch add-to-wishlist-fastchat
[11:52:31] <Yunohost Git/Infra notifications> [apps] @yunohost-bot opened [pull request #1938](https://github.com/YunoHost/apps/pull/1938): Add FastChat to wishlist
[11:52:32] <Yunohost Git/Infra notifications> [apps] @yunohost-bot pushed 1 commit to add-to-wishlist-fastchat: Add FastChat to wishlist ([70b93e00](https://github.com/YunoHost/apps/commit/70b93e000e0e88294d5959344d0f3ac2fcca3a7b))
[13:51:12] <lapineige> I have some news, with hints on how to solve this : https://github.com/YunoHost-Apps/peertube\_ynh/pull/397#issuecomment-1870324009
Does that ring a bell for someone ? 🤔
[13:57:09] <Aleks (he/him/il/lui)> he's right, could be about the db not initialized ... the sql message souns like it ("plugin" table not existing, something like this)
[13:57:26] <Aleks (he/him/il/lui)> in fact I'm reading the script and i'm suspicious about : https://github.com/lapineige/peertube_ynh/blob/master/scripts/install#L203
[13:58:03] <Aleks (he/him/il/lui)> `line_match="Started"` is probably meaningless / counterproductive .. I'm thining it will match very early because the systemd log contains the word Started maybe as soon as it's started
[13:58:26] <Aleks (he/him/il/lui)> but maybe the first launch of the service is precisely when the DB is initialized (?) and it probably needs a few seconds for this ...
[14:06:41] <orhtej2> > <@Alekswag:matrix.org> but maybe the first launch of the service is precisely when the DB is initialized (?) and it probably needs a few seconds for this ...

There's rest endpoint to check if service is up a bed running
[14:07:11] <Aleks (he/him/il/lui)> sounds good, we should probably use that
[14:12:35] <Yunohost Git/Infra notifications> [apps] @ericgaspar approved [pull request #1937](https://github.com/YunoHost/apps/pull/1937#pullrequestreview-1797226072) cjdns working status
[14:12:41] <Yunohost Git/Infra notifications> [apps] @ericgaspar merged [pull request #1937](https://github.com/YunoHost/apps/pull/1937): cjdns working status
[14:12:41] <Yunohost Git/Infra notifications> [apps] @ericgaspar pushed 2 commits to master ([9dac790e4013...a46ff5bc40a4](https://github.com/YunoHost/apps/compare/9dac790e4013...a46ff5bc40a4))
[14:12:46] <Yunohost Git/Infra notifications> [apps/master] cjdns working status - dkoukoul
[14:13:50] <lapineige> Can someone guide me on how to do it ? Or make the commit for me ? 😅
Then I can run a manual test (it takes half an hour on my RPI 3)
[14:14:55] <lapineige> > <@Alekswag:matrix.org> but maybe the first launch of the service is precisely when the DB is initialized (?) and it probably needs a few seconds for this ...

It's very slow hardware (hence my commit about yarn install timeouts) so that could be it… more than a few seconds
[14:15:39] <Yunohost Git/Infra notifications> [apps] @alexAubin changes_requested [pull request #1938](https://github.com/YunoHost/apps/pull/1938#pullrequestreview-1797228750) Add FastChat to wishlist: Feels weird, its not like you realistically gonna train AI on a YunoHost server ..
[14:16:31] <Aleks (he/him/il/lui)> uuh, maybe your hardware is slow but it also fails on the CI ?
[14:18:14] <Yunohost Git/Infra notifications> [apps] @alexAubin [commented](https://github.com/YunoHost/apps/pull/1936#discussion_r1437069977) on pull request #1936 Add Gramps Web to wishlist: suggestion description = "Collaborative genealogy, based on and interoperable with Gramps."
[14:18:30] <Yunohost Git/Infra notifications> [apps] @alexAubin pushed 1 commit to add-to-wishlist-gramps-web: Update wishlist.toml: simplify description ([748b7117](https://github.com/YunoHost/apps/commit/748b7117058a9ed755cf2265cf66cf1b9d51ddf5))
[14:18:39] <Yunohost Git/Infra notifications> [apps] @alexAubin merged [pull request #1936](https://github.com/YunoHost/apps/pull/1936): Add Gramps Web to wishlist
[14:18:39] <Yunohost Git/Infra notifications> [apps] @alexAubin pushed 3 commits to master ([a46ff5bc40a4...4a21509302b9](https://github.com/YunoHost/apps/compare/a46ff5bc40a4...4a21509302b9))
[14:18:43] <Yunohost Git/Infra notifications> [apps/master] Add Gramps Web to wishlist - yunohost-bot
[14:18:44] <Yunohost Git/Infra notifications> [apps] @alexAubin deleted branch add-to-wishlist-gramps-web
[14:18:48] <Yunohost Git/Infra notifications> [apps/master] Update wishlist.toml: simplify description - Alexandre Aubin
[14:22:02] <lapineige> CI isn't fast enough ? ^^
[14:23:09] <lapineige> But the CI issue seems different. It's a 502 error, but install of the LDAP plugin doesn't raise any issue
[14:23:20] <Aleks (he/him/il/lui)> my point is the issue isnt about slow or fast hardware, the issue seems to be that the code should be waiting for the service to be up and the DB initialized before continuing
[14:23:37] <lapineige> Yes, I meant that on slow hardware, it's more prone to failure for that special issue
[14:24:38] <Aleks (he/him/il/lui)> ah i didnt realize the log with `"relation \"plugin\" does not exist"` is from your personal test and not from the CI @_@
[14:24:58] <lapineige> I'm trying the dirty fix with a `sleep 1m` in the code ^^
https://github.com/YunoHost-Apps/peertube_ynh/pull/399
[14:25:21] <lapineige> I wasn't clear either 🙂
[14:26:45] <Aleks (he/him/il/lui)> I'd rather replace `--line_match="Started"` with something like `--line_match="HTTP server listening on localhost"` (taken from the service log)
[14:28:07] <lapineige> Nice idea 🙂
[14:28:46] <lapineige> I'm currently trying my dirty fix and I'll let you know how it went 🙂 (and I will be able to test the better one in half an hour 😂)
[14:38:24] <lapineige> > I'm trying the dirty fix with a `sleep 1m` in the code ^^
> https://github.com/YunoHost-Apps/peertube_ynh/pull/399

Doesn't solve CI issue
[14:40:32] <Aleks (he/him/il/lui)> yeah i don't know what the CI issue is, it 502 despite that the service is apparently up and running ... but doesnt mean that it doesnt fix the issue for slow hardware ...
[14:42:52] <lapineige> PR created 🙂
https://github.com/YunoHost-Apps/peertube_ynh/pull/400
[15:03:50] <Yunohost Git/Infra notifications> [apps] @dkoukoul [commented](https://github.com/YunoHost/apps/pull/1937#issuecomment-1870386490) on [issue #1937](https://github.com/YunoHost/apps/pull/1937) cjdns working status: Hello @ericgaspar , thanks for your help with this. I just noticed after local testing that the cjdns service now can no...
[15:08:55] <Aleks (he/him/il/lui)> do we really need to add this empty json as configuration : https://github.com/YunoHost-Apps/peertube_ynh/blob/testing/conf/local-production.json which hasnt been upgraded in 4 years x_x
[15:09:48] <Aleks (he/him/il/lui)> hmm or maybe it's used but we could create it with just `echo '{}' > /path/to/conf` maybe
[15:12:30] <lapineige> Neither the real patch. So there is something else 🙂
[15:12:33] <lapineige> Neither the real patch. So there is something else 🤔
[15:14:11] <Aleks (he/him/il/lui)> the real patch ?
[15:14:57] <lapineige> > PR created 🙂
> https://github.com/YunoHost-Apps/peertube_ynh/pull/400

↑ this one. Using the suggested line for service startup
[15:15:27] <Aleks (he/him/il/lui)> yes it's designed to fix the issue on slow hardware, not the 502 error on the CI
[15:20:45] <lapineige> Leave my dreams untouched please 😂
[15:21:56] <lapineige> > I'm trying the dirty fix with a `sleep 1m` in the code ^^
> https://github.com/YunoHost-Apps/peertube_ynh/pull/399

It worked !
[15:23:16] <lapineige> So in theory it the advanced patch (which install on CI, so the syntax is correct) should work too !
I'm unsure I'll test it too, it took 1h to install Peertube…
[15:28:14] <lapineige> And now I have the 502 bad gateway too ! 😂
[15:28:33] <Aleks (he/him/il/lui)> 😬
[15:28:38] <Aleks (he/him/il/lui)> well then maybe we can debug
[15:29:28] <Aleks (he/him/il/lui)> can you check which port it's supposed to use with `grep port /etc/yunohost/apps/peertube/settings.yml` and compare to `netstat -tulpn | grep PORT` (replacing the port number)
[15:30:36] <Aleks (he/him/il/lui)> maybe something to do with the fact that it listens on `localhost:PORT` but the nginx conf is binded to `127.0.0.1:PORT` ? https://github.com/YunoHost-Apps/peertube_ynh/blob/master/conf/nginx.conf#L19
[15:44:07] <lapineige> > <@Alekswag:matrix.org> can you check which port it's supposed to use with `grep port /etc/yunohost/apps/peertube/settings.yml` and compare to `netstat -tulpn | grep PORT` (replacing the port number)

It's there
[15:45:09] <Aleks (he/him/il/lui)> What if you try to 'curl http://127.0.0.1:PORT'
[15:45:16] <lapineige> > <@Alekswag:matrix.org> maybe something to do with the fact that it listens on `localhost:PORT` but the nginx conf is binded to `127.0.0.1:PORT` ? https://github.com/YunoHost-Apps/peertube_ynh/blob/master/conf/nginx.conf#L19

Gotcha !!
[15:45:18] <Aleks (he/him/il/lui)> Versus using localhost ?
[15:45:30] <lapineige> Changed the nginx line to localhost, it worked !
[15:45:59] <Aleks (he/him/il/lui)> :0
[15:46:08] <lapineige> > <@Alekswag:matrix.org> What if you try to 'curl http://127.0.0.1:PORT'

`curl: (7) Failed to connect to 127.0.0.1 port 8095: Connection refused`
[15:46:23] <lapineige> > <@Alekswag:matrix.org> Versus using localhost ?

An HTTP page as the anwser
[15:46:58] <lapineige> > Changed the nginx line to localhost, it worked !

Then should I submit a patch with this change ?
[15:54:42] <Aleks (he/him/il/lui)> but wait what about changing localhost to 127.0.0.1 inside the production.yaml (+ restarting the service) ?
[15:54:59] <Aleks (he/him/il/lui)> (and keeping 127.0.0.1 in the nginx conf)
[15:55:50] <Aleks (he/him/il/lui)> imho 127.0.0.1 has less ambiguity, compared to localhost which is just some sort of alias defined in /etc/hosts and may resolve to ipv6 in some cases
[15:58:18] <lapineige> > <@Alekswag:matrix.org> but wait what about changing localhost to 127.0.0.1 inside the production.yaml (+ restarting the service) ?

502 bad gateway
[15:58:26] <Aleks (he/him/il/lui)> 🤔
[15:58:41] <Aleks (he/him/il/lui)> and you reverted the changes in nginx right ?
[15:58:51] <lapineige> Also the database setting is also localhost
[15:59:17] <lapineige> > <@Alekswag:matrix.org> and you reverted the changes in nginx right ?

Uh wait maybe not, reloading it again worked
[15:59:38] <Aleks (he/him/il/lui)> *sweating intensifies*
[16:00:02] <lapineige> Maybe it was too soon ? 🤔
[16:00:16] <Aleks (he/him/il/lui)> nah just gotta reload the nginx conf after you edit it
[16:00:38] <Aleks (he/him/il/lui)> so tl:dr, imho we should change localhost to 127.0.0.1 in production.yaml and voila
[16:01:05] <lapineige> I did reload it, twice (worked on 2nd time)
[16:01:05] <Aleks (he/him/il/lui)> bit confused about why it works in master tho
[16:01:05] <lapineige> > <@Alekswag:matrix.org> so tl:dr, imho we should change localhost to 127.0.0.1 in production.yaml and voila

I'm doing it in a PR right now 😄
[16:05:31] <lapineige> Can you give me write right on the Peertube app ?
Then I can merge my changes and test the other awaiting PR.
(pinging eric_G for information)
[16:06:07] <Aleks (he/him/il/lui)> damn are you meaning you don't have full permission over yunohost-apps yet
[16:07:11] <lapineige> It's best for the apps safety 🤣
[16:08:00] <Aleks (he/him/il/lui)> is it 😬
[16:10:13] <Aleks (he/him/il/lui)> whoops i stumbled on my mouse and it made a bunch of random click
[16:10:19] <Aleks (he/him/il/lui)> who knows what it did 😬
[16:15:08] <lapineige> you shouldn't have ^^
[16:16:38] <Aleks (he/him/il/lui)> _Lapineige remove its hood and reveals he is in fact Caerbannog_
[16:20:26] <lapineige> > you shouldn't have ^^

And I didn't merge the commits properly (should have squashed them). I told you 😂
[16:33:16] <lapineige> So now the CI is full 😅
See you in 2 days 🤣
[16:51:46] <Aleks (he/him/il/lui)> The current DISCLAIMER.md mentions a special `ovh_fix` branch which exists but has no diff with respect to master x_x ...
[17:09:15] <Aleks (he/him/il/lui)> https://i.imgflip.com/8alj95.jpg /me everytime i open any app script file
[17:26:09] <lapineige> So much activity in the Peertube_ynh repo right now 🤣
[17:52:09] <lapineige> General feedback, seeing the failled upgrade log of Peertube : there is almost zero information during the backup restoration, which takes forever… it's really hard to tell if anything is still going on.
Should that log each step as during install ?
[17:55:25] <Aleks (he/him/il/lui)> can you share the link to the actual log ?
[18:00:34] <lapineige> https://paste.yunohost.org/raw/ofosifimox
The thing is, during the upgrade on CLI, everything after `getcwd() failed: No such file or directory` is not shown.
[18:02:47] <Aleks (he/him/il/lui)> yeah that's related to the fact that it's the app's script launching the backup restauration
[18:02:53] <Aleks (he/him/il/lui)> this is implicitly fixed by packaging v2
[18:02:57] <orhtej2> > https://paste.yunohost.org/raw/ofosifimox
> The thing is, during the upgrade on CLI, everything after `getcwd() failed: No such file or directory` is not shown.

This getcwd is a red herring
[18:02:58] <Aleks (he/him/il/lui)> the app is still using packaging v1
[18:03:13] <Aleks (he/him/il/lui)> also yes, red herring
[18:03:19] <Aleks (he/him/il/lui)> also fixed by packaging v2
[18:09:58] <Yunohost Git/Infra notifications> [apps] @alexAubin pushed 1 commit to master: Update apps.toml: mark garradin as deprecated, replaced by another software ([81a11f90](https://github.com/YunoHost/apps/commit/81a11f905c676e8530c45ac015c9bfd257134ae5))
[18:33:19] <eric_G> crap, I broke my system while trying opencast package 🥶
[18:33:47] <eric_G> I am getting an error while trying to use apt
[18:34:32] <eric_G> ```
$ sudo apt upgrade
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
Calculating upgrade... Done
The following packages will be upgraded:
crystal crystal1.10 yunohost
3 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
Need to get 0 B/33.0 MB of archives.
After this operation, 4,096 B of additional disk space will be used.
Do you want to continue? [Y/n] Y
Reading changelogs... Done
dpkg: unrecoverable fatal error, aborting:
unknown system user 'opencast' in statoverride file; the system user got removed................]
before the override, which is most probably a packaging bug, to recover you
can remove the override manually with dpkg-statoverride
E: Sub-process /usr/bin/dpkg returned an error code (2)
```



[18:34:54] <Aleks (he/him/il/lui)> ah yeah it's kind of random and annoying but we can fix it
[18:35:48] <Aleks (he/him/il/lui)> does `grep opencast /var/lib/dpkg/statoverride` shows something ?
[18:36:08] <eric_G> ```
opencast opencast 755 /etc/opencast
opencast opencast 755 /srv/opencast
opencast opencast 755 /var/lib/opencast
opencast opencast 755 /usr/share/opencast
opencast opencast 755 /var/log/opencast

[18:36:32] <Aleks (he/him/il/lui)> let's try : `sed -i '/opencast/d' /var/lib/dpkg/statoverride`
[18:38:30] <eric_G> ah! upgrade starting
[18:38:46] <Aleks (he/him/il/lui)> https://www.youtube.com/watch?v=IFS-njpx8SI
[18:40:05] <eric_G> what does `sed -i` do?
[18:40:38] <Aleks (he/him/il/lui)> `sed -i` acts directly on the file content, it means "inline", instead of just outputting the modified version to the terminal
[18:40:56] <Aleks (he/him/il/lui)> and `/opencast/d` means "delete any line containing 'opencast'
[18:41:17] <Aleks (he/him/il/lui)> stolen from https://askubuntu.com/questions/335538/unknown-user-in-statoverride-file/872369#872369
[18:41:49] <eric_G> thanks for the help!
[20:55:19] <Yunohost Git/Infra notifications> [nextcloud_ynh] @alexAubin [commented](https://github.com/YunoHost-Apps/nextcloud_ynh/pull/593#issuecomment-1870620867) on [issue #593](https://github.com/YunoHost-Apps/nextcloud_ynh/pull/593) Update install to avoid previewgenerator error: (Randomly passing by but this should be superseded by the packaging v2 PR, because the system_user resource reapplies th...
[20:56:14] <Yunohost Git/Infra notifications> [nextcloud_ynh] @alexAubin pushed 1 commit to enh-faster-permissions-setter: Update scripts/upgrade: fix final_patg/install_dir typo Co-authored-by: Thomas <51749973+Thovi98@users.noreply.github.c... ([0eace994](https://github.com/YunoHost-Apps/nextcloud_ynh/commit/0eace9945be385c0c8db4fcc58edd548f40a5c18))
[20:56:37] <Yunohost Git/Infra notifications> [nextcloud_ynh] @alexAubin [commented](https://github.com/YunoHost-Apps/nextcloud_ynh/pull/637#issuecomment-1870621645) on [issue #637](https://github.com/YunoHost-Apps/nextcloud_ynh/pull/637) [enh] Faster permissions setter on big datadir: testme
[20:56:40] <Yunohost Git/Infra notifications> [nextcloud_ynh] @yunohost-bot [commented](https://github.com/YunoHost-Apps/nextcloud_ynh/pull/637#issuecomment-1870621675) on [issue #637](https://github.com/YunoHost-Apps/nextcloud_ynh/pull/637) [enh] Faster permissions setter on big datadir: :carousel_horse:
[[Test Badge](https://img.shields.io/endpoint?url=https://ci-apps-dev.yunohost.org/ci/api/job/12196/ba...
[21:16:27] <Aleks (he/him/il/lui)> lapineige: if that's okay i'm cancelling https://ci-apps-dev.yunohost.org/ci/job/12188 because ive included your commit in the v2 PR, otherwise this is wasting too much CI time ..
[22:30:24] <Yunohost Git/Infra notifications> [gitlab_ynh] @kay0u pushed 1 commit to pages: WIP gitlab pages ([003ff553](https://github.com/YunoHost-Apps/gitlab_ynh/commit/003ff5534128bc15273eca0b04d8a03eb33cfbbf))
[22:30:24] <Yunohost Git/Infra notifications> [gitlab_ynh] @kay0u created new branch pages
[22:31:02] <Yunohost Git/Infra notifications> [gitlab_ynh] @kay0u opened [pull request #237](https://github.com/YunoHost-Apps/gitlab_ynh/pull/237): WIP: Pages
[23:02:12] <Yunohost Git/Infra notifications> [example_ynh] @OniriCorpe opened [pull request #220](https://github.com/YunoHost/example_ynh/pull/220): add shellcheck config