14:31:00 <acozine> #startmeeting Docs Working Group aka DaWGs
14:31:00 <zodbot> Meeting started Tue Sep  1 14:31:00 2020 UTC.
14:31:00 <zodbot> This meeting is logged and archived in a public location.
14:31:00 <zodbot> The chair is acozine. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:31:00 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
14:31:00 <zodbot> The meeting name has been set to 'docs_working_group_aka_dawgs'
14:31:05 <acozine> #topic opening chatter
14:31:06 <felixfontein> hi!
14:31:13 <acozine> good morning, good afternoon, good evening!
14:31:19 <acozine> felixfontein: hi!
14:31:25 <acozine> who all is around?
14:31:29 <acozine> #chair felixfontein
14:31:29 <zodbot> Current chairs: acozine felixfontein
14:32:17 <samccann> me!
14:32:33 <felixfontein> #chair samccann
14:32:33 <zodbot> Current chairs: acozine felixfontein samccann
14:32:48 <acozine> gundalow: samccann aminvakil2 andersson007_ briantist geerlingguy madonius Tas-sos tremble wangbaoshan zbr you talking docs today?
14:32:58 * gundalow waves
14:33:36 <acozine> #chair gundalow
14:33:36 <zodbot> Current chairs: acozine felixfontein gundalow samccann
14:34:28 <acozine> today's agenda: https://github.com/ansible/community/issues/521#issuecomment-680112093
14:34:46 <acozine> as a reminder, everyone is welcome to add items to the agenda
14:35:38 <acozine> if you open a PR and would like reviews, or have a question, or anything - you can add it to the agenda ahead of time, or you can bring it up in the `open floor` at the end of the meeting
14:35:44 * zbr not really
14:35:53 * acozine waves at zbr
14:36:31 <acozine> let's talk about redirects first
14:36:51 <acozine> #topic redirects across the pre-collections/post-collections divide
14:37:08 <samccann> do we have a certain someone who may still be asleep? HE WHO SHALL NOT BE NAMED :-)
14:37:50 <acozine> heh, I think/hope he is still snoozing
14:38:18 <acozine> but I can present his latest work, I htink
14:38:46 <samccann> kewl
14:38:49 <acozine> he's running a web server to showcase the redirects from https://toshio.fedorapeople.org/ansible/2.10/collections/
14:38:54 <felixfontein> I saw there is a PR
14:39:20 <acozine> felixfontein: of the script? I haven't seen that yet
14:39:22 <acozine> link?
14:39:47 <felixfontein> #info https://github.com/ansible-community/antsibull/pull/185
14:39:57 <samccann> #info redirects in place on this webserver for test https://toshio.fedorapeople.org/ansible/2.10/collections/
14:41:28 <acozine> ah, he linked that in yesterday afternoon (US afternoon)
14:41:41 <acozine> looks like the builtin redirects work now, too
14:41:45 <felixfontein> just tested the links for docker_container, they seem to work fine
14:42:20 <samccann> so is this a general call for people to test out  that link and report problems here?
14:42:49 <acozine> samccann: yes, that sounds great
14:42:54 <felixfontein> I think so
14:43:19 <samccann> #action all - please test redirects to/from different versions at https://toshio.fedorapeople.org/ansible/2.10/collections/
14:43:23 <acozine> note that the version-switcher does not work
14:43:32 <acozine> on that test server
14:43:35 <felixfontein> 2.10 is not in the version selector though; you'll have to switch between devel and latest
14:43:37 <gundalow> acozine: Bullhorn will go out tomorrow, would it be worth including anything in that?
14:44:01 <acozine> gundalow: about testing the site?
14:45:19 <samccann> acozine - so should people hack the url from devel/2.10 to 2.9 etc?
14:45:37 <acozine> well, now I'm not so sure
14:45:46 <acozine> it might just not work back to 2.8
14:45:52 * acozine tries again
14:46:05 <felixfontein> it might be too early for throwing the public at it
14:46:37 <acozine> yeah
14:46:43 <felixfontein> 2.7 and 2.8 and 2.9_ja aren't there at all
14:46:53 <acozine> okay, that makes sense
14:47:15 <gundalow> acozine: yup, about testing site (or anything else we'd like extra eyes/help on)
14:47:50 <samccann> so, do we keep it in the minutes and say hack the url to 2.9/2.10? or remove the action entirely?
14:47:57 <acozine> if we can get this up on the semi-official testing site today, then yes, a request in the Bullhorn would be useful
14:48:05 <felixfontein> I think having it in the minutes is totally fine
14:48:12 <felixfontein> acozine: +1
14:48:14 <acozine> I don't want to send tons of traffic to Toshio's private server, though
14:48:31 <samccann> you think tons of people read our meeting minutes??? :-)
14:48:46 <acozine> no, but tons of people might read the Bullhorn
14:49:01 <samccann> ok so meeting minutes are a +1  bullhorn a -1 ?
14:49:04 <acozine> welcome snowolf
14:49:15 <acozine> meeting minutes +1
14:49:43 <samccann> #info you may have to hack the url on that test server from 2.9 to 2.10 and back to 2.9 to test. Other versions not available yet
14:49:51 <acozine> bullhorn also +1 but only if I can get this stuff generated and posted to docs.testing.ansible.com
14:50:37 <samccann> kewl
14:51:33 <samccann> also something to consider - blasting the test site to tons of people via the Bullhorn.  There was some unease about this in the past. I know we have the 'this is a test site' banner up, but is that enough?
14:52:50 <felixfontein> it's probably better to not spread it too much actively
14:55:07 <acozine> I'd like to get basic auth on that site, with the password listed on the screen where you log in
14:55:18 <felixfontein> :+1:
14:55:25 <acozine> just as a "really, this is not your regular docs site" reminder
14:58:12 <acozine> gundalow: I'll try to get the redirects up on the testing site by lunchtime here
14:58:25 <acozine> anything else on redirects?
14:58:41 * acozine is very happy to have them!
14:59:21 <acozine> #topic changelogs and porting guide updates
14:59:58 <acozine> I know we modified the Ansible 2.10 timeline, so we'll get more changelogs coming in this week, right?
15:00:07 <felixfontein> I was hoping so
15:00:18 <felixfontein> wasn't it today that there would be a bunch of releases?
15:00:41 <acozine> samccann: do you remember when those network collections were going to be released?
15:00:59 <samccann> the beta effort starts today, afaik yes. and yes, the network collection goal was to have things in place by yesterday I think
15:02:36 <acozine> sounds good
15:03:26 <acozine> other than merging PRs as they come in, how can we help?
15:04:06 <samccann> did a quick check and I see some of the network collections are rev'd as of yesterday or Friday.
15:05:59 <acozine> samccann: awesome
15:08:06 <acozine> gundalow: will the community team and the community handle generating changelog PRs?
15:08:10 <samccann> so afaik we have a beta build happening (either today or tomorrow? dunno if beta takes longer than alphas)
15:08:37 <acozine> yes, it's very exciting
15:09:03 <samccann> and afai also k - we don't have the ..erm...bitflip in place. So we will need to manually coordinate the build, the changelog and porting guide updates and backports, and the docs republishing to get everythhing in place hopefully the same time.
15:09:19 <acozine> right
15:09:24 <samccann> was there anything else that had to be coordinated?
15:09:34 <acozine> I plan to be glued to my chair on release day, pretty much
15:09:39 <samccann> #info  need to manually coordinate the build, the changelog and porting guide updates and backports, and the docs republishing to get everythhing in place hopefully the same time
15:09:49 <gundalow> acozine: changelog fragments should be added in the PRs as code is added. Is there something specific you are wanting to check?
15:09:50 <samccann> that could be today btw.
15:10:19 <acozine> gundalow: I'm not sure
15:10:43 <samccann> acozine are you asking about the Ansible changelog generation? or individual collection changelogs?
15:10:53 <acozine> just trying to make sure we aren't forgetting things
15:11:20 <acozine> but whatever we forget, someone will point out, and we can fix it then
15:11:47 <felixfontein> as long as the released collections include changelogs/changelog.yaml, everything should be fine - I hope :)
15:12:25 <acozine> I worry that there are some "oh I thought someone else was doing that part" moments ahead
15:12:39 <acozine> felixfontein: that's very reassuing
15:12:44 <acozine> reassuring
15:12:47 <samccann> so what is the order of the manual steps again? HE WHO SHALL NOT BE NAMED does a bunch of stuff, but doesn't publish to pypi until the Ansible changelogs are regenerated, the PR merged and backported?  Then pypi can happen at the same time as we republishe 2.10?
15:13:06 <felixfontein> I'm right now building a new ansible version, to see how changelogs are going
15:13:17 <acozine> samccann: yes, i think that's right
15:13:40 <felixfontein> samccann: yes, I think so too
15:13:48 <acozine> I figure as long as the new `latest` docs go up within 24 hours, we are okay
15:13:57 <acozine> we can hone the process next release
15:13:57 <felixfontein> devel should also be refreshed soon
15:14:07 <samccann> latest doesn't change at beta does it?
15:14:12 <acozine> nope
15:14:17 <felixfontein> both ansible-base 2.10 and ansible 2.10 link to devel for porting guides, not to 2.10
15:14:19 <acozine> I'm thinking of the final release
15:14:21 <samccann> so at the moment we are talking beta
15:14:51 <acozine> felixfontein: I think linking to devel for porting guides sounds fine
15:15:17 <felixfontein> acozine: I think so too. except that sometimes the porting guide in stable-2.x is different from the version in devel
15:15:23 <samccann> where are you seeing that felixfontein?
15:15:31 <acozine> in some ways, the porting guides section is like the release-and-maintenance page
15:15:37 <samccann> not really
15:15:47 <samccann> you port to a specific versin for Ansible.
15:15:54 <felixfontein> samccann: I saw that during some backports; changes made to the latest porting guide were changed to be changes in the previous version porting guide inside the backport
15:16:21 <acozine> felixfontein: ah, that happens when a bugfix gets backported
15:16:26 <felixfontein> so afterwards the porting guide for ansible-base 2.10 in devel was not having that info, while the porting guide in stable-2.10 had it
15:16:35 <felixfontein> acozine: yes
15:16:56 <acozine> oh . . . it shouldn't disappear from the later version porting guides though, should it?
15:16:56 <felixfontein> acozine: it's probably a good idea to compare them from time to time to see whether something is missing
15:17:08 <samccann> so i'm confused now.  the 2.10 porting guides should be identical in 2.10 and devel (once the backport is done)
15:17:23 <felixfontein> devel: change made to porting_guide_base_2.10.rst
15:17:33 <felixfontein> stable-2.10: change made to porting_guide_base_2.9.rst
15:17:37 <felixfontein> argh sorry
15:17:40 <felixfontein> stable-2.9: change made to porting_guide_base_2.9.rst
15:17:42 <samccann> :-)
15:18:00 <acozine> samccann: you're right, but we do keep the porting guides for older versions even in the devel docs
15:18:04 <felixfontein> but now porting_guide_base_2.9.rst in stable-2.9 has some additions which porting_guide_base_2.9.rst in devel does not have
15:18:13 <felixfontein> (I should have used 2.11 / 2.10, and not 2.10 / 2.9))
15:18:19 <acozine> ah, I see
15:18:20 <samccann> ^^ that is what shouldn't happen
15:18:52 <samccann> so devel could have 2.11 porting guide stuff that never shows up in stable-2.10
15:19:18 <samccann> stable-2.10 porting guide FOR 2.10 should never have more info than the devel version of the 2.10 porting guide
15:19:35 <acozine> yeah, we'll need to cross-check those
15:19:58 <felixfontein> samccann: yep, it shouldn't have, but that happened in the past
15:20:32 <felixfontein> the problem is the way backporting such fixes works...
15:20:59 <felixfontein> after the backport is merged, you need to make another PR on devel to also add the porting guide change from the backport to devel
15:21:33 <acozine> heh
15:21:51 <acozine> backports and foreports, it's a bit chaotic
15:21:53 <samccann> i'm just not following this, sorry
15:21:56 <felixfontein> fortunately such changes do not happen a lot :)
15:22:06 <samccann> oh I see, so a fix to JUST 2.10.
15:22:22 <samccann> and then, as acozine said - it's a foreport that is needed?
15:22:30 <acozine> fix happens in devel, porting guide entry goes in for 2.11 on devel
15:22:37 <samccann> ok
15:23:06 <acozine> backport of bugfix goes into 2.10, porting guide entry goes in for 2.10 on stable branch, but no porting guide entry for 2.10 on devel branch
15:23:18 <felixfontein> acozine: exactly
15:23:19 <acozine> (well, right now, it's 2.9, usually)
15:23:36 <felixfontein> I haven't checked for 2.9, but I've seen it for 2.10
15:23:59 <acozine> for now it's "let's fix this after the release", but eventually it would be good if we could come up with a process
15:24:23 <acozine> not sure what that would look like, with luck some brain cells will rejuvenate after the release
15:24:47 <samccann> #info problem to solve in the future- porting guides in devel may not be as up to date as a porting guide in earlier branches for released versions of ansible.
15:25:21 <acozine> we've got five minutes left in the official meeting time
15:25:45 <acozine> shall we skip to an open floor?
15:26:31 <acozine> #topic open floor
15:27:04 <acozine> we are down to 66 open docs PRs, which is close to our record low of 63 from last year sometime
15:27:13 <acozine> thanks to everyone who has reviewed PRs!
15:27:16 <samccann> woot woot!!
15:27:43 <samccann> ooo and down to 122 open issues!
15:27:55 <acozine> ooh, nice
15:28:04 <felixfontein> cool!
15:28:26 <acozine> that means 3 pages of docs PRs and 5 pages of docs issues on GitHub
15:29:07 <samccann> yep. progress for sure
15:29:35 <samccann> meanwhile, as we are open floor - does anyone have anything they want to bring up? favorite open issues or prs? some other thought about docs etc?
15:31:11 <samccann> ok gonna close out the meeting then ...going once...
15:31:31 <felixfontein> :)
15:31:36 <felixfontein> almost in time!
15:31:38 <samccann> oh wait! I had one item - felixfontein  - whjere should we coordinate the beta release
15:31:44 <samccann> here or on ansible-community?
15:32:06 <felixfontein> hmm, both is fine for me. especially since it's mainly T who is doing the work
15:32:37 <samccann> ok I'll ping him and then we can decide one channel to use - so we know where to look etc.
15:32:46 <samccann> going twice....
15:33:08 <samccann> #endmeeting