<@james:fedora.im>
16:01:50
!startmeeting fpc
<@meetbot:fedora.im>
16:01:50
Meeting started at 2024-03-28 16:01:50 UTC
<@meetbot:fedora.im>
16:01:51
The Meeting name is 'fpc'
<@james:fedora.im>
16:01:55
!topic Roll Call
<@tibbs:fedora.im>
16:02:00
Hey.
<@decathorpe:fedora.im>
16:02:04
!hi
<@zodbot:fedora.im>
16:02:05
Fabio Valentini (decathorpe) - he / him / his
<@limb:fedora.im>
16:02:15
!hi
<@zodbot:fedora.im>
16:02:17
Gwyn Ciesla (limb) - she / her / hers
<@james:fedora.im>
16:02:42
Wooo ... all the people
<@conan_kudo:matrix.org>
16:02:59
!hi
<@zodbot:fedora.im>
16:03:01
Neal Gompa (ngompa) - he / him / his
<@james:fedora.im>
16:04:13
!hi
<@carlwgeorge:matrix.org>
16:04:15
!hi
<@zodbot:fedora.im>
16:04:15
James Antill (james)
<@zodbot:fedora.im>
16:04:17
Carl George (carlwgeorge) - he / him / his
<@james:fedora.im>
16:10:15
!topic FPC Issue 1343: https://pagure.io/packaging-committee/issue/1343
<@james:fedora.im>
16:10:38
I think this should be an easy +1 ... anyone have any objections?
<@tibbs:fedora.im>
16:11:09
+1
<@decathorpe:fedora.im>
16:11:32
+1 but who is going to actually make the change?
<@tibbs:fedora.im>
16:11:37
But no corresponding PR.
<@tibbs:fedora.im>
16:11:56
I'm working on GAP at the moment but I can look at this later if I get the time.
<@decathorpe:fedora.im>
16:12:25
I already commented with "please submit a PR with the proposed to change" a week ago and there was no response yet, so *shrug*
<@limb:fedora.im>
16:12:46
+1
<@tibbs:fedora.im>
16:13:28
Do we think they would want to ack any proposed PR?
<@decathorpe:fedora.im>
16:13:51
I don't know
<@decathorpe:fedora.im>
16:14:04
if FESCo has approved changes, then they should be the ones making the text ...
<@carlwgeorge:matrix.org>
16:14:26
Seems like a reasonable assumption, to ensure that it matches what they're looking for
<@tibbs:fedora.im>
16:14:28
I would tend to agree, but everyone is busy.
<@decathorpe:fedora.im>
16:15:12
we're busy too :)
<@decathorpe:fedora.im>
16:15:29
I'd rather not "guess" what parts of the guidelines need to be changed and how
<@james:fedora.im>
16:17:10
Okay, I'll comment that we are basically +5 ... but nobody has time to write the change for them
<@james:fedora.im>
16:18:05
!topic FPC Issue 1345: https://pagure.io/packaging-committee/issue/1345
<@james:fedora.im>
16:19:01
Seems okay, but as tibbs said ... come back in 9 months or so.
<@decathorpe:fedora.im>
16:20:16
this looks more like "here's what we would like to work", not "this is working now"
<@tibbs:fedora.im>
16:20:27
Cwertainly they are welcome to experiment. I don't object to them using the ticket to record that but then they will need to tell us when there is something for us to look at.
<@tibbs:fedora.im>
16:21:03
I'm certainly no expert in MPI and I'm probably the closest one here to actual use of it (since I support computational mathematics here at work).
<@carlwgeorge:matrix.org>
16:21:04
I'm guessing `module load` has nothing to do with modularity, right?
<@tibbs:fedora.im>
16:21:23
No, that's "environment modules".
<@tibbs:fedora.im>
16:22:10
It's sort of like user-level alternatives. Or switching $PATH around.
<@tibbs:fedora.im>
16:23:47
Indeed, though one came first and it wasn't the "modularity" thing we had for a while.
<@james:fedora.im>
16:23:56
!topic FPC PR 1348: https://pagure.io/packaging-committee/pull-request/1348
<@james:fedora.im>
16:24:06
Going to merge this unless someone speaks up quickly
<@tibbs:fedora.im>
16:24:58
I tried to verify that there were no changes in the output besides the four things I escaped.
<@james:fedora.im>
16:25:57
Okay, I'll leave it then and you can hit merge when you check the output?
<@james:fedora.im>
16:26:51
!topic FPC PR 1342: https://pagure.io/packaging-committee/pull-request/1342
<@james:fedora.im>
16:27:04
Same deal ... Going to merge this unless someone speaks up quickly.
<@tibbs:fedora.im>
16:27:17
Sure, I'll merge it later.
<@decathorpe:fedora.im>
16:27:22
yes please, I have a few more small clarifications for Rust guidelines pending.
<@tibbs:fedora.im>
16:27:30
+1 to 1342.
<@tibbs:fedora.im>
16:28:06
I wonder how the different apostrophes got in there.
<@decathorpe:fedora.im>
16:29:02
no idea. maybe copy-paste issue from a different editor I used ...
<@james:fedora.im>
16:29:27
Done
<@james:fedora.im>
16:29:36
!topic FPC PR 1346: https://pagure.io/packaging-committee/pull-request/1346
<@decathorpe:fedora.im>
16:29:55
looks good to me
<@decathorpe:fedora.im>
16:30:35
I always found it weird that some people mix up two different use cases - using macros to know where to *install* stuff, but they shouldn't be used to check where to *find* stuff
<@tibbs:fedora.im>
16:31:02
I think it may be a good idea to actually write that down somewhere.
<@james:fedora.im>
16:32:06
Yeh, the patch seems fine as it fixes things ... but I feel like we should be changing something somewhere to tell people how to not have this problem
<@tibbs:fedora.im>
16:32:36
Yes, this is just part of it.
<@tibbs:fedora.im>
16:32:52
Also wasn't Fedora going to merge sbin and bin at some point?
<@limb:fedora.im>
16:33:12
Didn't we alrady?
<@decathorpe:fedora.im>
16:33:25
no, that's just a pending change proposal
<@limb:fedora.im>
16:33:28
I that was /usr
<@limb:fedora.im>
16:33:37
Oh that was /usr
<@tibbs:fedora.im>
16:33:46
This line (after the patch) kind of shows why it's confusing: /usr/sbin/update-alternatives --set mta %{_sbindir}/sendmail.sendmail
<@tibbs:fedora.im>
16:34:19
So the macro was removed from the command being executed, but not from the argument.
<@tibbs:fedora.im>
16:34:52
And honestly a clean explanation of why is kind of difficult for me to come up with.
<@decathorpe:fedora.im>
16:35:23
`/usr/sbin/update-alternatives` is something from a different package, and you know where that must be installed
<@decathorpe:fedora.im>
16:35:43
`%{_sbindir}/sendmail.sendmail` is a path where the *current* package installed something
<@decathorpe:fedora.im>
16:36:20
even if you defined a custom `%_prefix` in sendmail, the updates-alternatives binary would still be at the same place
<@tibbs:fedora.im>
16:39:40
So the remaining question is where that needs to be said.
<@tibbs:fedora.im>
16:40:14
And +1 to merging the PR.
<@decathorpe:fedora.im>
16:41:21
maybe here? https://docs.fedoraproject.org/en-US/packaging-guidelines/RPMMacros/#macros_installation
<@tibbs:fedora.im>
16:41:53
Seems reasonable.
<@decathorpe:fedora.im>
16:42:09
the text already mentions that these macros are for determining where to *install* stuff, but not that they shouldn't be used for other purposes
<@tibbs:fedora.im>
16:49:05
It used to be very common to use macros for commands within RPM sections but fortunately that has gone away. This is probably just an extension of that.
<@tibbs:fedora.im>
16:49:18
And I think the alternatives guidelines are pretty old.
<@decathorpe:fedora.im>
16:51:27
AFAIK stuff like `%{__rm}` really should only be used when actually implementing your own RPM macros
<@tibbs:fedora.im>
16:52:56
It was a dark time back then.
<@james:fedora.im>
16:56:02
ha
<@james:fedora.im>
16:56:10
On that note ...
<@james:fedora.im>
16:56:15
!topic Open Floor
<@james:fedora.im>
16:56:23
Anything anyone wants to talk about before we end?
<@decathorpe:fedora.im>
16:56:47
nothing that is not controversial and will take longer than three minutes :)
<@tibbs:fedora.im>
16:57:10
I will finish up semantic breaks for the GAP guidelines in a bit. I thought I had done more of these way back then but I guess I didn't.
<@tibbs:fedora.im>
16:58:46
And now that I look, there are a whole bunch of them that still need that treatment. Oh, well, I'll consider it job security.
<@james:fedora.im>
16:59:30
Sisyphus ftw
<@james:fedora.im>
16:59:39
!endmeeting