| *** markvoelker has joined #openstack-tc | 00:24 | |
| *** markvoelker has quit IRC | 00:51 | |
| *** markvoelker has joined #openstack-tc | 00:52 | |
| *** markvoelker has quit IRC | 01:01 | |
| *** ricolin has joined #openstack-tc | 01:03 | |
| *** markvoelker has joined #openstack-tc | 01:05 | |
| *** markvoelker has quit IRC | 01:11 | |
| *** markvoelker has joined #openstack-tc | 01:16 | |
| *** spsurya has joined #openstack-tc | 01:28 | |
| *** markvoelker has quit IRC | 03:10 | |
| *** altlogbot_0 has quit IRC | 03:16 | |
| *** altlogbot_1 has joined #openstack-tc | 03:18 | |
| *** markvoelker has joined #openstack-tc | 03:55 | |
| *** markvoelker has quit IRC | 04:00 | |
| *** markvoelker has joined #openstack-tc | 04:50 | |
| *** markvoelker has quit IRC | 04:54 | |
| *** jaosorior has joined #openstack-tc | 05:16 | |
| *** markvoelker has joined #openstack-tc | 05:35 | |
| *** markvoelker has quit IRC | 05:41 | |
| *** tdasilva has quit IRC | 06:47 | |
| *** tdasilva has joined #openstack-tc | 06:47 | |
| *** ricolin_ has joined #openstack-tc | 07:03 | |
| *** ricolin has quit IRC | 07:05 | |
| *** jaosorior has quit IRC | 07:23 | |
| *** ricolin_ is now known as ricolin | 07:33 | |
| *** e0ne has joined #openstack-tc | 08:15 | |
| *** jaosorior has joined #openstack-tc | 08:17 | |
| *** markvoelker has joined #openstack-tc | 08:20 | |
| *** markvoelker has quit IRC | 08:25 | |
| *** tdasilva has quit IRC | 08:48 | |
| *** tdasilva has joined #openstack-tc | 08:48 | |
| *** ricolin has quit IRC | 09:07 | |
| asettle | o/ hi all | 09:50 |
|---|---|---|
| asettle | Back in the northern hemisphere | 09:50 |
| gmann | asettle: welcome back. | 10:00 |
| asettle | Thanks :D | 10:01 |
| asettle | Nice to be bac | 10:01 |
| asettle | back* | 10:01 |
| jroll | welcome back asettle :) | 10:04 |
| asettle | \o/ | 10:04 |
| *** markvoelker has joined #openstack-tc | 10:36 | |
| *** markvoelker has quit IRC | 10:41 | |
| *** jaosorior has quit IRC | 11:18 | |
| *** markvoelker has joined #openstack-tc | 11:40 | |
| *** tdasilva has quit IRC | 11:51 | |
| *** jaosorior has joined #openstack-tc | 11:58 | |
| *** ijolliffe has joined #openstack-tc | 12:27 | |
| *** tdasilva has joined #openstack-tc | 12:27 | |
| openstackgerrit | Merged openstack/governance-sigs master: Ansible SIG https://review.opendev.org/676428 | 12:28 |
| openstackgerrit | Merged openstack/governance master: linters: enforce ATC membership after January 2020 https://review.opendev.org/675902 | 12:33 |
| openstackgerrit | Merged openstack/governance master: Add Michał Dulko as Kuryr PTL https://review.opendev.org/674624 | 12:33 |
| *** smcginnis is now known as smcginnis|pto | 12:34 | |
| *** ijolliffe has quit IRC | 12:36 | |
| *** ricolin has joined #openstack-tc | 12:37 | |
| *** tdasilva has quit IRC | 12:42 | |
| *** ijolliffe has joined #openstack-tc | 12:43 | |
| *** mriedem has joined #openstack-tc | 13:18 | |
| asettle | tc-members - the PDF train goal may need a little technical help. stephenfin is pretty underwater, like we all are, but it would be appreciated if some could help us out with testing. Also https://etherpad.openstack.org/p/pdf-goal-train-common-problems | 13:18 |
| asettle | Help review this list that Akihiro is working through | 13:18 |
| *** tdasilva has joined #openstack-tc | 13:19 | |
| amotoki | asettle: I read your mail which hits me a question. we are exploring two approaches: sphinx latex builder and rst2pdf. Does your mail mean we abandon the sphinx latex builder way? | 13:20 |
| asettle | amotoki, good question. stephenfin thoughts? I didn't mean the email to suggest that. But at the same time, we're putting a lot of effort into this and I think it would be hard to just abandon it | 13:24 |
| amotoki | asettle: my main point is that exploring two approaches at this moment means individual teams need to explore their own ways (or duplicated works). | 13:27 |
| amotoki | asettle: IMO this kind of thing should be explored before it was accepted as the community goal. Otherwise the goal requires individual teams to spend duplicated efforts.. | 13:27 |
| asettle | Yep, and that's a good point. | 13:27 |
| stephenfin | amotoki, asettle: To be honest, I think a bit of duplication is a good thing. Much of what we fix for the LaTeX-based build will be used regardless of what we opt for, since the constraints are very similar (single doc, can't use HTML-style links, etc.) | 13:28 |
| stephenfin | So I was hoping people could look at the different approaches and see which one is ultimately the better option | 13:29 |
| amotoki | yeah, it makes sense for me to explore two approaches if it is in the early stage, but we are in the milestone 3 so I am afraid this goal fits as train goal. | 13:29 |
| amotoki | stephenfin: but such thing is expected as what individual teams explore as part of the community goal? | 13:30 |
| stephenfin | I understand, but I hadn't personally expected the latex builder to result in so many issues, especially since it's part of the main Sphinx package | 13:31 |
| amotoki | stephenfin: Problems I see now are only two things: (1) too large literalinclude for oslo.config sample files and (2) relative links | 13:33 |
| *** ijolliffe has quit IRC | 13:33 | |
| stephenfin | This is for neutron, yes? There are many more problems in nova, unfortunately :( | 13:33 |
| amotoki | ah... | 13:34 |
| stephenfin | For me, it was easier to add Python 3 support to rst2pdf than it was to learn LaTeX :) | 13:34 |
| amotoki | sounds fair | 13:34 |
| stephenfin | But because I can't build LaTeX-based PDFs for nova, I don't have anything to compare the output of rst2pdf with | 13:34 |
| stephenfin | Hence my request for someone with a little more time than I have right now to check out rst2pdf and see if it's "good enough" | 13:35 |
| stephenfin | If that makes sense? | 13:35 |
| amotoki | stephenfin: yes, it makes sense. It is a reasonable reason to explore another approach (rst2pdf). | 13:36 |
| *** jeremyfreudberg has joined #openstack-tc | 13:46 | |
| mnaser | o/ | 15:00 |
| mnaser | bonjour | 15:00 |
| fungi | apologies in advance, i can't be around for office hour today | 15:01 |
| mnaser | seems like a quiet one | 15:03 |
| * mnaser is still struggling to land any code to openstack/magnum | 15:04 | |
| corvus | stephenfin, asettle: hi, i'm a latex expert | 15:04 |
| fungi | i'll do my best to catch up on scrollback once i get home | 15:04 |
| corvus | if there is some technical challenge with latex that i can help with, i would like to | 15:04 |
| jroll | \o | 15:05 |
| corvus | however, i do not want to take on yet another lost cause, so if ya'll have decided to do rst2pdf just let me know and i won't worry about it | 15:06 |
| stephenfin | corvus: Finally someone that actually knows _anything_ about it :) | 15:06 |
| corvus | but i haven't followed the pdf goal *at all* so would need some pointers to catch up :) | 15:06 |
| stephenfin | corvus: I think I'd really like to see what rst2pdf can do for us first. The issue with the latex approach is that there are so many issues and they're so varied between projects | 15:06 |
| asettle | corvus, WELL HELLOOOO | 15:07 |
| asettle | Sorry, bit late to the party | 15:07 |
| corvus | i'm pretty sure *i'm* the one late to the party :) | 15:07 |
| zaneb | I'm sort-of here | 15:07 |
| stephenfin | corvus: tl;dr: we're building docs using Sphinx's latex builder instead of the html builder. That results in a load of auto-generated tex sources that are compiled with latexmk (or mklatex, I don't recall which) | 15:07 |
| asettle | zaneb, are you ever in the party? | 15:07 |
| zaneb | no | 15:07 |
| stephenfin | *we're attempting to build docs | 15:08 |
| asettle | Okay well that solved that then | 15:08 |
| * mnaser doesn't know much but perhaps those 99% of those latex issues are probably going to be common fixes that we can document, and then the 1% that projects can clean up | 15:08 | |
| mnaser | (ill go back to being quiet about things i dont know about :)) | 15:08 |
| stephenfin | corvus: but this is failing to compile for larger projects and because we don't know LaTeX, we don't know if it's because (a) our docs are using some construct that isn't supported by the latex builder, (b) one of the many extensions is doing something wrong, or (c) the core latex has some actual bugs | 15:09 |
| corvus | oh, we should be so lucky to have (c) :) | 15:10 |
| dhellmann | do we have a list of the blockers? | 15:10 |
| dhellmann | everything I see in https://etherpad.openstack.org/p/pdf-goal-train-common-problems could be worked around in the short term | 15:10 |
| stephenfin | dhellmann: amotoki has been putting together one for neutron. I couldn't get the nova docs to generate a single thing though | 15:11 |
| dhellmann | remember, the goal is not to build "perfect" PDFs, but to successfully produce *any* PDF | 15:11 |
| dhellmann | then later we iterate and improve | 15:11 |
| dhellmann | nova had an extension that imported an svg, and we were going to have to convert that image to another format. Was there something else? | 15:11 |
| stephenfin | so the blocker for nova is pretty much "it doesn't work", and I don't know enough to refine it any further /o\ | 15:11 |
| corvus | so should i take "get something readable for nova" as a first task? | 15:12 |
| dhellmann | ok, having some error messages would help in debugging :-) | 15:12 |
| stephenfin | corvus: That would be a good target, yeah | 15:12 |
| fungi | also a clear set of steps to follow after cloning nova to reproduce the problem maybe? | 15:12 |
| dhellmann | is the doc job running the pdf build steps, yet? | 15:12 |
| dhellmann | having that in place will make it easier for more people to experiment and iterate | 15:13 |
| stephenfin | I have the patch up but I haven't touched it in weeks | 15:14 |
| corvus | stephenfin: have a link to that? | 15:14 |
| stephenfin | sec | 15:15 |
| stephenfin | https://review.opendev.org/#/c/664555/ | 15:15 |
| corvus | cool, i think i have enough to go and make problems and try to fix them | 15:15 |
| stephenfin | Yeah, https://review.opendev.org/#/c/664555/2/roles/build-pdf-docs/tasks/main.yaml@15 is the thing you'd want if you were running this locally | 15:16 |
| amotoki | question on the doc job. do we install latex packages as part of this job or do each project need to install them via bindep.txt? | 15:16 |
| stephenfin | amotoki: I was hoping to leave that to someone more informed to decide | 15:16 |
| corvus | i'll try to report back with something (even if it's "i'm still stumped") next week | 15:17 |
| amotoki | stephenfin: okay | 15:17 |
| dhellmann | amotoki : projects should NOT need to modify their bindep files | 15:18 |
| fungi | which of those is more correct can be a bit of a philosophical discussion. if we think folks are ever going to want to try and build pdfs locally, then including the texlive or similar package dependencies for some common platform in bindep.txt (perhaps with a specific/custom profile name associated with them like "pdf") would be nice | 15:18 |
| stephenfin | amotoki: The aim for this was to try do it without having to do anything in the project (so no new tox targets or, I guess, bindep files) | 15:18 |
| stephenfin | but then again, we recently got rid of that default bindep file so by not changing bindep, perhaps we're regressing a little | 15:18 |
| fungi | but it's perfectly fine in my opinion for the pdf builder job to have a list of packages it installs, especially as an early solution | 15:18 |
| dhellmann | right, the point of this phase of implementation is to try to get the automated systems to build PDFs with projects only changing content | 15:18 |
| *** jeremyfreudberg has quit IRC | 15:18 | |
| stephenfin | like I said, I'll let someone more informed than me make the call | 15:18 |
| dhellmann | later we can provide instructions or tools for folks to do the build locally | 15:19 |
| corvus | doing it in a pre-run playbook should be fine | 15:19 |
| asettle | corvus, if you could track your issues here: https://etherpad.openstack.org/p/pdf-goal-train-common-problems that would be amazing | 15:21 |
| amotoki | corvus: if you work on the pdf doc job, some existing PDF build patches (per project) would help too. https://review.opendev.org/#/c/667345/ and https://review.opendev.org/#/c/667968/ | 15:23 |
| amotoki | bindep.txt there is a working set at least. | 15:23 |
| corvus | asettle, amotoki: got it, thx :) | 15:28 |
| asettle | Thank you corvus for reaching out :D | 15:29 |
| corvus | i just hope i can come back with something other than "turns out i am not a latex expert after all" :) | 15:30 |
| fungi | if nothing else, that might also bait mtreinish ;) | 15:30 |
| asettle | LOL fungi | 15:31 |
| asettle | Just place little nibbits along the way until he takes the bait | 15:31 |
| *** markvoelker has quit IRC | 15:44 | |
| *** markvoelker has joined #openstack-tc | 15:48 | |
| *** tdasilva has quit IRC | 15:55 | |
| *** markvoelker has quit IRC | 16:02 | |
| *** markvoelker has joined #openstack-tc | 16:11 | |
| mtreinish | fungi: heh, you just need to say LaTex 3 times and I'll appear :P | 16:11 |
| *** ricolin has quit IRC | 16:17 | |
| *** e0ne has quit IRC | 16:28 | |
| *** e0ne has joined #openstack-tc | 17:52 | |
| *** e0ne has quit IRC | 18:36 | |
| *** spsurya has quit IRC | 18:53 | |
| *** camelCaser has quit IRC | 19:28 | |
| *** gouthamr is now known as gouthamr|brb | 20:02 | |
| *** gouthamr|brb is now known as gouthamr | 20:30 | |
| *** camelCaser has joined #openstack-tc | 20:38 | |
| *** mriedem has quit IRC | 21:00 | |
| *** tdasilva has joined #openstack-tc | 21:27 | |
| *** markvoelker has quit IRC | 21:29 | |
| *** markvoelker has joined #openstack-tc | 22:05 | |
| *** markvoelker has quit IRC | 22:09 | |
| *** markvoelker has joined #openstack-tc | 22:11 | |
| *** markvoelker has quit IRC | 22:40 | |
| *** markvoelker has joined #openstack-tc | 23:45 | |
| *** markvoelker has quit IRC | 23:50 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!