14:35:14 <randomuser> #startmeeting Docs Project Meeting - Agenda: https://fedoraproject.org/wiki/Docs_Project_meetings
14:35:14 <zodbot> Meeting started Mon Nov  6 14:35:14 2017 UTC.  The chair is randomuser. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:35:14 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
14:35:14 <zodbot> The meeting name has been set to 'docs_project_meeting_-_agenda:_https://fedoraproject.org/wiki/docs_project_meetings'
14:35:14 <randomuser> #meetingname Fedora Docs
14:35:15 <zodbot> The meeting name has been set to 'fedora_docs'
14:35:15 <randomuser> #topic Roll Call
14:35:19 <bexelbie> .hello bex
14:35:24 <zodbot> bexelbie: bex 'Brian (bex) Exelbierd' <bex@pobox.com>
14:35:28 <randomuser> .hello immanetize
14:35:29 <zodbot> randomuser: immanetize 'Pete Travis' <me@petetravis.com>
14:35:39 <randomuser> moving right along,
14:35:46 <porfiriopaiz> .hello porfiriopaiz
14:35:47 <zodbot> porfiriopaiz: porfiriopaiz 'Porfirio Andrés Páiz Carrasco' <porfiriopaiz@gmail.com>
14:35:47 * porfiriopaiz from Managua in Nicaragua
14:35:52 <randomuser> #topic Centos CI update from Bex
14:35:56 <randomuser> #chair bexelbie
14:35:56 <zodbot> Current chairs: bexelbie randomuser
14:36:03 <x3mboy> .hello2
14:36:04 <zodbot> x3mboy: x3mboy 'Eduard Lucena' <eduardlucena@gmail.com>
14:36:16 <randomuser> go go gadget bexelbie
14:36:49 <bexelbie> ack
14:36:50 <bexelbie> ok
14:37:04 <bexelbie> I had a meeting with Nigel and bstinson (separately)
14:37:06 <bexelbie> I have hte access
14:37:09 <bexelbie> I have some templates
14:37:12 <bexelbie> I have nothing running yet
14:37:17 <bexelbie> two blockers (one sort of resolved)
14:37:40 <bexelbie> 1. I still don't see the trigger mechanism - this requires a follow up with bstinson -- I am now setting my goal on something I can push a button and get usefulness out of
14:37:54 <bexelbie> 2. the council-docs needs ditaa to build which is a non-starter for CentOS
14:38:06 <bexelbie> so I have temporarily removed that build dep by building the one diagram and committing it
14:38:24 <bexelbie> so I have a path forward
14:38:34 <randomuser> one CI issue, one content issue; that's not too bad
14:38:36 <bexelbie> and I think I have a new idea for the CI job structure that will simplify builds
14:38:47 <williamjmorenor> .fas  williamjmorenor
14:38:48 <zodbot> williamjmorenor: williamjmorenor 'William Moreno' <williamjmorenor@gmail.com>
14:38:53 <bexelbie> short-version: Each repo that is published on docs.fp.o
14:39:02 <bexelbie> will have a "trigger" for a CI job that will do two things.
14:39:09 <bexelbie> 1. build all PRs and stage them for review
14:39:49 <bexelbie> 2. Build all commits on production branches .. if it passes it will commit the output to a master source repo ... if it fails it will complain to the repo owners
14:40:04 <bexelbie> a second job will watch the master source repo and rebuild and republish as needed
14:40:16 <bexelbie> this prevents us from getting blocked by a random documentation collection having a bad publication branch
14:40:24 <bexelbie> this master source repo doesn't exit yet --- but will soon
14:40:24 <randomuser> sounds good so far
14:40:25 <bexelbie> :)
14:40:41 <randomuser> that's not https://pagure.io/fedora-docs/docs-fp-o ?
14:41:04 <bexelbie> it will be :)
14:41:10 <randomuser> i see
14:41:11 <bexelbie> right now, no real source lives there
14:41:13 <bexelbie> soon it will
14:41:35 <randomuser> like, in an output branch?
14:41:36 <bexelbie> basically it will clean the build up and get us more functionality at the expense of a bit of disk
14:41:41 <bexelbie> input branch
14:41:45 <bexelbie> branches
14:41:48 * randomuser nods - disk is cheap
14:41:54 <bexelbie> output will still live in the output repo
14:41:58 <bexelbie> fedora-docs-web, iirc
14:42:01 <randomuser> ah
14:42:22 <randomuser> and intermediary builds would go in docs-fp-o
14:42:53 <bexelbie> not really
14:43:04 <bexelbie> stage builds would be hosted on the artifacts server in the CI
14:43:13 <bexelbie> at least that is my understanding
14:43:19 <bexelbie> there is no need to commit those anywhere, imho
14:43:25 <bexelbie> barring objections
14:43:56 <randomuser> ok, rereading, I think I follow you
14:44:18 <bexelbie> I'll try to write something sane to the ML
14:44:25 <bexelbie> however, I am traveling most of this week and next week
14:44:29 <randomuser> do the rest of us have any direct visibility on the CI jobs?
14:44:34 <bexelbie> so I may write an "I did this, and now we can debatee it and I will fix it"
14:44:47 <bexelbie> I haven't been told how to get the jobs to feedback to PRs - but that should happen
14:44:58 <randomuser> ok, cool
14:45:00 <bexelbie> as for rebuilds - I am wondering about that too - I think we may have to do status messages or fedmsg watches
14:45:21 <bexelbie> I don't think you can see the build details on the CI without a login ... I'll need to find out
14:45:43 <randomuser> listening on fedmsg would be the most efficient, if the system has a queueing concept of some sort
14:45:51 <bexelbie> exactly
14:46:10 <bexelbie> I was thinking we might want something (email? irc pings?) when a new site is staged so humans know aobu it
14:46:12 <bexelbie> about it
14:46:40 <bexelbie> but in an idea world, a valid publication branch commit leads to an updated site with no human intervention (5 min stage refresh, 60 min prod refresh - by cron - not demand :( )
14:46:42 <randomuser> if CI can build PRs, i'd like to get signal about that from a comment on the PR
14:46:48 <bexelbie> def
14:47:05 * bexelbie would also like to eventually hook in rebuild of the release-notes rpm .. but that is def version 3.0 stuff
14:47:18 <randomuser> cron seems fine IMO for the time being
14:47:35 <bexelbie> that is an infra decisoin
14:47:41 <bexelbie> they are considering other optoins but aren't there yet
14:48:02 <randomuser> infra decides what triggers our builds?
14:48:14 <bexelbie> no
14:48:28 <bexelbie> they do the publishing to stage and prod for docs.fp.o
14:48:28 <randomuser> i know infra's cron job to pull build output..
14:48:34 <bexelbie> build triggers will be by fedmsg based on the repos
14:48:39 <randomuser> ah, i see
14:48:43 <bexelbie> and staging of prs is not an infra related issue if we use the artifacts server
14:49:52 <randomuser> very encouraged by your efforts, bexelbie, thanks for making time to keep us updated
14:49:54 <bexelbie> that is the end result of my brain dump so far
14:50:03 * bexelbie can also update on the RPM if we add that topic
14:50:11 <bexelbie> thank you all for your support :)
14:50:20 <randomuser> sure, but let's move through quickly so porfiriopaiz has time
14:50:26 <randomuser> #topic release notes RPM update
14:50:37 <randomuser> i have a feeling I should spend time on this one
14:51:14 <bexelbie> short version is this
14:51:20 <bexelbie> I worked with ignatenkobrain and we have an rpm
14:51:24 <bexelbie> it published from the release-notes repo
14:51:26 <bexelbie> it works
14:51:28 <bexelbie> we can remake
14:51:30 <bexelbie> it
14:51:38 <bexelbie> it has one bug and needs to be rebuilt with new content
14:51:55 <bexelbie> the bug: it needs to include bootstrap locally so a truly offline machine can see hte docs in a way that looks nice
14:52:05 <bexelbie> the rebuild: I will work on it this week - and document :)
14:52:23 <randomuser> I'll make an effort to be around on IRC more to help
14:52:24 <bexelbie> we have a package in the build, so I am not worried at this point
14:52:26 <bexelbie> eom
14:52:31 <bexelbie> +1 to help
14:52:33 <randomuser> ty, bexelbie
14:52:36 <randomuser> ignatenkobrain++
14:52:37 <zodbot> randomuser: Karma for ignatenkobrain changed to 21 (for the f26 release cycle):  https://badges.fedoraproject.org/tags/cookie/any
14:52:46 <randomuser> #topic open floor
14:52:53 <porfiriopaiz> !
14:52:56 <randomuser> porfiriopaiz, still around?
14:53:00 <porfiriopaiz> yeap!
14:53:04 <randomuser> welcome@
14:53:17 <porfiriopaiz> Hi everyone!
14:53:24 <bexelbie> Hi porfiriopaiz
14:53:33 <porfiriopaiz> Well I just wanted to introduce myself to the team.
14:53:59 <porfiriopaiz> Here is the mail I sent to the mailing list: https://lists.fedoraproject.org/archives/list/docs@lists.fedoraproject.org/thread/5EU3JHFRN7UAG5SHL273PONUDXTAHK6C/
14:54:24 <porfiriopaiz> I have some doubts about the workflow, what are you all working on and how I can help.
14:54:46 <randomuser> constructive feedback is always welcome :)
14:55:16 <porfiriopaiz> I already forked a project under pagure and tried to make some edits.
14:55:25 <bexelbie> which project porfiriopaiz ?
14:55:26 <randomuser> porfiriopaiz, I agree with you about Atomic, I've been spending a bunch of my lab time lately trying to learn how to best use it
14:55:34 <porfiriopaiz> bexelbie: give me a sec...
14:55:53 <porfiriopaiz> bexelbie: this one: https://pagure.io/fedora-docs/asciidocs-sysadmin
14:56:04 <bexelbie> that is actually not active
14:56:14 <bexelbie> https://pagure.io/fedora-docs/
14:56:17 <bexelbie> has the active ones
14:56:25 <bexelbie> we need to update those repos as inactive :(
14:56:45 <bexelbie> you want this one, I think: https://pagure.io/fedora-docs/system-administrators-guide
14:56:47 <randomuser> there's been lots of repo creation, I've honestly lost track
14:56:53 <porfiriopaiz> hahah, also I wanted to know what are the actives ones because I got a litle confused with this two https://pagure.io/projects/fedora-docs/* and https://pagure.io/fedora-docs/
14:57:10 <bexelbie> the one with the * lists all repos we have
14:57:16 <bexelbie> the other is a curated list of what we actually have active
14:57:33 <porfiriopaiz> The thing is that I found some formatting errors, tried to fix them but did not work, the errors are still.
14:57:53 <porfiriopaiz> This is the page from the docs I found the errors: https://docs.fedoraproject.org/f26/system-administrators-guide/basic-system-configuration/System_Locale_and_Keyboard_Configuration.html
14:58:37 <porfiriopaiz> bexelbie: for clarification, what is the repo active, I got lose, my bad.
14:58:42 <bexelbie> that would be on the f26 branch of the repo I mentioned
14:58:44 <randomuser> bexelbie, I can see how the two URLs con be confusing; https://pagure.io/projects/fedora-docs/* is the target when clicking "Fedora Docs" in context
14:58:46 * bexelbie gets url again
14:59:00 <randomuser> ...i think we broke pagure
14:59:03 * bexelbie is not getting responses from pagure
14:59:06 <bexelbie> ok, not just me :)
14:59:23 <bexelbie> there is a stagnated proposal to move all of the non-active repos into a new name space
14:59:33 <porfiriopaiz> phew, It is back :)
14:59:48 <bexelbie> it is blocked by three things, afaik: 1) agreement from all of us :) ; 2) a script to do the renames - pending but known tech ; 3) a name space decision
15:00:07 <randomuser> let's take that to the list, please
15:00:25 <randomuser> or revive it there, more likely
15:00:33 <bexelbie> porfiriopaiz, this is hte url for hte page you want to fix: https://pagure.io/fedora-docs/system-administrators-guide/blob/f26/f/en-US/basic-system-configuration/System_Locale_and_Keyboard_Configuration.adoc
15:00:39 <randomuser> gotta get moving before adamw beats me up
15:00:43 <bexelbie> it probably also needs to be done for f27 branch and master branch
15:00:57 * bexelbie will be in #docs during my next call
15:01:04 <randomuser> let's hop over to #fedora-docs to continue the conversation
15:01:13 <randomuser> thanks for coming, everyone
15:01:16 <randomuser> #endmeeting