15:01:16 <decathorpe> #startmeeting Stewardship SIG Meeting (2020-05-05)
15:01:16 <zodbot> Meeting started Tue May  5 15:01:16 2020 UTC.
15:01:16 <zodbot> This meeting is logged and archived in a public location.
15:01:16 <zodbot> The chair is decathorpe. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:01:16 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:01:16 <zodbot> The meeting name has been set to 'stewardship_sig_meeting_(2020-05-05)'
15:01:21 <decathorpe> #meetingname stewardship-sig
15:01:21 <zodbot> The meeting name has been set to 'stewardship-sig'
15:01:26 <decathorpe> #topic Knock Knock
15:01:37 <mhroncok> who's there?
15:01:49 <decathorpe> you got it :)
15:01:52 <decathorpe> #chair mhroncok
15:01:52 <zodbot> Current chairs: decathorpe mhroncok
15:02:49 <decathorpe> cipherboy, sillebille: meeting time
15:03:08 <sillebille> \o
15:03:20 <decathorpe> hello!
15:03:23 <decathorpe> #chair sillebille
15:03:23 <zodbot> Current chairs: decathorpe mhroncok sillebille
15:05:46 <decathorpe> #topic Agenda
15:05:52 <decathorpe> #link https://pagure.io/stewardship-sig/issue/87
15:06:04 <decathorpe> I think we might want to start with the OpenJDK 11 Change?
15:07:19 <mhroncok> decathorpe: I don't know where to start there actually
15:08:12 <decathorpe> yeah it looks pretty bad :O
15:08:55 <decathorpe> honestly I have no idea how other distros have dealt with this? I doubt it's our tooling alone that's responsible for all the failures
15:09:06 <decathorpe> #topic OpenJDK 11 fedora 33 Change
15:09:42 <decathorpe> #link https://fedoraproject.org/wiki/Changes/Java11
15:10:34 <decathorpe> #link https://copr.fedorainfracloud.org/coprs/jvanek/java11/monitor/
15:10:54 * mhroncok doesn't consider the results any good, it's Fedora 32 :/
15:11:10 <decathorpe> yeah I think they're starting to rebuild for rawhide now
15:11:59 <decathorpe> I've looked at some of the failures, and enforcing a 1.6 or 1.8 target / source level should fix at least some packages
15:12:56 <decathorpe> xmvn used to enforce a minimum of 1.6, but that has been dropped from xmvn 3.1.0 (which we dealt with when we updated it): https://github.com/fedora-java/xmvn/commit/c64c35a
15:13:04 <decathorpe> I guess that's biting us again
15:14:50 <decathorpe> the change owners recommend bumping to 1.8 for all packages ... which is probably a good idea anyway
15:14:50 <mhroncok> this is the point where my knowledge seems not up to speed
15:15:06 <mhroncok> the sources have hardcoded constraints to not use Java 11?
15:15:07 <mhroncok> or?
15:15:28 <decathorpe> no, some maven projects specify that their sources are Java 4 or something. OpenJDK does only support Java 8+
15:15:57 <mhroncok> decathorpe: I still don't follow, sorry :(
15:16:11 <mhroncok> java-11-openjdk
15:16:19 <mhroncok> is what we have now whenwe BR java
15:16:21 <mhroncok> right?
15:16:25 <decathorpe> yep
15:16:31 <mhroncok> so what does that have to do with 1.6 -> 1.8?
15:16:58 <decathorpe> some packages define in pom.xml that their sources are Java 5 or something. OpenJDK only supports building for Java 8+
15:17:07 <decathorpe> *OpenJDK 11
15:17:26 <decathorpe> and instead of enforcing the newer version, it fails.
15:17:30 <mhroncok> OpenJDK 11 isn't Java 11?
15:17:42 <mhroncok> oh
15:17:49 <decathorpe> it is, but it supports building sources for Java 8+
15:18:02 <mhroncok> so the soruces say: build me like I am one of the french girls
15:18:03 <mhroncok> sorry
15:18:09 <decathorpe> yeah
15:18:12 <mhroncok> so the soruces say: build me like I am java 5
15:18:26 <mhroncok> and the openjdk 8 said: sure, eneteryng my legacy mode or whatever
15:18:38 <mhroncok> but openjdk 11 says: java 5? go home, you are drunk
15:18:58 <decathorpe> as far as I understand, yes
15:19:25 <decathorpe> so dropping those broken specifications from pom.xml files should fix those packages
15:19:31 <mhroncok> and if we patch it to say: build me like I am java 8 it will either build (and therfore we assume it works) or we have an actinable error
15:19:39 <decathorpe> yes.
15:20:13 <mhroncok> decathorpe: is there some %magic for this, or it's plain old sed/patch game?
15:20:31 <decathorpe> uh .. sed/patch. unless you want to use XPATH
15:20:38 <decathorpe> which I hope you don't
15:21:02 <mhroncok> haha
15:21:04 <mhroncok> you bet
15:21:11 <mhroncok> parsing xml by regex is better
15:21:24 <decathorpe> but XML isn't a regular language, you say? ;)
15:21:35 <mhroncok> #info it seems the java11 copr has automation set, so when we open a PR, it will build in it
15:21:56 <decathorpe> oh, great. so we can start submitting PRs to proactively fix things?
15:22:08 <mhroncok> decathorpe: we can try with one thing, to see if the above is true
15:22:24 <mhroncok> but the header in https://copr.fedorainfracloud.org/coprs/jvanek/java11/builds/ indicates that
15:22:28 <decathorpe> yeah
15:22:46 <decathorpe> #action decathorpe to try if PR automation for java 11 rebuilds works
15:23:29 <decathorpe> Change owners also say if we encounter any packages that are really broken with Java 11 they will help
15:23:34 <decathorpe> sorry, OpenJDK 11
15:24:00 <mhroncok> decathorpe: try to PR, I'll copy pasta what you have and we can continue from there
15:24:18 <mhroncok> decathorpe: do we know of any real failures?
15:24:33 <decathorpe> there seem to be some ...
15:25:11 <decathorpe> glassfish-jsp-api: https://download.copr.fedorainfracloud.org/results/jvanek/java11/fedora-32-x86_64/01355164-glassfish-jsp-api/build.log.gz
15:25:41 <decathorpe> looks like this one fails in JavaDoc generation, because it references some missing / private symbols
15:25:58 <decathorpe> I seem to remember that private symbols are now *really* private with OpenJDK 11 :)
15:27:14 <mhroncok> decathorpe: as far as I am concerned, we can disbale the docs
15:27:21 <decathorpe> mhroncok: exactly
15:27:34 <decathorpe> javaparser is also broken, but newer versions support / need OpenJDK 11
15:27:59 <decathorpe> yeah most of the failures I see are JavaDoc problems
15:28:38 <decathorpe> if nothing else helps, we can just disable those
15:29:45 * decathorpe sighs at broken GCC on f32
15:30:27 <mhroncok> decathorpe: I must admit, I won't have time to help with actual changes
15:30:32 <mhroncok> sorry :(
15:30:38 <decathorpe> no problem
15:30:43 <mhroncok> I can help deciding things :D
15:30:54 <decathorpe> well, it's something. :)
15:31:05 <decathorpe> #action decathorpe to open TODO items for Stewardship SIG packages that are broken with OpenJDK 11
15:31:31 <decathorpe> I think I can depend on jvanek's COPR from another COPR, right?
15:31:49 <decathorpe> I might just start a rawhide rebuild of all stewardship SIG packages
15:33:22 <mhroncok> decathorpe: we can depnd on their copr, yes
15:33:44 <decathorpe> I'll see what I can do
15:33:55 <decathorpe> fixing SIG packages is our priority anyway
15:34:27 <mhroncok> getting rid of them is :/
15:34:34 <decathorpe> yeah ...
15:34:40 <decathorpe> oh great. java-8-openjdk fails to build on rawhide because of missing oh great. java-8-openjdk fails to
15:34:52 <decathorpe> copypasta #include <sys/sysctl.h>
15:34:57 <mhroncok> :D
15:35:10 * decathorpe sighs
15:35:55 <decathorpe> I wanted to poke the Java SIG, and it looks like there's https://pagure.io/java-sig , but it's empty for 6 months
15:37:31 <mhroncok> decathorpe: next topic?
15:37:44 <decathorpe> yeah this is making me sad
15:38:01 <decathorpe> #topic Review Open RHBZs
15:38:06 <decathorpe> #link https://bugzilla.redhat.com/buglist.cgi?bug_status=__open__&email1=stewardship-sig%40lists.fedoraproject.org&emailassigned_to1=1&emailcc1=1&emailtype1=substring&list_id=11039307&product=Fedora&query_format=advanced
15:38:19 <decathorpe> there's a new CVE for log4j
15:38:30 <decathorpe> other than that, I think the bugs are only anitya stuff
15:39:26 <mhroncok> decathorpe: low prio CVEs? let's ignore them?
15:39:54 <mhroncok> decathorpe: some of the anitya bgs are assigned to orphans, some for you
15:40:11 <mhroncok> decathorpe: shoudl we drop the orphane dones and make stewardship sig default assignee for the rest?
15:40:37 <mhroncok> This package is now maintained by the Stewardship SIG.
15:40:37 <mhroncok> https://fedoraproject.org/wiki/SIGs/Stewardship
15:40:37 <mhroncok> Any non-security updates will probably not be handled, if no volunteer steps up to do that. If you wish to volunteer assign this bug to you and reset the priority.
15:40:45 <mhroncok> should I mass post that to the anitya ones?
15:41:27 <decathorpe> I'm wondering why some bugs are assigned wrong
15:41:57 <decathorpe> though with new pagure we can at least set default assignee now :)
15:42:01 <mhroncok> decathorpe: the oprhaned ones are just listed becasue we are cced, I've fixed it now
15:42:13 <mhroncok> decathorpe: for mass changes, the PR thing was easier
15:42:13 <decathorpe> ah, good. thanks
15:42:33 <decathorpe> true ...
15:43:14 <decathorpe> I'll look through the bugs later, I think some of them might actually be fixed already
15:43:38 <decathorpe> #action decathorpe to go through RHBZs with fine comb and close fixed ones
15:44:53 <decathorpe> moving on?
15:46:18 <mhroncok> yes
15:46:27 <decathorpe> #topic Review Open PRs
15:46:32 <decathorpe> #link https://fedora-stewardship.github.io/pr-report/
15:46:51 <decathorpe> junit 4.13 breaks some things, and there's no other PRs right now. so not much to talk about
15:47:23 <sillebille> thanks for the reviews on felix-* though :-)
15:47:41 <decathorpe> sure. looks like we didn't break anything
15:47:53 <sillebille> yeah :-)
15:48:14 <decathorpe> #topic Review (SIG) Leaf Packages
15:48:27 <decathorpe> #link https://pagure.io/stewardship-sig/issue/84
15:48:46 <decathorpe> cipherboy has not orphaned his packages yet, and I can't do it for him
15:49:16 <mhroncok> decathorpe: I can
15:49:41 <decathorpe> oh right. that would be great
15:49:44 <mhroncok> decathorpe: is the stewardship sig removed?
15:49:48 <decathorpe> yes
15:49:55 <mhroncok> decathorpe: codemodel, glassfish-dtd-parser, glassfish-fastinfoset, istack-commons, and xsom?
15:50:00 <decathorpe> codemodel, glassfish-dtd-parser, glassfish-fastinfoset, istack-commons, and xsom
15:51:09 <decathorpe> mhroncok: many thanks :)
15:51:13 <decathorpe> #link https://fedora-stewardship.github.io/report/#total-leaves
15:51:39 <decathorpe> cipherboy wants us to keep jboss-el-3.0-api and jboss-servlet-3.1-api for now, since new resteasy versions will require them, I think
15:52:23 <mhroncok> Giving rpms/codemodel to orphan
15:52:24 <mhroncok> Giving rpms/glassfish-dtd-parser to orphan
15:52:24 <mhroncok> Giving rpms/glassfish-fastinfoset to orphan
15:52:24 <mhroncok> Giving rpms/istack-commons to orphan
15:52:24 <mhroncok> Giving rpms/xsom to orphan
15:52:45 <decathorpe> mhroncok++
15:52:45 <zodbot> decathorpe: Karma for churchyard changed to 3 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
15:53:03 <decathorpe> with the latest felix-* updates, felix-framework is now a leaf, which will also free up felix-osgi-obr-resolver
15:53:22 <decathorpe> so we should be able to drop those two packages with the next cleanup
15:56:39 <mhroncok> good
15:56:52 <decathorpe> #topic glassfish-jaxb subpackages
15:57:01 <decathorpe> #link https://pagure.io/stewardship-sig/issue/85
15:57:37 <decathorpe> the neuro SIG would like us to re-enable some subpackages we dropped, and they would be willing to help with maintenance
15:58:20 <decathorpe> I'll work them to figure out a way to do that without pulling in ancient dependencies again
15:58:27 <mhroncok> decathorpe++ thnaks
15:58:31 <mhroncok> *thanks
15:59:02 <decathorpe> #topic Dropping isorelax
15:59:06 <decathorpe> #link https://pagure.io/stewardship-sig/issue/86
15:59:26 <decathorpe> that's one of those ancient cruft packages, not touched upstream in 15 years or so
15:59:54 <decathorpe> we should be able to drop it, since the only way we depend on it is because it's BRd in jdom2 (but not Required)
16:00:06 <decathorpe> I'll see what I can do about that
16:01:32 <decathorpe> #action decathorpe to work on dropping isorelax from jdom2 build
16:01:36 <decathorpe> #topic Open Floor
16:04:52 <decathorpe> I guess there's nothing else :) let's wrap up.
16:04:55 <decathorpe> #endmeeting