16:09:30 <quidam> #startmeeting Fri, 12/Jul/2019
16:09:30 <Trisnity> Meeting started Fri Jul 12 16:09:30 2019 UTC.  The chair is quidam. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:09:30 <Trisnity> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:09:36 <quidam> hi everybody
16:10:06 <quidam> sorry I was missing last week, I forgot to say I was going camping. That morning I had no connection to join in
16:12:06 <quidam> not many people around, I see :D
16:12:41 <dctrud> hello - first time here
16:12:43 <daroal> hi! sorry I was missing... a lot of weeks now
16:13:25 <quidam> daroal: no problem, glad to see you back!
16:13:35 <quidam> dctrud: hi, welcome!
16:14:18 <davidpgil[m]> hello -- hello
16:14:48 <quidam> nice, now we are getting quorum
16:16:42 <quidam> so, news from my side are that we did move the fsf server stack to a new location and it went really well
16:17:11 <quidam> there is a lot of space for community servers in there, and we can start populating it now
16:17:55 <quidam> I'm going to request 2U of space for the trisquel server, and prep the hardware we have for it
16:18:00 <davidpgil[m]> :rocket
16:18:05 <davidpgil[m]> 🚀
16:18:28 <daroal> sounds great!
16:18:45 <quidam> yeah, the current machine is way too clunky
16:20:20 <DiivaaD> sounds good quidam, will you use one of the ASUS KGPED-16 board?
16:22:13 <quidam> yes, one we already have
16:22:36 <quidam> it needs some extra components, which I will order
16:23:19 <quidam> and a chassis, which I can look into recycling from a decomissioned fsf server, or order a new one
16:23:31 <davidpgil[m]> noob question -- so this stuff helps at all because building packages from source takes a very long time?
16:24:58 <davidpgil[m]> ... the server would be dedicated to handling build requests or something?
16:25:12 <quidam> yes, it will host the build farm
16:25:44 <quidam> and other development services like gitlab (or whatever we choose)
16:26:10 <quidam> the key is that it will have so much ram all packages will be able to be built without using disk IO
16:26:22 <quidam> that speeds up builds by orders of magnitude
16:26:33 <davidpgil[m]> neat. i recently discovered through using Guix package manager that building your own software from source can take ridiculously long -- Emacs for example.
16:27:02 <davidpgil[m]> right cuz RAM is so much faster. Gotcha.
16:27:15 <quidam> but the secondary effect is that we will be rebuilding the development server, improving it
16:28:20 <davidpgil[m]> There was mention that Trisquel might become easier to contribute to -- this is part of that i would imagine...
16:29:04 <davidpgil[m]> How what else is needed for Trisquel 9 to be ready for release?
16:29:17 <quidam> yes, that is part of the plan too. an improved workflow on the git repo
16:29:28 <davidpgil[m]> *what else is needed for Trisquel 9 toi be ready for release?
16:29:40 <quidam> mainly to allow any devel to review and sign off other's commits
16:30:00 <davidpgil[m]> Cool.
16:30:12 <quidam> for the release, there are several key packages missing
16:30:25 <quidam> the kernel and the netinstall-related ones, to begin with
16:30:45 <quidam> or is the kernel done? I forget. one sec
16:31:12 <davidpgil[m]> I was supposed to test Trisquel 9's audio production tools. Have not gotten around to it. chaosmonk has been helping me get past snags.
16:32:07 <quidam> do you have a test installation of 9?
16:32:21 <davidpgil[m]> i had quite a few issues at first with doing a netinstall via "text mode". it seems many people have been having this issue.
16:32:32 <davidpgil[m]> is someone dedicated to fixing these issues?
16:32:43 <davidpgil[m]> yes, chaosmonk made a script to help me get it installed
16:33:21 <davidpgil[m]> let me check where i left off with him about my issuess... 1 sec
16:33:56 <dctrud> I'm keen to assist wherever, but have been struggling to find any kind of ToDo list.
16:34:21 <dctrud> Have submitted a couple of package-helper changes I've worked up as a result of using etiona as main desktop now
16:34:57 <quidam> dctrud: what DE are you using?
16:35:07 <dctrud> I'm using MATE
16:35:30 <dctrud> straight MATE on a desktop machine. MATE with i3 as the window manager on a small laptop
16:35:46 <quidam> where there installation issues with mate?
16:36:11 <davidpgil[m]> seems im at the point where i just need to test the ubuntu studio controls app with ardour and other sound tools.
16:36:31 <dctrud> Upgrading from an 8.0 install the main issue is the Trisquel theme being broken due to GTK3.20 breaking changes.
16:37:42 <dctrud> mate-tweak is missing - but the package-helper that was used previously is probably not needed, as it was doing a patch that is now in the upstream ubuntu version
16:39:13 <quidam> then I'll remove the helper and let it come from upstream
16:40:19 <dctrud> quidam: https://devel.trisquel.info/trisquel/package-helpers/merge_requests/224
16:41:45 <dctrud> other than that I looked at apache2, liferea, quassel - and have merge requests that may not be right... but did lead to working packages for me
16:44:34 <quidam> oh damn
16:44:47 <quidam> I deleted it myself before seeing your link
16:46:50 <quidam> ah, yours still applied
16:46:53 <quidam> all good
16:47:18 <quidam> now removing from the repo manager, one min
16:49:29 <quidam> done, now downloading
16:49:56 <quidam> I'll check your other MR now
16:51:59 <quidam> dctrud: checking the apache one, not sure I get the md5 one
16:52:57 <dctrud> The upstream package added a check that the `index.html` matches a specific md5 in the postrm script
16:53:22 <quidam> I got that, I didn't understand the sed command at first
16:53:27 <quidam> but it looks good
16:53:27 <dctrud> The patch inserts the trisquel index.html md5sum at the top of the list that the postrm is checking
16:54:22 <dctrud> it took me a bit.... it looks weird with all the "\ \ \ " to preserve leading spaces in the appended line
16:54:24 <quidam> merged
16:58:25 <quidam> and two others
16:58:39 <quidam> now checking the hpk one for security concerns
16:59:48 <dctrud> it's not ideal.... both networks I use block me pulling keys on default port 11371 for some reason
17:00:04 <dctrud> I can always just change it locally for my circumstances
17:00:46 <quidam> my suggestion, since we are upgrading that, is to use this instead
17:00:53 <quidam> hkps://keyserver.ubuntu.com:443
17:01:10 <quidam> it should work for constrained networks about the same as :80
17:01:15 <quidam> but adds extra safety
17:01:20 <dctrud> yeah, makes sense
17:01:35 <quidam> do you mind changing that in the MR?
17:01:53 <quidam> also, you should add your copyright line to helpers you contribute to
17:02:57 <dctrud> sure - the code is on desktop and I'm on laptop eating lunch, so will get to that in a bit
17:03:11 <quidam> no problem, I'm about to go for lunch too
17:03:19 <quidam> but first, about the apt sources list one
17:03:49 <quidam> I don't completely see the point, doesn't that disable sources.list.d completely?
17:04:10 <quidam> ohhh, never mind, that is for the helpers config
17:04:18 <quidam> I thought it was for general application
17:04:33 <quidam> merged
17:04:54 <dctrud> cool - glad these things were hopefully useful, if only small
17:04:54 <quidam> dctrud: excelent work! thanks so much
17:06:02 <quidam> any other topics? otherwise I'll go for a lunch break
17:06:39 <daroal> I'm still trying to keep up with all those nice MRs :-D
17:06:43 <dctrud> I'll step away now to finish lunch. Will get to that hkp small change later. Thanks, and have a good weekend.
17:08:09 <quidam> thanks everybody, see you in a bit!
17:08:17 <quidam> dctrud: thanks to you!
17:10:35 <daroal> thanks, enjoy your lunch(es)
17:56:07 <davidpgil[m]> 😋
17:58:48 <quidam> back
17:58:55 <quidam> I forgot to #endmeeting
17:59:00 <quidam> #endmeeting