14:59:58 #startmeeting "NeuroFedora - 2019-09-05" 14:59:58 Meeting started Thu Sep 5 14:59:58 2019 UTC. 14:59:58 This meeting is logged and archived in a public location. 14:59:58 The chair is dan1mal. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:59:58 Useful Commands: #action #agreed #halp #info #idea #link #topic. 14:59:58 The meeting name has been set to '"neurofedora_-_2019-09-05"' 14:59:58 dan1mal: https://fedoraproject.org/wiki/Meeting:Guide 15:00:11 #meetingname neurofedora 15:00:11 The meeting name has been set to 'neurofedora' 15:00:20 #topic Introductions and Roll call 15:00:21 Please use #info 15:01:08 #chair MeWjOr mhough FranciscoD alciregi blackfile 15:01:08 Current chairs: FranciscoD MeWjOr alciregi blackfile dan1mal mhough 15:01:10 #info ankursinha (FranciscoD), UTC+1 (Europe/London), packaging, neurofedora, and other Fedora tidbits 15:01:45 .fas dan1mal 15:01:45 dan1mal: dan1mal 'Danny Lee' 15:01:58 #info fas dan1mal, UTC-04:00, NeuroFedora Spin, QA, Design 15:01:59 #info major IST (+5:30) packaging, figuring out how the linux kernel works, optimizing the OS for my old laptop 15:02:29 #action Dan1mal sends out logs to mailing lists 15:02:33 .hello2 15:02:34 alciregi: alciregi 'Alessio Ciregia' 15:03:10 Sorry for the late reminder, I hope it didnt cause too much confusion. 15:03:21 * dan1mal likes more than 1 day's notice usually 15:03:34 #info mhough, UTC-8:00, pakaging, pipelines and brains 15:03:40 No, had it on our calendars already :) 15:04:11 Reminders are an opportunity to get more people involved. We got a reply on the -devel list (to be discussed after the meeting maybe) 15:04:19 chair zbyszek 15:04:27 #chair zbyszek 15:04:27 Current chairs: FranciscoD MeWjOr alciregi blackfile dan1mal mhough zbyszek 15:04:29 .hello2 15:04:30 zbyszek: zbyszek 'Zbigniew Jędrzejewski-Szmek' 15:04:54 I would have liked to get more info together before hand. All the moving parts makes it challenging 15:05:38 Do you think we can start or another minute for anyone else 15:06:00 I think it's better to start. 15:06:05 #topic Today's agenda 15:06:05 #info Pagure tickets at: https://pagure.io/neuro-sig/NeuroFedora/issues?status=Open&tags=S%3A+Next+meeting 15:06:05 #info Open bugs at: https://bugzilla.redhat.com/buglist.cgi?bug_status=NEW&bug_status=ASSIGNED&email1=neuro-sig%40lists.fedoraproject.org&emailassigned_to1=1&emailcc1=1&emaildocs_contact1=1&emaillongdesc1=1&emailqa_contact1=1&emailreporter1=1&emailtype1=substring&list_id=10455921&query_format=advanced 15:06:05 #info Review of current progress 15:06:06 #info Action items from previous meeting 15:06:08 #info Planning what we want to do next 15:06:10 #info Neuroscience query of the week 15:06:12 #info Next meeting chair selection 15:06:14 #info Open floor 15:06:59 Starting with the first: 15:06:59 #topic Review of pagure tickets 15:07:55 ticket #276 - major filed ticket https://pagure.io/fedora-infrastructure/issue/8159 , response from @zbyszek, @kevin and @pingou 15:08:03 chair #mewjor 15:08:11 #chair mewjor 15:08:11 Current chairs: FranciscoD MeWjOr alciregi blackfile dan1mal mewjor mhough zbyszek 15:08:36 I guess we can talk about the tickets that were worked on this past week 15:08:39 You seem to have added me twice (as MeWjOr and mewjor) 15:08:57 won't matter :) 15:09:04 * zbyszek didn't do anything during the last week 15:09:06 * dan1mal blushes, sorry first time trying to run meeting 15:09:22 dan1mal: could you info the thing about ticket #276 too please---so it shows up in the logs 15:09:33 #info https://pagure.io/neuro-sig/NeuroFedora/issue/276 -> Issue #276: Package descriptions at Fedora Package Sources 15:09:36 #276 is in action, and are waiting for an update from @Pingou 15:09:38 Haha, take your time. The whole point of rotating chairs is for everyone to learn how to go about it :) 15:10:47 MeWjOr: if you have the cycles, maybe worth asking pingou if we can help with the script---they're quite busy managing all the infra, so any thing we can do to reduce their workload would be nice :) 15:11:26 The best is propose a script 15:11:32 Maybe i can help 15:11:38 #chair bt0dotninja 15:11:38 Current chairs: FranciscoD MeWjOr alciregi blackfile bt0dotninja dan1mal mewjor mhough zbyszek 15:11:50 I would love to do that... But, I am a bit occupied the next two weeks (exam weeks) 15:11:53 should I action it to you @bt0dotninja? :P 15:12:00 Sure 15:12:03 Uh, MeWjOr, you'd better focus on the exams then 15:12:11 :P 15:12:17 #info MeWjOr has exams so avoid volunteering them for tasks :P 15:12:18 Right Compadre 15:12:29 Good luck MewjOr! Fighting! 15:12:40 #action bt0dotninja check out script required for https://pagure.io/fedora-infrastructure/issue/8159 15:12:46 Eyy... Thanks guys! <3 15:13:03 dan1mal: #274 next? 15:13:11 #info https://pagure.io/neuro-sig/NeuroFedora/issue/274 -> Issue #274: Flock: conference reports 15:13:15 (so I can poke everyone to write their Flock reports again? :P) 15:13:33 :) 15:13:45 Please write your reports and include neurofedora in them---that'll be good outreach/marketing for the team 15:13:55 You can use the NeuroFedora blog if you wish :) 15:14:27 I would have to reconfigure my whole "blog", so that it can send RSS feeds as well... So, it will take some time for me to send it on planetfedora 15:14:39 speaking of @zbyszek did any of the fenics packages from the hackathon get submitted? 15:15:18 I see an old dolfin in bugzilla 15:15:21 MeWjOr: ah? let's discuss that after the meeting---whatever platform you're on, it should have an rss feed generator 15:15:45 mhough: should we discuss that in the open floor so dan1mal can proceed with the agenda, do you think? 15:15:47 It's just an extension from my static website... *let's talk after the meeting* 15:15:59 ok, next ticket then? 15:16:04 okay 15:16:06 Yes please :) 15:16:07 #info https://pagure.io/neuro-sig/NeuroFedora/issue/199 -> Issue #199: #199 Computational neuroscience labs image 15:16:25 The good news is the trademark vote passed right as we finished our meeting 15:16:36 dan1mal: the meeting manager is also required to keep all us talkative folks in check and on topic :P 15:16:54 The next step was to try and generate a new live cd and see if selinux was an issue 15:17:05 +1 15:17:07 by bt07? 15:17:35 the selinux is enabled and enforcing in my last build 15:17:50 Yes, I remember you mentioning it somewhere. This is after the PR, right? 15:17:54 https://pagure.io/neuro-sig/neuro-scripts/pull-request/4 15:18:09 in the ticket, also added an screenshot 15:18:47 #info bt0dotninja's latest build has selinux enabled and enforced: https://pagure.io/neuro-sig/NeuroFedora/issue/199#comment-595246 15:19:03 that's great! 15:19:06 :D 15:19:23 Thats very good news, and will help move forward everything towards the deadline 15:19:25 :) 15:19:39 dan1mal: the next step would be the Change proposal page, would it? 15:19:57 issue #198 right? 15:20:36 I don't think a new comps group requires a change proposal---a PR to comps should be enough 15:20:41 Im not really sure what a change proposal page is 15:20:48 the new lab image will reqiuire a change proposal 15:21:06 dan1mal: I've done a change proposal before, I can set it up and then we can fill the details in together. How does that sound? 15:21:22 that sounds great! 15:21:37 #action FranciscoD set up a new change proposal page for the lab 15:21:53 dan1mal: for comps, a PR to comps would be the next step---anything blocking that? 15:21:53 #action dan1mal to follow change proposal page for lab process 15:22:07 ticket #198 - preparing to file PR for comps.xml - how to group software 15:22:08 #info https://pagure.io/neuro-sig/NeuroFedora/issue/198 -> Issue #198: #198 Computational neuro comps group 15:22:34 The only thing left was how to arrange the software into groups 15:22:45 Ah, yeh: https://pagure.io/neuro-sig/NeuroFedora/issue/198#comment-594981 15:22:52 zbyszek what do you think? 15:23:39 I'm leaning towards a neuron-modelling group, and a brain-modelling group (which will repeat some of the neuron-modelling packages, since one can't include a group in another in comps) 15:23:56 Or, non-neuron-brain-modelling? 15:23:59 XD 15:24:07 too long! :P 15:24:42 I need to revise how the other groups are 15:24:45 * FranciscoD runs dnf grouplist 15:26:09 dan1mal: if the packages aren't too large, we could just have one group for "brain modelling" and put everything in it 15:26:10 maybe brain-imaging 15:26:32 mhough: yes, that'll be a separate group for the neuro-imaging packages 15:26:32 dan1mal: it is possible to include a group in another group 15:27:06 Also, I wouldn't sweat the division into subgroups too much. 15:27:10 zbyszek: it doesn't look like it, there are categories which can contain multiple groups but they are broader 15:27:12 mhough: we're hoping to do the modelling group first so that it simplifies the kickstarts for the ISO image (we'll have a different ISO image for neuro-imaging) 15:27:20 ahh ok 15:27:38 ugh, I don't think I know the difference between a category and a group.. 15:27:49 so tvb-* would be brain-modeling for instance? 15:27:51 or "environment groups" and "groups" 15:28:00 zbyszek: individual pkgs can be marked required, default, optional. 15:28:04 mhough: when we do get all done (after they move to python3), yes 15:28:40 #info Comps lives here: https://pagure.io/fedora-comps 15:29:40 dan1mal: so, do I have this correct: Categories are the highest level, and they can contain different groups 15:29:53 #info Some info on using comps.xml: https://fedoraproject.org/wiki/How_to_use_and_edit_comps.xml_for_package_groups 15:30:04 So, "Neuroscience" will be a category, and it will include the comps groups 15:30:33 And these categories are only used by Anaconda, the installer. DNF uses groups, not categories 15:31:48 ah, they dont define category there 15:32:20 dan1mal: let's forget about categories for the moment---they're relevant for the installer only, and we're going to provide ready-to-use images 15:33:19 I see, okay, so just putting the software into groups, but how many groups? 15:33:20 Please look how Fedora Workstation is defined 15:33:21 https://pagure.io/fedora-comps/blob/master/f/comps-f32.xml.in#_6350 15:33:31 It includes gnome-desktop group. 15:33:50 * FranciscoD waits for the large file to load 15:33:51 * dan1mal waits for the link to load. 15:34:04 probably easier for me to clone the repo and use vim.. 15:34:07 haha ! it took so long 15:34:12 And dnf grouplist shows "Fedora Workstation" 15:34:33 It doesn't show GNOME because it has uservisible=false, but it otherwise would. 15:34:35 * dan1mal just learned you can URL link to lines in files on pagure.io 15:34:50 So you can just use the same structure if necessary to put groups together. 15:34:58 yep, my idea is use that structure to define the neuro fedora image 15:35:35 @zbyszek well I learned something today 15:35:38 adding groups and removing/tunning at package level 15:36:53 OK, so how about the environments map to our 4 types of software: Neuro + {Modelling, Imaging, Analysis, and Utilities (optional maybe?)} 15:37:07 (since they are sufficiently different) 15:37:34 Perhaps, we should document on the ticket and move on, just to cover other topics 15:37:41 and neuron-simulators can be a group to .. group them together 15:37:41 +1 15:37:43 it make me sense 15:37:43 +1 15:37:48 +1 15:37:53 * FranciscoD copy-pastes comment 15:38:06 Ok, moving on to: 15:38:09 #topic open bugs on bugzilla 15:38:16 #link https://bugzilla.redhat.com/buglist.cgi?bug_status=NEW&bug_status=ASSIGNED&email1=neuro-sig%40lists.fedoraproject.org&emailassigned_to1=1&emailcc1=1&emaildocs_contact1=1&emaillongdesc1=1&emailqa_contact1=1&emailreporter1=1&emailtype1=substring&list_id=10455921&query_format=advanced 15:38:33 #action zbyszek dan1mal bt0dotninja mhough MeWjOr comment on environment/group organization on #198 15:38:54 dan1mal: I made a tinyurl to make it easier in the future: https://tinyurl.com/neurofedora-bugs 15:38:59 #last meeting we had new releases of biosig and airspeed 15:39:04 Much better 15:39:08 .thank FranciscoD 15:39:08 dan1mal thinks FranciscoD is awesome and is happy they are helping! (Please also type FranciscoD++ since that is what gives them a cookie) 15:39:12 #info short URL for neurofedora bugs: https://tinyurl.com/neurofedora-bugs 15:39:18 biosig is yet to be merged 15:39:23 ah, that's on me 15:39:29 #action FranciscoD review and merge biosig PR 15:39:36 #action airspeed merged and in testing 15:39:42 #undo 15:39:42 Removing item from minutes: ACTION by FranciscoD at 15:39:36 : airspeed merged and in testing 15:39:48 #info airspeed merged and in testing 15:40:13 Anything else on this? Or next topic form last week: why major is unable to assign bugs to self 15:40:18 Question: How long does a package stay in testing? 15:40:46 dan1mal: we can skip that one for the time being---I expect MeWjOr will be able to work with the new bugs, otherwise we'll speak to infra 15:41:02 ok on to: 15:41:10 until the karma reach +3 and non -1 15:41:16 #fsleye broken because of sip and wpython 15:41:24 #info fsleye broken because of sip and wpython 15:41:34 #info Updates policy set out by FESCo: https://docs.fedoraproject.org/en-US/fesco/Updates_Policy/#stable-releases 15:41:37 I read this: wpython changes made in rawhide, removed sed patches and reverted to wx.siplib 15:42:08 FranciscoD: Thanks! 15:42:10 dan1mal: we're waiting on the changes to make it to the stable releases: https://bugzilla.redhat.com/show_bug.cgi?id=1739469#c39 15:42:38 Ok, any other bug related topics? 15:42:44 that would be great 15:43:05 There are a few FTBFS (failed to build from source) bugs, but I think they've all been reported upstream already 15:43:46 https://github.com/INCF/nineml-python/issues/45, for example, if anyone wants to do some python hacking 15:44:10 or https://github.com/nipy/nitime/issues/136 -> nitime 15:44:12 Quick question, are you guys all python programmers? 15:44:28 * FranciscoD dabbles in lots of languages 15:44:37 Python, C++ and a bit of Java 15:44:45 once you know one or two, it's easy to learn others 15:44:49 Everybody is a python programmer. Maybe they just don't know it yet. 15:44:55 XD 15:44:56 lol! 15:45:01 .thank zbyszek 15:45:01 MeWjOr thinks zbyszek is awesome and is happy they are helping! (Please also type zbyszek++ since that is what gives them a cookie) 15:45:09 zbyszek++ 15:45:09 FranciscoD: Karma for zbyszek changed to 4 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 15:45:33 some of the python in these packages is quite involved, but you learn so much while reading their code 15:45:47 ok then, moving on, bug-talk, going once... 15:45:54 haha 15:46:02 Not bug related, but I forgot to mention it in the Pagure issues topic. I will be working on python-pymanopt. I have a spec file ready, but am waiting for an update from @mhough for some minor changes. 15:46:07 zbyszek++ 15:46:08 mhough: Karma for zbyszek changed to 5 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 15:46:24 #action FranciscoD update bug link on blog and docs with short link 15:46:29 thanks for libsbml too 15:46:36 zybszek++ 15:47:01 #topic neuroscience of the week section 15:47:27 This is a time for any neuroscience related topics that anyone might have 15:47:56 Because Fedora Neuro-sig isnt just about software, but getting to know more about neuroscience and computational neuroscience! 15:48:13 and neuroimaging! and how the brain works! 15:48:16 (in general) 15:48:30 :D 15:48:37 well just for this section, i was looking around and found this fun link: 15:48:41 #link MarI/O - https://www.youtube.com/watch?v=qv6UVOQ0F44 15:49:30 Oo! More machine learning 15:49:39 Fun. 15:49:44 Ahh is it RL? 15:49:47 * dan1mal is barely treading water with fedora, but adding neuro-science is threatening to pull me under 15:50:01 haha 15:50:25 lots of work there nowhttps://deepmind.com/blog/article/alphastar-mastering-real-time-strategy-game-starcraft-ii 15:51:13 from what I see, most neuroscience folks tend to stay away from ML/Deeplearning because it's considered an application of what we learn about the brain. 15:51:26 (only anecdotal evidence, so does not generalise to everyone) 15:51:43 in terms of the neurofedora software, its more involved in modelling the way brains work? 15:52:01 yes, modelling actual neurons, for example 15:52:14 there are some interesting applications of RL in my psychiatric field though 15:52:31 RL = reinforcement learning? 15:52:35 yes 15:52:39 Most of these new algos aren't even RL 15:52:46 they're unsupervised! 15:52:52 that's what makes them "cool" 15:53:03 this is typically used by people who classify themselves as computational psychiatrists 15:53:08 Isn't RL supposed to be unsupervised? 15:53:19 chuck lots of data at them, and let them figure out patterns themselves 15:53:22 Not when we do it with animals 15:53:40 exciting stuff! 15:53:59 ok, next up: 15:53:59 #topic Next meeting chair selection 15:54:03 RL is a third category, along with supervised and unsupervised from what I remember 15:54:10 Oo, fun topic! 15:54:20 So, who wants to manage us next week? :P 15:54:24 Actually some of the more recent work I have been following adds "curiosity" and "exploration" 15:54:48 * dan1mal thinks a meeting about just neuroscience/updates in the field might go over well 15:55:16 If someone can suggest a reading for me on irc admin I can do it 15:55:47 mhough: https://fedoraproject.org/wiki/Meeting:Guide#MeetBot_Commands 15:55:48 For more curiosity, we read about GANs today. An example is here: https://thispersondoesnotexist.com/ 15:55:49 there you go :P 15:56:05 if found these links that helped me 15:56:09 cool thanks 15:56:12 #link https://github.com/coreos/fedora-coreos-tracker 15:56:21 #link https://fedoraproject.org/wiki/Meeting:Guide 15:56:29 #link https://fedoraproject.org/wiki/Zodbot 15:56:36 so mhough chairs the next meeting then? :D 15:56:50 in a week's time? 15:57:04 ok happy to do that 15:57:11 and then I took our last log at https://meetbot.fedoraproject.org/ copied it to text file for cut and paste 15:57:16 pagure.git.receive -- lbazan pushed 1 commit to neuro-sig/documentation (master) https://pagure.io/neuro-sig/documentation/branch/master 15:57:28 dan1mal: I reckon the first neuroscience oriented meeting will need to differentiate the scope of neuroscience and ML etc 15:57:37 mhough: nice, thanks :) 15:57:54 :) 15:57:56 #info mhough to chair next meeting on 12th September, 2019 15:58:03 FranciscoD: yeah, this probably should be somewhere for volunteers to read for initial FAQs/intro 15:58:19 maybe could be generated organically by collecting irc talks 15:58:36 .thank mhough 15:58:36 dan1mal thinks mhough is awesome and is happy they are helping! (Please also type mhough++ since that is what gives them a cookie) 15:58:45 #topic Open floor 15:58:50 I have to take my daughter to school so bye 15:58:51 dan1mal: Yeh, we'll try and discuss a bit of it each meeting 15:59:01 https://www.cnsorg.org/computational-neuroscience -> computational neuroscience 15:59:01 Ok, thanks mhough! Have a safe trip! 15:59:14 ++ 15:59:24 thanks bye everyone 15:59:30 bubye! 15:59:33 bye 15:59:45 ok, closing in on 12 16:00:00 zbyszek: I think we can proceed with https://bugzilla.redhat.com/show_bug.cgi?id=1739786 now 16:00:28 pagure.git.receive -- lbazan pushed 1 commit to neuro-sig/documentation (master) https://pagure.io/neuro-sig/documentation/branch/master 16:00:42 an announcement on the macros will be made soon: https://pagure.io/packaging-committee/issue/918 16:01:02 * FranciscoD will comment on the bug too 16:01:13 #action FranciscoD comment on https://bugzilla.redhat.com/show_bug.cgi?id=1739786 16:01:25 FranciscoD: sorry, I was busy, and didn't have any time to look at the reviews. 16:01:27 I noticed a lot of the tickets on pagure (like 75%) are software tickets. Perhaps, down the road we could think about separating software from SIG development issues... 16:01:52 Hopefully next week. 16:01:57 dan1mal: yeh, we didn't want to start by over-organising 16:02:08 I did break the tags down to different dimensions, though: https://pagure.io/neuro-sig/NeuroFedora 16:02:31 So, if you filter by the right tag, it'll limit your view 16:02:38 zbyszek: awesome, no hurry :) 16:03:01 yeah, its totally fine. i just noticed a lot of software. 16:03:02 pagure.git.receive -- lbazan pushed 1 commit to neuro-sig/documentation (master) https://pagure.io/neuro-sig/documentation/branch/master 16:03:19 dan1mal: at the moment, that's our primary task, unfortunately---packaging it all up 16:03:22 Ok, if there's nothing else maybe I could close the meeting and anyone who wants can stay and talk about further issues? 16:03:32 once we've crossed that milestone, we move on to other tasks 16:03:35 +1 for close 16:03:54 Hereee 16:04:01 Finally... 16:04:10 #info mhough to chair next meeting on 12th September, 2019 16:04:20 Welcome! 16:04:52 Question, when you see +1 for close, how many do you need to close? 16:05:00 one? 16:05:19 Or when you see this in other places on fedora...is it as long as there are no -1? 16:05:33 dan1mal: different teams have different agreement processes 16:05:56 so, could be: "no net negatives in a week", for example 16:06:21 for closing a meeting, you neednt wait for votes. If we're over the 1 hour time, you go "Ok, time up, closing!" :P 16:06:26 dan1mal: just wait 30 seconds for anyone to speak up, and close. 16:06:31 +1 16:06:34 ok, time up! closing! :) 16:06:35 #endmeeting