15:00:18 #startmeeting Stewardship SIG Meeting (2019-07-23) 15:00:18 Meeting started Tue Jul 23 15:00:18 2019 UTC. 15:00:18 This meeting is logged and archived in a public location. 15:00:18 The chair is decathorpe. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:18 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:00:18 The meeting name has been set to 'stewardship_sig_meeting_(2019-07-23)' 15:00:23 decathorpe: hi 15:00:25 #meetingname stewardship-sig 15:00:25 The meeting name has been set to 'stewardship-sig' 15:00:31 #chair mhroncok 15:00:31 Current chairs: decathorpe mhroncok 15:00:34 hi :) 15:02:04 decathorpe: JFYI https://churchyard.fedorapeople.org/orphans.txt is now generated ina loop so should always be at most hours old 15:02:12 great 15:02:24 I updated the Agenda with the latest status a few hours ago 15:02:55 (unfortunately it takes ~10 hours to genearte) 15:03:18 wow that's bad 😅️ 15:03:37 is that constant time or is it faster if there are fewer orphans? 15:03:47 decathorpe: it depands on number of rohans 15:03:50 orphans 15:03:58 rohans? wtf 15:04:13 so at least there's hope it will get faster over time 15:05:34 by the way, there's now an up-to-date page with statistics of outdated packages 15:05:36 https://decathorpe.fedorapeople.org/stewardship-sig-stats.html 15:06:08 (some packages aren't tracked in anitya, and that's been broken for a few weeks anyway) 15:06:33 do we have the energy to update the packages? 15:06:55 I have boring holidays now, which is why I've been working on updates 15:07:01 decathorpe: ok 15:07:03 good luck 15:07:46 thanks 😣️ 15:08:00 let's start. maybe someone else will show up later 15:08:05 #topic Agenda 15:08:17 #link https://pagure.io/stewardship-sig/issue/38 Agenda 15:09:18 any preference what we should talk about first? 15:09:35 I'd liek to talk about gradle 15:09:45 that thing is totally broken 15:09:54 yeah it's not looking good 15:10:15 this is also disheartening: https://decathorpe.fedorapeople.org/stewardship-sig-stats.html#gradle-stats 15:10:35 still a better story than zinc 15:10:55 anyway, what exactly happens if we orhpan gradle and groovy? 15:11:14 there's a few non-orphan packages that rely on gradle for building 15:11:31 either they take gradle and fix it, or those will stop working 15:12:57 I've looked at gradle, but I know if something's above my paygrade when I see it 15:13:04 decathorpe: non-orphan but not ours, correct? 15:13:09 correct 15:13:15 gimme a second 15:13:24 I guess we should offer gradle/groovy to them and orphan it otherwise 15:13:31 that was my idea 15:13:36 ok, let's do it 15:14:25 so the packages built with gradle-local are: 15:14:35 archaius (already broken) 15:14:57 ecc-25519-java (quasi-orphan / gil) 15:15:10 fernflower 15:15:30 forbidden-apis 15:15:42 gpars (ours, needed only for gradle and groovy) 15:15:44 orphaned ecc-25519-java 15:15:59 gradle (BRs itself) 15:16:17 groovy (ours, outdated) 15:16:33 hystrix 15:16:56 mongo-java-driver (should probably be retired now mongo is gone?) 15:17:08 openjfx 15:17:28 procyon 15:17:45 reactive-streams (quasi orphan / gil) 15:17:59 rsyntaxtextarea 15:18:24 spock (already broken, already orphaned) 15:18:25 Orphaned reactive-streams 15:18:39 sshj 15:18:44 (gil) 15:19:05 Orphaned sshj 15:19:10 zkclient (already broken, gil) 15:19:31 orphaned zkclient 15:19:46 so, these are left: hystrix, mongo-java-driver, openjfx, procyon, rsyntaxtextarea 15:20:01 fernflower, forbidden-apis... 15:20:06 right 15:20:15 7 packages 15:20:19 what needs groovy? 15:20:40 stand by ... ... 15:21:12 some more packages, most of them I recognise as already broken 15:21:25 I can post a list to the ticket, it's too many to check live 15:22:03 ok 15:22:52 anything else gradle-related for now? 15:24:00 decathorpe: not really, but unless we can safely drop groovy, we are probably stuck with it, right? 15:24:17 I think so 15:24:48 some packages already have bconds for groovy, and for others we can probably disable scripting functionality 15:25:14 would you mind prioritizing that over updates? 15:25:24 not at all 15:25:45 I'm already checking if we can safely drop spring 15:25:55 gradle/groovy would be the next task 15:26:38 thanks 15:26:50 I'm sory that i don't hav emore tiem to actually help you with that 15:27:07 no need to apologize 15:27:33 I wonder what the mass rebuild will uncover 15:27:42 😱️ right 15:28:21 maybe some more people will realize in how bad a shape things are 15:29:50 decathorpe: new required orphaned packages... ? 15:30:14 erm .... newly orphaned packages that SIG packages depend on 15:30:34 decathorpe: sure, let's review them? 15:30:54 #topic Newly Orphaned Packages 15:31:17 I'd like to postpone deciding on anything spring-related 15:31:29 until after we've decided if we can drop spring or not 15:31:53 \o Sorry, I'm around now. Was on PTO two weeks ago. 15:31:55 google-oauth-java-client is also gradle related 15:32:07 hi cipherboy 15:32:13 #chair cipherboy 15:32:13 Current chairs: cipherboy decathorpe mhroncok 15:32:51 well, there's no downside to taking packages now, and once we don't need them anymore, we can orphan them again 15:33:12 it seems that everything listed is just 0-1 weeks old, we can defintivelly go trough them later 15:33:24 absolutely 15:33:27 Sure sounds good to me. 15:34:15 but I think we'll definitely need: apache-common-vfs, lz4-java, mina-ftpserver, spock, testng 15:34:31 decathorpe: OK, shall I rewuest them? 15:34:57 if you want to 15:35:02 I don't :D 15:35:06 I need to schedule some time to put into the SIG again. I can try and review these new packages and what we require them for. Maybe we can drop a few dependencies. 15:35:08 haha then I'll do it 15:35:10 I can take them. 15:35:42 good. just ping me in the ticket, so I can update my data 15:36:16 decathorpe: anything else in the agenda to go trough? I don't feel like reviewing PRs now 15:36:26 I understand completely :) 15:37:47 cipherboy: do you think we can work on dropping some JBoss packages? 15:38:11 dogtag-pki is pulling these in, but some of them only on fedora and not in EPEL 15:38:27 decathorpe: Probably -- is that something DT requires? I think we never fully supported running on JBoss. We can take a look at that and see what breaks. :) 15:38:46 that would be great. I think on EPEL it uses resteasy instead 15:38:54 (which is why I've fixed resteasy) 15:39:23 Ah, we should only need resteasy on Fedora / RHEL 8+. Let me check something quick... 15:39:51 decathorpe: Yeah, I think we can drop a portion of those. Let me look at them more closely and I'll get back to you. 15:39:57 great 15:40:16 it would be great if you could comment there: https://www.pagure.io/stewardship-sig/issue/33 15:40:33 ACK will do. 15:40:37 +1 15:40:58 just for completeness, is there any reason you can think of that would prevent us from dropping springframework support? 15:40:59 Oh I don't actually get notifications when I'm @mentioned for some reason. 15:41:35 Spring? Isn't that the GUI front end? The only thing I can think of is perhaps ESC (smart card management in Dogtag), but I'll talk to someone and see what that uses. 15:41:44 great, thanks 15:41:56 If Dogtag doesn't use it, then yeah, I'm all for dropping that :D 15:42:13 I don't think it does 15:43:11 Then yeah, I'm fine dropping that. 15:43:17 do you know if we can patch jackson to not require groovy? I think it's only needed for tests 15:43:56 I haven't looked at jackson much, I'll see if I get time to look at that. 15:44:20 it's a prerequisite for dropping groovy/gradle 15:44:22 https://www.pagure.io/stewardship-sig/issue/35 15:44:24 thanks .) 15:47:23 decathorpe: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/thread/UNTGT65DKYADSJTPQUKTSTU2ERGGLFT6/ 15:47:34 yeah I saw 15:48:16 once we own javapackages-tools, we can finally start *really* messing with things :P 15:48:45 fwiw, we tried poking this internally and got nowhere, so... 15:49:47 honestly I'm getting a bit tired of this. everybody expects these things to keep working, but nobody wants to spend time to keep it working 15:50:34 And when it doesn't keep working, the people who are stuck with the side effects get the burden, not those that caused the problem.. 15:50:56 or, in this case, people just left 15:51:08 That too. 15:51:27 there's going to be about 500 more orphaned Java packages soon 15:51:57 500?! 15:52:10 I'm also juggling two more non-responsive maintainer processes, which will increase that number further ;) 15:52:12 What's getting removed from the ecosystem? 15:52:14 https://pagure.io/releng/issue/8498 15:52:38 Oh wow. 15:52:44 Uhm. 15:53:01 We might want to try standardizing how we maintain packages if we do need most of these. 15:53:05 so effectively, the entire Java stack is getting orphaned in the next few weeks 15:53:12 right 15:54:04 I'll be sure to tell management. I'm working on critical stuff internally as well (on top of a move), so time might get tight. 15:54:14 Doubt we can do much though. 15:54:32 thanks 15:54:50 * cipherboy shrugs 15:54:54 I'll be talking about this at flock as well. 15:55:03 maybe I can make people realize that this is not sustainable 15:55:36 decathorpe: you have a talk, or just want to talk to people? 15:55:58 I have a talk 15:56:10 https://pagure.io/flock/issue/208 15:56:52 good 15:57:04 I'll try to shame people into helping us as hard as I can 😂️ 15:57:15 not sure if shaming will help 15:57:16 :D g'luck! 15:57:28 mhroncok: I know. that was sarcasm ;) 15:57:34 ok :D 15:59:39 let's end? 15:59:45 sure 15:59:50 thanks guys 15:59:56 Thanks! 16:00:13 I'll update the gradle/groovy ticket with more impact data later today 16:01:09 #endmeeting