18:22:59 <geppetto> #startmeeting fpc
18:22:59 <zodbot> Meeting started Wed Jan 24 18:22:59 2018 UTC.  The chair is geppetto. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:22:59 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
18:22:59 <zodbot> The meeting name has been set to 'fpc'
18:22:59 <geppetto> #meetingname fpc
18:22:59 <zodbot> The meeting name has been set to 'fpc'
18:22:59 <geppetto> #topic Roll Call
18:23:02 <geppetto> #chair limburgher
18:23:02 <zodbot> Current chairs: geppetto limburgher
18:23:06 <geppetto> #chair tibbs
18:23:06 <zodbot> Current chairs: geppetto limburgher tibbs
18:23:10 <geppetto> #chair orionp
18:23:10 <zodbot> Current chairs: geppetto limburgher orionp tibbs
18:23:13 <geppetto> #chair mbooth
18:23:13 <zodbot> Current chairs: geppetto limburgher mbooth orionp tibbs
18:23:33 <geppetto> #topic Schedule
18:23:34 <geppetto> https://lists.fedoraproject.org/archives/list/packaging@lists.fedoraproject.org/message/HQV64UZOCVCNVAX6H74R7K4PVY337DTS/
18:23:44 * limburgher still here
18:23:50 <tibbs> I'm still around.
18:23:55 <mbooth> Damn
18:24:07 <limburgher> mbooth that wasn't nice
18:24:15 <mbooth> Sorry
18:24:15 <limburgher> you'll hurt tibbs' feelings.
18:24:20 <geppetto> #topic #729 How to handle non-informative package reviews? .fpc 729
18:24:29 <mbooth> Accidental up arrow+enter
18:24:43 <limburgher> :)
18:25:10 <tibbs> As I wrote in the ticket, I've struggled with the issue in 729 since we started doing package reviews.
18:25:17 <limburgher> Indeed.
18:25:28 <tibbs> I see no acceptable solution besides having the humans processing the SCM requests kicking things back.
18:25:31 <limburgher> I mean it's a GIGO process.
18:25:41 <tibbs> But right now that's just one human, isn't it?
18:25:42 <geppetto> I think we can just close this? … it's one of the least problematic parts of reviews, as you should be able to get someone else to take it over if the reviewer isn't doing anything.
18:25:49 <limburgher> Not entirely.
18:26:11 <limburgher> But at a certain point that leads to the SCM processor basically re-reviewing each package.
18:26:13 <tibbs> geppetto: The issue is that occasionally someone just checks the review+ box and doesn't bother to add anything useful.
18:26:25 <tibbs> I don't think we can solve this with process.
18:26:38 <orionp> Just close it, not our concern
18:26:55 <tibbs> I do think we could modify the review process document to say what reviewers should say.
18:26:57 <limburgher> Yeah, maybe just have something that says if you find a person doing this chronically yell at FESCO.
18:27:08 <tibbs> But right now we trust them to have done the right thing.
18:27:25 <limburgher> tibbs: which at a certain point we have to do.
18:27:27 <geppetto> Well, it's bad … and we should be concerned … but I'm not sure what we can do "please be better" +666
18:27:41 * geppetto shrugs
18:27:50 <tibbs> And if we find out that they couldn't possibly have checked anything (like if they approved a package that doesn't even build) then someone can start a discussion.
18:28:24 <geppetto> #action Close as we can't really do much unless the reviewer is very obviously doing bad things.
18:28:26 <limburgher> Right, and if someone's repeatedly problematic, maybe look at disallowing them from reviewing, but as a last resort.
18:28:36 <geppetto> #info Maybe start a discussion on fedora-devel-list
18:28:48 <limburgher> Their sponsor could work with them.
18:29:03 <geppetto> #topic #731 Testing guidelines .fpc 731
18:29:09 <geppetto> #topic #731
18:29:16 <geppetto> .fpc 731
18:29:18 <zodbot> geppetto: Issue #731: Testing guidelines - packaging-committee - Pagure - https://pagure.io/packaging-committee/issue/731
18:29:56 <tibbs> Does anyone think we need to say more than we say at https://fedoraproject.org/wiki/Packaging:Guidelines#Test_Suites?
18:30:01 <geppetto> I'm def. -1 on us doing anything without a draft
18:30:22 <limburgher> I say we just drop the Perl section.
18:30:29 <geppetto> As we know, we don't have a magic roundtuit tree
18:30:39 <limburgher> Test suites vary, we don't have to document all of them.
18:30:50 <limburgher> "Test as much as possible" should be sufficient.
18:30:52 <tibbs> Domain-specific guidelines can of course suggest how you might run test suites in ways common to packages in that domain.
18:31:23 <limburgher> Can, sure.
18:31:32 <geppetto> Yeh, I'm happy with the one line we have
18:31:35 <tibbs> I don't think we want to get more complicated than that, though.  Instructions on setting up a virtual X server to test graphical apps is definitely not the kind of thing we'd want.
18:31:51 <orionp> I don't see the need to change anything.  Domain specific help for running test suites seems welcome, but not needed in the main guidelines
18:32:05 <limburgher> orionp: nods
18:32:19 <tibbs> I think the Perl guidelines are pretty reasonable, though; they just expand upon the "whenever reasonable" part with perl-specific info.
18:32:22 <geppetto> #action Everything seems fine as it is
18:32:50 <geppetto> #info If you want to create a draft to propose a change feel free to open a ticket (however we are generally against listing every single testing framework)
18:33:02 <tibbs> Maybe other domain-specific guidelines could add test suite information if that is useful.  I have no idea how to test python things properly, for example.
18:33:04 <geppetto> #topic #733 'users and groups' should link to prealloc. list
18:33:10 <geppetto> .fpc 733
18:33:14 <zodbot> geppetto: Issue #733: 'users and groups' guidelines need a link to the current list of preallocated users - packaging-committee - Pagure - https://pagure.io/packaging-committee/issue/733
18:33:32 <tibbs> I honestly don't know why someone needs that, honestly.
18:34:12 <geppetto> Both of those files seem worthless
18:34:23 <tibbs> Maybe if someone could present some rationalization for why that specific information needs to be available directly from the guidelines.
18:34:25 <limburgher> I agree with tibbs's comment in the ticket.
18:34:47 <tibbs> I mean, who cares what the numbers are?  And if you've been allocated a number, surely you know that.
18:35:11 <tibbs> Maybe I'm missing something.
18:35:28 <geppetto> Proposal: Remove the link entirely, as it doesn't seem useful.
18:35:29 <geppetto> +1
18:35:38 <limburgher> And if the lack of this link makes it that much harder to go for a static allocation and encourages dynamic allocation. . .good.
18:35:41 <limburgher> +1
18:35:52 <mbooth> Link already went, no?
18:35:55 <tibbs> There's no link there now, is there?
18:36:06 <mbooth> Page now says "pass it on to the maintainers of the setup package for implementation."
18:36:06 <tibbs> He is referring to an old version of the wiki page before it got cleaned up.
18:36:14 <mbooth> We don't care where setup is maintained :-)
18:36:36 <geppetto> Ahh, my bad … nevermind me
18:36:59 <tibbs> Maybe just ask for some justification as to why this needs to be in the guidelines?
18:37:07 <geppetto> #action Link will stay removed, as it doesn't seem useful.
18:37:30 <orionp> +1
18:37:34 <geppetto> #info If you have reasons for needing this information, feel free to open another ticket with a draft policy change.
18:37:55 <geppetto> #topic #737 Allocating a soft static uid and gid 389 for dirsrv
18:37:59 <geppetto> .fpc 737
18:38:00 <zodbot> geppetto: Issue #737: Allocating a soft static uid and gid 389 for dirsrv - packaging-committee - Pagure - https://pagure.io/packaging-committee/issue/737
18:39:00 <tibbs> Sorry I didn't get back to answering that last comment.
18:39:02 <geppetto> This seems weird, I think they just confused about what to do
18:39:37 <limburgher> Agreed, I see no need to approve at this point, they just need examples..
18:39:41 <tibbs> Yes, this is one of those things where it wasn't being done right initially and now doing it right migh annoy someone.
18:39:59 <tibbs> I guess I should have linked to the actual guidelines in my last comment there.
18:40:03 <limburgher> I'm strongly in favor of annoying people in these cases.
18:40:27 <tibbs> It just means that for new installs they won't get their vanity UID.
18:40:34 <limburgher> :tiny_violin:
18:40:48 <tibbs> 389 is the port number for unencrypted LDAP.
18:41:37 <tibbs> I will answer them since I was having the discussion.  If there's anything else we actually need to do then I'll readd the meeting tag.
18:42:02 * limburgher nods
18:42:27 <geppetto> #action tibbs Will point to the guidlines with the process to do what they want.
18:42:48 <limburgher> guidlines is literally the best possible typo for this ticket.
18:42:55 <geppetto> #topic #726 Review for SELinux Independent Policy packaging Draft
18:42:58 <geppetto> .fpc 726
18:43:00 <zodbot> geppetto: Issue #726: Review for SELinux Independent Policy packaging Draft - packaging-committee - Pagure - https://pagure.io/packaging-committee/issue/726
18:43:17 <tibbs> Sadly I had something to do here that I didn't do.
18:43:20 <geppetto> So I got pinged about this, they think it's ready to go
18:43:58 <geppetto> tibbs: on 726?
18:44:05 <tibbs> Yeah, on 726.
18:44:37 <tibbs> Last time I checked this, it was in a state that they thought was good but really wasn't a guideline.
18:44:58 <tibbs> I had volunteered to extract the guideline bits.
18:45:17 <geppetto> Ok
18:45:36 <tibbs> It still has instructions on setting up a github repository, so I don't think that's changed.
18:46:18 <tibbs> FPC agreed that stuff didn't belong in there, so either they didn't see our discussion about it or they disagree with us.
18:46:43 <geppetto> #action tibbs Will work with you to try to extract the guideline parts of the document (removing stuff about setting up github etc.)
18:47:06 <geppetto> I think it was one of those thing where they didn't read what we discussed, only the summary I posted in the ticket
18:47:40 <geppetto> And I didn't spell it out there, so everyone was confused.
18:47:48 <geppetto> #topic Open Floor
18:47:49 <tibbs> If I had done what I said I'd do then they could at least have that document to look over.
18:48:01 <tibbs> I had a proposal in a recent ticket.
18:48:12 <tibbs> .fpc 744
18:48:13 <zodbot> tibbs: Issue #744: Make packaging guidelines Fedora only - packaging-committee - Pagure - https://pagure.io/packaging-committee/issue/744
18:48:26 <tibbs> Just clarifies what releases the guidelines apply to.
18:48:45 <tibbs> And also, we really have to get back to 710 at some point.
18:49:07 <geppetto> #topic  #744 Make packaging guidelines Fedora only
18:49:11 <geppetto> .fpc 744
18:49:13 <zodbot> geppetto: Issue #744: Make packaging guidelines Fedora only - packaging-committee - Pagure - https://pagure.io/packaging-committee/issue/744
18:49:33 <geppetto> I'm happy to +1 your comment
18:49:42 <limburgher> Ditto.
18:50:05 <tibbs> +1 obviously.
18:50:42 <mbooth> +1
18:50:57 <geppetto> orionp: It's go time ;)
18:52:22 <tibbs> Can vote in the ticket, too.
18:57:01 <tibbs> Did I lag out?
18:57:06 <geppetto> no
18:57:08 <limburgher> Dunno. Did I?
18:57:17 <tibbs> OK, just got quiet it in here.
18:57:34 <geppetto> yeh, I ping'd orionp directly … but I guess he's afk
18:57:53 <tibbs> I guess we're about done.  Can ask for additional votes in the ticket.
18:58:22 <limburgher> Works for me.
18:58:22 <geppetto> #info +4 on the proposal, geppetto; limburgher; tibbs; mbooth
18:58:32 <geppetto> #endmeeting