19:37:05 <sumpfralle> #startmeeting
19:37:05 <MeetBot> Meeting started Wed Mar  4 19:37:05 2020 UTC.  The chair is sumpfralle. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:37:05 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
19:37:24 <sumpfralle> It is Wednesday evening again - time for our weekly IRC meeting.
19:39:28 <TheSnide> sumpfralle: hi
19:39:30 <TheSnide> indeed.
19:40:21 * TheSnide started (slowly) to work again on munin last week. Merely some bugfixes, and travis refactoring.
19:40:34 <sumpfralle> Sounds good!
19:40:50 <TheSnide> The testing pipeline should now be much more stable. no more (fragile) perlbrew.
19:41:18 <sumpfralle> I guess, I should port these changes to stable-2.0, correct?
19:55:51 <TheSnide> the travis ones ?
19:56:02 <TheSnide> yeah... might be a good idea indeed.
19:56:31 <TheSnide> Btw, explaining the rationale of the change here (so it is kept & archived)
19:57:05 <TheSnide> 1. historically one shall test with many perl versions
19:57:18 <TheSnide> 2. those perl version where moving quite fast in the past
19:57:38 <TheSnide> 3. nowadays, perl versions are pretty much stable.
19:58:04 <TheSnide> 4. we do _not_ use much exotic features of Perl. So for _us_ it should be super stable.
19:58:20 <sumpfralle> sounds reasonable to me
19:58:43 <TheSnide> 5. noone sane installs his own perl anymore. I mean, in our target users.
19:59:17 <TheSnide> 6. it makes total sense to ensure that a "distro based install" works nicely.
19:59:52 <TheSnide> 7. we cannot test all distro, so I picked the travis default.
20:07:30 <travis-ci> munin-monitoring/munin (stable-2.0 - ce8fbfa : Lars Kruse): The build has errored.
20:07:30 <travis-ci> Build details: https://travis-ci.org/munin-monitoring/munin/builds/656178812
20:07:53 <sumpfralle> Regarding the initial problem to be solved: the module Alien::RRDtool seems to have become unavailable for some reason. Do you have an idea, why?
20:08:11 <sumpfralle> (I just tested it above right now)
20:09:27 <TheSnide> yeah, it doesn't compile anymore. I tried to debug, but gave up and refactored it instead ;)
20:09:46 <sumpfralle> fair enough
20:10:02 <TheSnide> Alien::RRDtool is compiling some wrrapper to the system RRDs.pm
20:10:48 <TheSnide> otherwise, you need to compile rrdtool with --perl pointing to your perlbrew install. which would also be possible.
20:11:02 <TheSnide> but... i didn't pursue that road for the reasons I gave.
20:11:04 <sumpfralle> I do not intend to dig down that deep :)
20:11:53 <TheSnide> So, now the only "downside" is that we cannot tell Travis that we are a Perl-based project. It has to look like a "shell" one.
20:12:19 <TheSnide> and frankly, i don't even see that as a downside, since we now own totally the pipeline
20:12:51 <sumpfralle> Sounds good to me - I prefer packaging-based testing over the cpan-based one.
20:13:00 <TheSnide> --> i even debianize a cpan package 'on-the-fly"
20:13:44 <TheSnide> https://github.com/munin-monitoring/munin/commit/d515e3fefddf08199dda1a9585509368ba2d1eba
20:14:23 <sumpfralle> In a perfect world, it will be packaged somewhen by a helpful soul.
20:15:22 <TheSnide> well, the package is only relevant to a specific SaaS (coverall). So i doubt it will be packaged.
20:15:54 <sumpfralle> ok
20:17:30 <sumpfralle> At the moment I am still finishing my investigation regarding test flakyness in Debian's autopkgtest environment. I identified some small issues with the tests and the init scripts. The last small details is quite narrowed down, now. Thus I guess, I will finish it tomorrow (the tests need to run for some more ccycles until I catch it). Then I will do another 2.0 release with the change
20:17:32 <sumpfralle> of travis setup based on yours.
20:17:32 <TheSnide> I'm also looking as why CDEF doesn't work in 2.999. I thought it wasn't implemented, but it is.
20:17:42 <sumpfralle> good point!
20:18:08 <TheSnide> Same with the "supersampling".
20:18:33 <TheSnide> I might have to reset some RRD in dmm0
20:19:12 <sumpfralle> Progress shall not be stopped by old data :)
20:26:03 <sumpfralle> That's it for tonight's meeting?
20:27:13 <TheSnide> yep
20:27:15 <TheSnide> thx
20:33:16 <sumpfralle> #endmeeting