ChanServ changed the topic of #freedesktop to: https://www.freedesktop.org infrastructure and online services || for questions about freedesktop.org projects, please see each project's contact || for discussions about specifications, please use https://gitlab.freedesktop.org/xdg or xdg@lists.freedesktop.org
eluks has quit [Remote host closed the connection]
eluks has joined #freedesktop
AbleBacon has quit [Remote host closed the connection]
ximion has quit [Remote host closed the connection]
swatish2 has joined #freedesktop
tzimmermann has joined #freedesktop
Kayden has quit [Ping timeout: 480 seconds]
krushia has quit [Quit: Konversation germinated!]
jsa1 has joined #freedesktop
jvaclav has joined #freedesktop
Arsen has joined #freedesktop
<jani>
fwiw I'm pretty pissed off about fdo gitlab advertizing paid gitlab features like: "Upgrade for advanced agile planning Unlock epics, advanced boards, status, weight, iterations, and more to seamlessly tie your strategy to your DevSecOps workflows with GitLab."
<jani>
(I'm hating the gitlab freemium model more and more every day)
olivial has quit [Ping timeout: 480 seconds]
<svuorela>
jani: where you see that ?
<slomo>
that randomly pops up in the bottom right corner sometimes, yes
<jani>
svuorela: open any issue, bottom right
<jani>
yes
<jani>
unless gitlab is paying for ads, that shouldn't be there
<jani>
daniels: ^
<svuorela>
my debian salsa gitlab does not have it. My KDE invent gitlab has it.
<kode54>
I thought fdo was sponsored for enterprise?
<bentiss>
kode54: we could, but we refused
<kode54>
oh
<svuorela>
and I applaud that.
<kode54>
didn't want to be beholden to sponsors likely
<svuorela>
more 'free software needs free tools' I hope
<kode54>
would hope so
<bentiss>
FWIW, not sure if this will be of any help, but I'm seriously considering making a dump of issues and MR available through lore.gitlab.freedesktop.org (with public-inbox), got the deployment to work out
<bentiss>
mupuf: should I consider this as a "please don't do it" or a "go ahead"?
<jani>
:)
<jani>
the issue search is one of my pet peeves with gitlab. can't have a simple boolean query with more than one label afaict.
<mupuf>
bentiss: Not sure how I feel about it :D On the one hand, emails are notoriously terrible so I don't want to go back there, but this is lore we are talking about so this is just another web service... but then, what's wrong setting notifications in projects or using RSS?
<mupuf>
I guess I just don't understand what you are trying to achieve, I guess
* mupuf
uses rss for mesa issues and MRs with a certain tag. Seems more flexible than downloading all the emails and then using local filters.
<bentiss>
mupuf: initially, one of the complains about gitlab was that it was forcing everybody to go through the web UI, and it's full of javascript and whatnots. So I always dreamed of having that public-inbox setup to shut up people who complain about that. And then, I've started using lei for kernel work, and this is just awesome
<mupuf>
happy to hear you like lei. Still need to set it up to give it a try, but at least I love that you would not have to subscribe to tons of MLs just in case you would need to answer
<bentiss>
basically, instead of pulling everything, you set a filter on the query, and when one filter matches, you get the whole thread in a local inbox
<mupuf>
and what do you use to view it? Mutt?
<bentiss>
yeah
<mupuf>
ack
<bentiss>
(neomutt actually)
<mupuf>
so, in this case, the filters would have to be regex based rather than actually using tags, right?
<bentiss>
the other ent advantage is that you can set it up locally and then you can disconnect (or connect remotely when you are not at home)
<mupuf>
so... we go back to processing on plain text
<mupuf>
not sure I understand your comment about disconnection/remote connection
<bentiss>
yeah, on the lkml, you can set filter like "when a patch touches that directory or any file below" -> no matter what mailing list it was sent
<jrayhawk>
That sounds nice. Much nicer than trying to Xapian-index RSS.
<jrayhawk>
It'd also be a decent place to export all the old text from Bugzilla.
<mupuf>
bentiss: oooooooh, so lore is actually parsing patches. Nice!
<bentiss>
for disconnecting, it can be also useful when you don't have a lot of connectivity or a long timeout (hello .au)
<bentiss>
you have a local cache of your work
<mupuf>
ah, I get it. You meant to say this is like a `git fetch`, all the data is pulled locally on demand
<mupuf>
yeah, that's important too
<bentiss>
mupuf: another filter I have on lei is "the bentiss" one: anytime I get mentioned on any patch/thread, it pops up in my lei-mailbox *with the full history* even if I wasn't cc-ed at the beginning
<mupuf>
nice
<bentiss>
mupuf: yeah, git fetch
<bentiss>
and FTR, I don't even look that much my kernel.org email for kernel patches, because I got everything on lei, without having to set up IMAP :)
<mupuf>
bentiss: I assume lore still can't deal with patch series, and patch series revisions though... but this is not applicable to gitlab since the MRs are tree-based anyway.
<bentiss>
mupuf: lore doesn't, but `b4` does (to some extend)
* mupuf
has kernel patches he needs to send but has not because he just can't be bothered to set git send-email up ever again
<bentiss>
mupuf: b4 is also a bliss to send patches
* bentiss
would never go back
<bentiss>
(though git push -f is much easier)
<mupuf>
I will give it another try, soon-ish)
<mupuf>
hehe, for sure! Git FTW.
<bentiss>
so.. never go back to git send-email when working on the kernel on the lkml
* bentiss
likes the gitlab approach
<bentiss>
but TBH, this is more like an experiment (public-inbox) than anything. But this can also allow to keep public backups in case we get to remove gitlab
<mupuf>
yeah
<jani>
I'm experimenting with replacing mailing list subscriptions with a tool that pulls messages from public-inbox/lore to local maildir
<bentiss>
jani: I strongly suggest using lei for that
<jani>
bentiss: need to look into that again
tuxifan_ has quit []
<jani>
bentiss: but the basic principle is the same, avoid the whole smtp->imap->maildir path
<bentiss>
yeah, except lei does the filtering on the server side
<jani>
well, I run notmuch locally, which indexes the whole thing anyway
<jani>
but lei does have the benefit of pulling in stuff from lists you're not "subscribed" to based on the query
<jani>
hmm, installing lei on debian gives me "lei-daemon /run/user/1000/lei/5.seq.sock" running
<jani>
:(
<bentiss>
I guess things have changed since 2021, but my current setup of lei hasn't changed since and is still working
<jani>
I'm just totally weirded out by having such spartan documentation for evidently opinionated reasons, and then having a perl daemon running for local use. idk.
<jani>
does not compute
<bentiss>
the weird part is here (fedora silverblue), I've got the man for lei-daemon, but 'lei-daemon' or 'lei daemon' are not working
swatish2 has quit [Ping timeout: 480 seconds]
<jani>
bummer, I really don't need or want the local queries with lei. I just want to get the messages from public-inbox delivered locally, and I'll take it from there myself
enunes has joined #freedesktop
vkareh has joined #freedesktop
<bentiss>
indeed, it seems lei-daemon is spawn when lei up is called
swatish2 has joined #freedesktop
__0x1eaf has joined #freedesktop
guludo has joined #freedesktop
ximion has joined #freedesktop
mahkoh has quit [Ping timeout: 480 seconds]
mahkoh has joined #freedesktop
mahkoh has quit [Read error: Connection reset by peer]
mahkoh has joined #freedesktop
swatish2 has quit [Ping timeout: 480 seconds]
mahkoh has quit [Ping timeout: 480 seconds]
<eric_engestrom>
bentiss: is annarchy down?
mahkoh has joined #freedesktop
<eric_engestrom>
debug1: Connecting to annarchy.freedesktop.org [131.252.210.176] port 22.
<eric_engestrom>
debug1: connect to address 131.252.210.176 port 22: Connection refused
swatish2 has joined #freedesktop
<eric_engestrom>
hmm nvm, it seems the problem was my wifi; worked around with a vpn, I'm now able to upload the release :)
mahkoh has quit [Ping timeout: 480 seconds]
mahkoh has joined #freedesktop
mahkoh has quit [Read error: Connection reset by peer]
<daniels>
also just do —target ‘.*’ and it’ll run all the pre-merge jobs. that will run strictly more than required (e.g. for a radv-only MR it will run GL/VAAPI tests and all drivers) but it’s incredibly hard to fix
<glehmann>
I guess it's fine for my MR, because it touches nir anyway
<daniels>
I’m not sure why the non-draft ones aren’t labelled (I’m up the side of a Basque mountain right now) but add the ‘relabel’ label to your MR and mr-l-m should find it and dtrt
tzimmermann has quit [Quit: Leaving]
valentine_ has joined #freedesktop
valentine_ has quit []
swatish2 has quit [Ping timeout: 480 seconds]
valentine has quit [Quit: Reconnecting]
valentine has joined #freedesktop
<valentine>
zmike: couldn't repro it locally, just in CI
<zmike>
very frustrating
a-l-e has quit [Quit: Leaving]
swatish2 has joined #freedesktop
enunes has quit [Ping timeout: 480 seconds]
swatish2 has quit [Ping timeout: 480 seconds]
ximion has quit [Remote host closed the connection]
enunes has joined #freedesktop
<bentiss>
glehmann, daniels: sigh... mr-label-maker is not working because of the invalid token mess that happened 1 year ago, when we had to recreate all of them, and they now all expired on Monday because we couldn't create unlimited token at that time
<bentiss>
(not going to fix this tonight FWIW)
<bentiss>
actually, should be quick enough
<bentiss>
fixed, now we just need to add the relabel in batch