16:00:04 #startmeeting fpc 16:00:04 Meeting started Thu May 19 16:00:04 2016 UTC. The chair is geppetto. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:04 Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:00:04 The meeting name has been set to 'fpc' 16:00:04 #meetingname fpc 16:00:04 The meeting name has been set to 'fpc' 16:00:04 #topic Roll Call 16:00:55 Hi 16:00:59 #chair mbooth 16:00:59 Current chairs: geppetto mbooth 16:01:02 Hey 16:01:07 * racor is here 16:01:28 #chair racor 16:01:28 Current chairs: geppetto mbooth racor 16:02:14 #chair tomspur 16:02:14 Current chairs: geppetto mbooth racor tomspur 16:02:25 Hi 16:02:31 Hey 16:03:04 hi 16:03:11 #chair Rathann 16:03:11 Current chairs: Rathann geppetto mbooth racor tomspur 16:03:39 wait a few more minutes to see if orion or tibbs can make it 16:03:49 hello 16:04:13 #chair orionp 16:04:13 Current chairs: Rathann geppetto mbooth orionp racor tomspur 16:05:12 #topic Schedule 16:05:15 #link https://lists.fedoraproject.org/archives/list/packaging@lists.fedoraproject.org/message/I2J2MQ4NK2RXFHDEMYHJOOCOWLRELZFJ/ 16:05:23 Ok 16:05:30 #topic #625 GAP Packaging Guidelines Draft 16:05:33 .fpc 625 16:05:35 geppetto: #625 (GAP Packaging Guidelines Draft) – fpc - https://fedorahosted.org/fpc/ticket/625 16:07:22 It could do with an example 16:08:23 Hey, folks. Sorry about that. 16:08:32 #chair tibbs 16:08:32 Current chairs: Rathann geppetto mbooth orionp racor tibbs tomspur 16:08:35 No problem 16:08:47 Wow, seven. Not bad. 16:08:50 yeah, sample spec would be good 16:08:56 TTH could be explained brieflx in the text. But maybe that's obvious if you know GAP packages 16:09:31 it'd be nice to actually map the 2-level GAP dep system to RPMs hard/soft/weak deps instead of leaving it to the discretion of the packager 16:09:42 that's what guidelines are for, after all 16:09:49 otherwise, nice work 16:09:55 maybe "should" instead of "must" but still 16:10:11 yeh 16:10:32 also, a dep generator that would parse this PackageInfo.g and spit them out automatically would be extra nice 16:10:39 tomspur: Yeh, I think that bit mostly says "if you need tth to build, add tth as a buildrequires" ... but maybe it's saying something else too 16:10:58 Rathann: What colour would you like your pony ;) 16:11:08 pink and blue 16:11:36 :) 16:11:45 Rathann: Purple. Got it. 16:11:54 well, anyway, the draft looks sane enough for +1 16:12:10 it's quite good for the first draft 16:12:17 I agree, it's relatively sane for a first go at it. 16:12:35 oh, just one thing 16:12:50 I don't think it's correct to exclude license files from the package 16:12:55 Sadly, someone didn't read the guidelines about naming things in lower case with they packaged GAPDoc. 16:13:01 they should go into %license line 16:13:23 Rathann: If you have the same license as something you depend on ... I think it's fine. 16:13:27 Rathann: I think the intent of that was that they wouldn't be installed down in the gap_dir/pkg directory. 16:13:32 Rathann: If you are an addon/plugin/etc. 16:13:44 But the wording suggests otherwise. 16:13:47 indeed 16:14:04 ah well 16:14:09 Honestly, a sample spec would kind of be nice. If for no reason other than letting me look at it. 16:14:21 * geppetto nods 16:14:30 two sentences earlier it says it's talking about files going into %{_gap_dir}/pkg 16:14:38 so maybe I'm exaggerating 16:15:18 Right, it's just confusing. 16:15:27 Let me make an edit. 16:16:44 https://fedoraproject.org/w/index.php?title=User%3AJjames%2FGAP_packaging_guidelines&diff=447141&oldid=446656 16:17:02 +1 16:18:17 +1 16:18:56 do we want to approve it as is though, or ask for the examples and cleanup the tth bit and the weak deps? 16:19:07 #1 16:19:22 +1 16:19:22 I don't think I mind +1'ing as is, and asking for improvements :) 16:20:10 yeah, tibbs change is better, but I still want the other changes 16:20:20 I'm happy to +1 now and work out kinks as we find them. 16:20:34 But I don't think it hurts to ask for a sample. 16:21:13 Also, maybe a macro to clean up some cruft from the installed package directory so people aren't all making up their own long find commands. 16:22:46 I wonder if the %docdir stuff actually is a packaging/configuration bug in some gap-package somewhere. 16:25:23 Down at the bottom? I don't think so. 16:26:16 This kind of thing isn't really uncommon with software suites which have their own documentation system. 16:26:31 tibbs|w: yes, the "documentation" section 16:26:50 Though maybe it would be possible to just make that internal documentation system look in the same place as we normally keep documentation. 16:27:23 tibbs|w: I wonder, what prevents them from using /usr/share/doc/gap or similar 16:27:41 Probably nothing besides the will to avoid figuring out how to make that work and patching it in. 16:28:43 tibbs|w: Possible, I don't know ;) 16:31:20 Ok, so if we want to vote now and ask for a few kinks to be fixed later ... who is +1 for that? 16:31:25 +1 16:31:39 (after tibbs fix, of course) 16:31:40 +1 16:31:45 +1 16:32:00 Sorry, a line of people appeared at the door. 16:32:06 I'm still +1. 16:32:29 +1 16:32:52 +1 (just to give people something to play with) 16:33:39 orionp: want to vote? 16:33:59 Sure +1 16:34:06 #action GAP Packaging Guidelines (+1:7, 0:0, -1:0) 16:35:27 #info While approved a couple of things: 1) An example or two would be nice. 2) Might be nice to make a choice on the week deps. mapping, even if it's SHOULD and not MUST. 16:35:39 Anything else we want to ask? 16:36:21 yes, suggest writing an auto dependency generator 16:36:24 How about multilibs in gap-packages? 16:36:28 s/week/weak/ 16:36:34 #info 3) A macro to clean up the cruft from installed package so all packagers don't do it 666 different ways. 16:36:42 that too 16:37:00 #info 4) An auto dep. generator would be cool :). 16:37:08 racor: What about multilibs? 16:38:04 geppetto: I suspect gab i386/x86_64 packages to conflict 16:38:12 s/gab/gap/ 16:38:34 maybe ... could use requires _isa? 16:39:45 geppetto: I don't know ... To be investigated, IMHO. 16:40:08 I don't think they have any intention of making this multilib. 16:40:26 .me nods 16:40:33 ok ... 16:40:35 #topic #610 Packaging guidelines: Check upstream tarball signatures 16:40:40 .fpc 610 16:40:44 geppetto: #610 (Packaging guidelines: Check upstream tarball signatures) – fpc - https://fedorahosted.org/fpc/ticket/610 16:40:45 gap-devel is the only thing that's currently multilib-ed. 16:40:47 tibbs: Anything new for this week? 16:41:03 tibbs|w: Quite likely, but this often applies to upstreams. 16:41:18 geppetto: Not really. Some people have decided they don't like the way something works. 16:41:36 And some bugs were found, which isn't surprising. I'm dealing with those in the pagure repo. 16:41:51 I blew pretty much the whole week working on quick-fedora-mirror instead. 16:43:15 ok, no problem ... what is quick fedora mirror though, that sounds cool :) 16:43:30 https://pagure.io/quick-fedora-mirror 16:44:10 Ahh, cool. Di dyou look at how debian do this? 16:44:28 Bah, this is probably off topic anyway. But cool. 16:44:41 #topic Open Floor 16:44:44 So anything else? 16:45:01 There was the naming guidelines thing. 16:45:33 I started making those trivial changes and found out that the python section of that document was terrible and rather wrong. 16:46:03 So I did some cleanup. The result is at https://fedoraproject.org/wiki/Packaging:NamingGuidelines#Addon_Packages_.28Python_modules.29 16:46:44 Missing the % 16:47:01 Sorry, what's missing the %? 16:47:37 Weird, the link didn't take me to the right section 16:47:40 looks good, tibbs|w 16:47:50 I assumed it was missing %'s, but it's not it ... just buggy firefox I think 16:48:00 geppetto: worksforme ;) 16:48:23 Anyway, looks good 16:48:49 Anyway, that's already in there but I did it pretty late last night after losing my edits several times (thanks, wiki) so if you see anything stupid please let me know. 16:49:39 * orionp is fixing some typos/grammar 16:50:21 To be fair I haven't even re-read it after sleeping. 16:50:31 But feel free to fix away. 16:51:13 I don't quite follow the: Python2 module packages were named using a python- prefix. 16:51:24 as being previous convention 16:51:33 Isn't that the current convention? 16:52:34 or is that referring to the binary rpm? 16:53:04 Yes, that should be talking about the binary package. 16:54:06 Anyway, orionp and I can probably hash this out if we want to finish off the meeting. 16:57:03 thanks for working on on that 16:57:12 Sure 16:57:24 well, I have one bit of news to share: my unbundling workshop was accepted for Flock 16:57:28 I'll close at 1 unless someone else wants to talk about something 16:57:34 Rathann: cool 16:57:39 examples to tackle are welcome ;) 16:58:13 I have to duck out 16:58:26 And will not be present at next meeting (vacation, yay) 16:58:41 good for you, get some rest! 16:58:46 Enjoy the rest of your days :-) 17:01:34 #endmeeting