13:01:36 <matjazp> #startmeeting training-guides 13:01:37 <openstack> Meeting started Tue Aug 8 13:01:36 2017 UTC and is due to finish in 60 minutes. The chair is matjazp. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:01:38 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:01:41 <openstack> The meeting name has been set to 'training_guides' 13:01:46 <matjazp> hi ianychoi 13:02:02 <matjazp> how are you? 13:02:21 <matjazp> #link Agenda at https://etherpad.openstack.org/p/training-guides-meeting-agenda 13:02:46 <ianychoi> fine although i am now a little busy 13:02:50 <ianychoi> how about u? 13:03:08 <matjazp> ok 13:03:22 <matjazp> I'm still in Dublin, I'll be back in Slovenia in September 13:04:07 <ianychoi> Aha.. and how was in Korea? 13:04:26 <matjazp> Great 13:04:37 <ianychoi> :) 13:04:47 <matjazp> too bad that we didn't have time to meet 13:05:03 <matjazp> but I was there only for a few days 13:05:34 <matjazp> if you're busy, lets just start 13:05:52 <matjazp> #topic Moving from Hieroglyph for rendering slides 13:06:09 <matjazp> #link http://lists.openstack.org/pipermail/openstack-dev/2017-August/120598.html 13:06:34 <matjazp> #link https://bugs.launchpad.net/openstack-training-guides/+bug/1708240 13:06:34 <openstack> Launchpad bug 1708240 in OpenStack Training Guides "Incompatible requirements" [Undecided,New] 13:06:44 <ianychoi> yep I reas it 13:06:52 <ianychoi> s/reas/read 13:07:28 <matjazp> we already talked about moving to something else 13:07:55 <matjazp> now it seems that we have to do it, otherwise we will have problems with new versions 13:08:12 <matjazp> of openstackdocstheme 13:08:35 <matjazp> there are few suggestions on the ML 13:09:13 <matjazp> rst2html5 13:09:28 <matjazp> pandoc 13:10:26 <matjazp> hovercraft 13:10:32 <matjazp> that's it I guess 13:11:23 <matjazp> does you know perhaps if the conversion with rst2html works? or Pandoc? I think I remember that we already tried a few (years?) back and that it wasnt the best output 13:11:31 <ianychoi> hovercraft.. I don't know 13:11:52 <ianychoi> I want to try them :) 13:12:25 <matjazp> hovercraft was suggested by Markus Zoeller on the Docs ML 13:12:31 <matjazp> its his tool 13:13:12 * ildikov steps in and hides at the back of the room :) 13:13:16 <matjazp> maybe the process of conversion is better now 13:13:24 <matjazp> hi ildikov 13:13:32 <ildikov> how quickly do we need to decide and move? 13:13:38 <ildikov> matjazp: hi :) 13:14:01 <ildikov> sorry for being late, I have a parallel meeting 13:14:16 <matjazp> well, we can stay on older version of the openstackdocstheme for a while, but the real problem is sphinx 13:14:29 <matjazp> np ildikov 13:14:55 <matjazp> hieroglyph does not work with sphinx >= 1.6.0 13:15:20 <ildikov> so it's not Sphinx but hieroglyph that got so out of date we cannot use it anymore 13:15:22 <matjazp> and honestly, I think it's more or less a dead projectso we should move from it 13:15:33 <matjazp> ildikov: yes 13:15:34 <ildikov> yeah, agreed 13:16:13 <matjazp> I think that reveal,js is the obvious candidate 13:16:31 <matjazp> but it has a problem of natively supporting MD, not RST 13:16:34 <ildikov> do we have any constraints from infra side? 13:16:42 <ildikov> besides the Sphinx version 13:17:14 <matjazp> well I remember that this was one of the reasons we chose hieroglyph in the first place 13:17:37 <matjazp> support for it was already there, since it's just a addon for sphinx 13:17:38 <ildikov> hmm, it would be great not to switch to another file format as well 13:18:01 <matjazp> yes, there are converters 13:18:06 <matjazp> rst2html 13:18:09 <matjazp> pandoc 13:18:41 <matjazp> maybe we could use this to just render slides from the same source 13:18:45 <matjazp> this would be ideal 13:19:09 <matjazp> but I expect problems with this - it never works OOTB 13:19:50 <matjazp> we also have this integration with bug reporting 13:20:16 <matjazp> you know - that little bug link on the slides that directs you directly to the proper bug queue 13:20:46 <matjazp> so users can report bugs on the particular slide with one click 13:21:39 <fungi> i thought rst2html5 looked promising since you could still maintain your content in rst but render it to reveal.js like you were wanting 13:22:12 <ildikov> I guess we need to try it out and see how it works 13:22:14 <matjazp> fungi: yes, but we have to see how it renders slides 13:22:45 <fungi> granted you'd likely need to reformat the current corpus of slides some, but at least wouldn't need to switch to a completely different markup language than is used for other documentation in the community 13:22:53 <matjazp> I remember that we tried that a long time ago and results were not so good and we stick with hieroglyph 13:23:07 <fungi> ahh, too bad :/ 13:23:11 <matjazp> fungi: yes, I don't want to switch to MD 13:23:28 <matjazp> fungi : but things could improve in the mean time 13:23:41 <matjazp> maybe even pandoc works now :) 13:24:14 <ildikov> fungi: yeah, it's a priority not to change to another format 13:24:34 <fungi> great 13:24:38 <matjazp> hmm.. iare there any hard time limits regarding sphinx version from the infra? 13:24:43 <ianychoi> Personally I prefer to use rst2html5 but I do not have experience. Pandoc means that we need to convert to another format? 13:25:00 <ildikov> I wonder whether we could work with pdf files 13:25:16 <matjazp> ildikov: how do you mean? 13:25:57 <ildikov> I was just thinking out loud on what format we could render the slides to that we could use for trainings 13:26:38 <ildikov> let's start with trying out what we have and see where we are first 13:27:09 <matjazp> fungi: is from the infra point of view older sphinx ok? or what is the time frame for us to switch to >1.6? So we plan 13:27:37 <fungi> matjazp: if you can successfully install it from pypi, i don't think we care what version of sphinx you're on 13:28:06 <matjazp> fungi: great, I believe that buys us more time 13:28:21 <matjazp> but we'll be stuck with older version of openstackdocstheme 13:28:45 <matjazp> so migration is inevitable 13:29:06 <fungi> yeah, as long as your job/tox env/whatever you're using specifies the version(s) of sphinx you want (in a requirements.txt file maybe? i haven't looked at how yours is set up) then it should work just dandy 13:30:36 <matjazp> great. So we should just try rst2html and pandoc... any volunteers? :) 13:31:29 <ildikov> matjazp: I can take one 13:31:31 * matjazp is still on secondment in Dublin (so unfortunately no time for this) 13:31:42 <matjazp> ildikov: great 13:31:55 <matjazp> ildikov: rst or pandoc? 13:31:56 <ildikov> has either of you tried either of the options? 13:32:08 <fungi> #link https://twitter.com/faidonl/status/894694365154738176 13:32:16 <fungi> #undo 13:32:26 <fungi> sorry, prepping link from the wrong buffer 13:32:33 <fungi> chair will have to #undo 13:32:47 <ildikov> fungi: np, good message :) 13:32:50 <matjazp> #undo 13:32:51 <openstack> Removing item from minutes: #link https://twitter.com/faidonl/status/894694365154738176 13:32:52 <ildikov> matjazp: rst 13:32:59 <fungi> thanks, sorry 13:33:01 <matjazp> np 13:33:44 <matjazp> #action ildikov tries the conversion of slides with rst2html 13:33:45 <ianychoi> I want to try them but not now - maybe from September.. 13:33:45 <ildikov> matjazp: I need to look into these, I'm less knowledgable when it comes to presentation formats 13:34:16 <ildikov> matjazp: do you know how big work it would be to fix up hieroglyph? 13:34:33 <matjazp> ildikov: no, but is it even worth it 13:34:44 <matjazp> it's not maintained anymore 13:34:59 <matjazp> so we'll run into troubles later for sure 13:35:06 <ildikov> well, it depends what we find to replace it 13:35:28 <ildikov> so i just asked as plan C 13:35:48 <matjazp> I don't see no other long term options besides migration to something that is maintained - unless WE adopt maintenance of hieroglyph 13:36:27 <fungi> which, while noble, is not a decision to be taken lightly 13:37:41 <matjazp> agree 13:37:45 <ildikov> sure 13:38:05 <ildikov> was just an innocent question really :) 13:38:09 <matjazp> this looks like the problem with hieroglyph https://github.com/nyergler/hieroglyph/issues/124 13:38:41 <matjazp> ildikov: sure, maybe we'll need plan C, so good thinking 13:39:32 <matjazp> ok, so for a while, we're OK, as long we use the older version of sphinx and openstackdocstheme 13:40:02 <ildikov> matjazp: maybe someone likes it enough to fix it :) 13:40:04 <ianychoi> matjazp, yep - just we might need to specify sphinx version in test-requirements in training-guides repository? 13:40:18 <ildikov> matjazp: +1, hopefully we find a good enough alternative 13:41:06 <ianychoi> oh already specifyed: http://git.openstack.org/cgit/openstack/training-guides/tree/test-requirements.txt 13:41:24 <matjazp> ildikov: I'm still hoping that conversion will save us ;) 13:41:42 <ildikov> matjazp: fingers crossed :) 13:42:01 <ianychoi> :) 13:42:19 <matjazp> that's it for this topic, I guess 13:42:30 <matjazp> switch to AOB? 13:42:59 <ianychoi> Nope from me 13:43:09 <ildikov> nope from me either 13:43:12 <matjazp> #topic any other business 13:43:40 <ildikov> if either of you played with either of the alternatives earlier and have any heads ups I'm happy to learn them :) 13:44:03 <ildikov> otherwise, I'll play with it and see how things go 13:44:05 <fungi> oh, the actual link i meant to post 13:44:09 <fungi> #link http://lists.openstack.org/pipermail/openstack-dev/2017-August/120771.html New code contributors no longer forced to join the foundation 13:44:17 <matjazp> ildikov: I'll try to look for that old stuff 13:44:42 <fungi> that presumably has some impact on the upstream institute materials 13:44:44 <ianychoi> I was out of my home for about 20 minutes but if rst2html5 installation and testing works well, then I think choosing rst2html5 would be best - the change of rst syntaxes (from hydroglpyh to rst2html5 supported syntaxes) would be less then other options 13:45:32 <ildikov> fungi: looks good, thanks for the reminder 13:45:41 <fungi> i wasn't sure if the content managers for the training guides had seen the announcement on the change to the onboarding workflow 13:45:48 <ianychoi> fungi, thanks :) 13:45:57 <ildikov> fungi: yep, we will need to update a few slides 13:46:11 <matjazp> was that even a problem? 13:46:21 <matjazp> "forcing" to join I mean 13:46:41 <ildikov> fungi: always a good approach to give a heads up considering how busy the mailing list is :) 13:47:18 <matjazp> ildikov: yes, especially now that we'll merge docs ML with dev ML 13:47:21 <ildikov> matjazp: there are always complaints about that and the ICLA as far as I remember 13:47:22 <fungi> matjazp: yes, because you had to join the foundation before trying to sign the icla, and if your e-mail address didn't match perfectly between the foundation member profile and gerrit or you chose the wrong member type then you got a cryptic error message 13:47:30 <matjazp> so A LOT more mails in ML 13:48:27 <matjazp> fungi: yes, I know about those problems, but what I mean is that ppl will still join the Foundation, I believe 13:48:35 <ildikov> fungi: at least a half an hour on the training to figure out the root case of those errors... :) 13:48:57 <fungi> matjazp: not to mention, casual contributors (those coming from other free software communities, not so much the upstream institute audience) were potentially turned off by being required to join an organization just to push patches 13:49:07 <ildikov> fungi: as you need to fill out the contact info in Gerrit as well and there are issues time to time too, all joy :) 13:49:17 <ianychoi> ildikov, you are encountering some errors? 13:49:29 <fungi> yeah, no more contact info bits in gerrit 13:49:40 <matjazp> fungi: huh, yes, that's true... I all you want is to contribute a small patch, than joining was a bit of overkill 13:50:08 <ildikov> ianychoi: just the process to have the students signed up everywhere is always tricky 13:50:26 <ildikov> ianychoi: easy to make a mistake and not always easy to figure out which step went wrong 13:50:31 <fungi> i'm setting my sights on finally being able to drop icla enforcement in gerrit in another cycle or two, but that's a pretty major undertaking 13:50:49 <ianychoi> ildikov, aha I see. I think ICLA signing user interfaces in Gerrit does not seem user friendly 13:51:06 <ildikov> fungi: sounds good, I think we are having good progress to the right direction 13:51:36 <ildikov> ianychoi: with ICLA it's more about corporate concerns and so forth 13:51:53 <ianychoi> ildikov, aha yep also.. got it - thanks :) 13:51:55 <matjazp> ok, anything else left to discuss? 13:52:10 <ildikov> I'm good 13:52:16 <ianychoi> Nope from me - thanks, matjazp for organizing today meeting 13:52:27 <fungi> yeah, we've cleared the board of directors/foundation legal hurdle on dropping icla enforcement and switching to the dco a while back, but the mechanics of making that happen will still take a while 13:52:50 <matjazp> thanks a lot for your time 13:53:04 <matjazp> see you around in 2 weeks 13:53:10 <ildikov> thanks all! 13:53:22 <ianychoi> Thank you all! 13:53:31 <matjazp> bye all 13:53:33 <ildikov> fungi: yeah, those things are never easy 13:53:54 <matjazp> #endmeeting