16:00:28 <geppetto> #startmeeting fpc
16:00:28 <zodbot> Meeting started Thu Sep 15 16:00:28 2022 UTC.
16:00:28 <zodbot> This meeting is logged and archived in a public location.
16:00:28 <zodbot> The chair is geppetto. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions.
16:00:28 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
16:00:28 <zodbot> The meeting name has been set to 'fpc'
16:00:28 <geppetto> #meetingname fpc
16:00:28 <zodbot> The meeting name has been set to 'fpc'
16:00:28 <geppetto> #topic Roll Call
16:00:38 <tibbs> Hey.
16:00:43 <geppetto> #chair tibbs
16:00:43 <zodbot> Current chairs: geppetto tibbs
16:00:55 <decathorpe> .hi
16:00:56 <zodbot> decathorpe: decathorpe 'Fabio Valentini' <decathorpe@gmail.com>
16:00:56 <decathorpe> good evening
16:01:05 <geppetto> #chair decathorpe
16:01:05 <zodbot> Current chairs: decathorpe geppetto tibbs
16:01:14 <GwynCieslasheher> hi!
16:01:18 <geppetto> #chair GwynCieslasheher
16:01:18 <zodbot> Current chairs: GwynCieslasheher decathorpe geppetto tibbs
16:01:47 <carlwgeorge> .hi
16:01:50 <zodbot> carlwgeorge: carlwgeorge 'Carl George' <carl@redhat.com>
16:03:32 <geppetto> #chair carlwgeorge
16:03:32 <zodbot> Current chairs: GwynCieslasheher carlwgeorge decathorpe geppetto tibbs
16:04:41 <TheExorcist[m]> Hello all, what is the agenda of this meeting please?
16:05:43 <geppetto> The only thing that's really moved and is tagged for meeting is: https://pagure.io/packaging-committee/pull-request/1163
16:05:53 <geppetto> But if you have something, now is the time to bring it up
16:07:10 <TheExorcist[m]> I'm not into packaging. I'm mostly into QA and Respins.
16:07:15 <tibbs> I think 1163 is clear, honestly.  I sort of understand why the people who complain would be inconvenienced but this is something that I don't think RPM will change.
16:07:40 <geppetto> yeh, I kind of want to merge it this meeting
16:08:03 <geppetto> So if anyone has objections/changes/etc. then hopefully we can work through it
16:08:15 <tibbs> Objections to the PR itself are mostly stylistic (line breaks, use of direct language like "you").
16:08:27 <geppetto> #topic conditionalization
16:08:28 <geppetto> https://pagure.io/packaging-committee/pull-request/1163
16:08:32 <decathorpe> Yeah looks good to me
16:09:37 <tibbs> Was hoping the submitter would fix up at least the line breaks but I guess we can do this post-merge.
16:09:50 <geppetto> Cool, it might be worth pointing out the autopatch range thing … but I'm also happy to merge this and wait for something to add/link that info. later.
16:10:29 <carlwgeorge> merging and touching up line breaks later sounds fine to me too
16:11:19 <geppetto> Ok, I clicked the merge button
16:12:34 <geppetto> #topic Open Floor
16:13:12 <decathorpe> I have nothing :)
16:13:36 <carlwgeorge> minor heads up, rpm is adding some helper macros to expose lua string stuff, and i plan on backporting at least %sub to epel-rpm-macros
16:13:48 * geppetto nods
16:13:53 <carlwgeorge> https://github.com/rpm-software-management/rpm/pull/2181 and https://github.com/rpm-software-management/rpm/pull/2179 for reference
16:14:00 <geppetto> You see people using it that quickly?
16:14:25 <carlwgeorge> hopefully yes, to stop shelling out to echo for %shortcommit
16:15:35 <tibbs> Yeah logs has changed with the lua stuff.  Only F35 is missing stuff that makes RPM macros available in a lua table.
16:16:09 <decathorpe> carlwgeorge: will that new functionality be backported to older RPM versions?
16:16:25 <tibbs> That probably means it's time to dust off that PR for documenting the build constraing stuff.
16:16:33 <carlwgeorge> panu said maybe for 4.17, but no earlier
16:16:39 <tibbs> You can presume that the fancy new stuff will never make it into old RPM.
16:16:45 <decathorpe> ok, so epel will have to wait?
16:17:25 <carlwgeorge> there are other less clean ways to implement it for epel, see my closed pr https://github.com/rpm-software-management/rpm/pull/2179
16:17:31 <tibbs> I mean, that's up to Red Hat but we've always had to live with EPEL RPM being very, very behind.
16:17:48 <decathorpe> Ah, I see.
16:18:26 <tibbs> So I could technically do the %constrain_build stuff for EPEL (and F35) but.... I don't particularly want to because it just means rewriting it completely.
16:19:31 <carlwgeorge> likely what i backport to epel-rpm-macros will be incomplete and flawed but will work for the most common use cases
16:20:56 <decathorpe> Does that stuff really need to work everywhere and for anybody and their dog before we document it?
16:21:06 <tibbs> I think the "single spec for everything" approach is untenable in the long term anyway, but that's because I remember packaging for RHEL4.
16:21:43 <tibbs> Here, we just need to worry about Fedora.
16:21:54 <carlwgeorge> epel has a mini set of guidelines for documenting things in fedora guidelines that don't apply or don't work
16:22:09 <tibbs> But it's nice to not be deliberately antagonistic towards people doing other things.
16:22:48 <tibbs> We've been pretty lucky since RHEL7 hasn't been that far behind, but the more useful stuff RPM adds, the farther behind EPEL is going to get.
16:23:00 <decathorpe> Just add NOTE: This depends on RPM 4.18+, which is not available on EPEL < X and Fedora < Y ...
16:23:01 <carlwgeorge> from the epel perspective i don't consider this one antagonistic at all
16:23:41 <carlwgeorge> oh i like that idea, transferring the epel special cases to notes in the main guidelines
16:24:06 <tibbs> We actually worked to achieve the opposite.
16:24:07 <carlwgeorge> as long as it's just a short note and not like a whole section (like ldconfig scriptlets)
16:24:14 <tibbs> The guidelines document Fedora.
16:24:53 <tibbs> https://docs.fedoraproject.org/en-US/packaging-guidelines/#_applicability
16:25:10 <geppetto> Yes, but…
16:25:12 <decathorpe> Yeah but we already mention if things don't work on old but still supprted Fedora releases ... and adding a single "and EPEL" to those warnings wouldn't hurt.
16:25:28 <tibbs> Nothing against having a note; if someone could make a convenient macro for it then that would be nice.
16:25:56 <tibbs> But I worry that such things would then either be expected, or would become out of date.
16:27:01 <decathorpe> same problem as with all other guidelines we have ...
16:27:19 * decathorpe shudders when thinking about some obscure stuff like Drupal packaging which has probably been out-of-date for decades
16:28:11 * geppetto nods
16:28:11 <tibbs> Yeah I try not to think about that.
16:28:24 <geppetto> carlwgeorge: Can you create an epel warning macro?
16:28:42 <tibbs> Honestly I think we may consider just nuking some of the worst stuff.  There is no need to have guidelines for everything, after all.
16:29:21 <carlwgeorge> i'm assuming this is an asciidoc macro you mean?  i'm not really familiar but i can try.
16:29:25 <decathorpe> tibbs: Yeah, I've been thinking about how to make the Guidelines more current ... and that would involve dropping some of the really old stuff (unless somebody objects and volunteers to update it, of course)
16:29:32 <geppetto> Might be worth at least a warning about "it's been X years since someone looked at this"
16:30:05 <geppetto> carlwgeorge: Yeh, I assume you can copy/paste the older versions of fedora ones and change the wording.
16:30:15 <tibbs> I think that such a warning would just be an admission that the page really should just be deleted.
16:30:33 <tibbs> It's not as if we can't get things back now anyway if someone decides to work on them.
16:31:00 <geppetto> I mean, I guess the data will still be in git … so it's not a real delete … just thinking that if someone came along it might still be a useful starting point.
16:31:24 <tibbs> And while this stuff might not be actively harmful to have just sitting, if it has no utility at all then I guess it sort of harms the image of the project.
16:33:29 <decathorpe> It's also kind of surprising that people get directed at some of those ancient pages by google searches etc.
16:35:26 <tibbs> I remember when the Internet was new and all of the information there was current.
16:35:30 <tibbs> But that was 1992.
16:36:05 <tibbs> Now you search for help on something and get info on a 15 year old version.
16:38:07 <geppetto> Could be 30 yr old info at this point ;)
16:38:24 <tibbs> So I guess the next step is to make a list or maybe tag documents that should probably go away, then talk to someone who actually knows the subject matter.
16:38:39 * geppetto nods
16:38:45 <tibbs> Most of the deadest documents in the guidelines are going to be in areas where none of us have specific knowledge.
16:38:46 <geppetto> Can probably sort by edit date
16:39:14 <decathorpe> Should I Write an email to the devel@ and packaging@ lists with some of the ancient pages?
16:39:23 <geppetto> Sure
16:39:29 <decathorpe> And ask of somebody would miss them or wants to update them?
16:40:51 <decathorpe> For now I think it's safe to assume the ones where we already added "last-reviewed" to AsciiDoc metadata are at least not "ancient"
16:41:30 * geppetto nods … fair enough
16:41:51 <geppetto> On that note, I'll close in a minute and give you all 15+ minutes back.
16:42:09 <decathorpe> good idea :) thanks for running the meeting, see you next week
16:42:15 <tibbs> Thanks.
16:42:30 <geppetto> #endmeeting