17:04:09 <ahf> #startmeeting gitlab status meeting, 8 april 2020
17:04:09 <MeetBot> Meeting started Wed Apr  8 17:04:09 2020 UTC.  The chair is ahf. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:04:09 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
17:04:10 <gaba> pad: https://pad.riseup.net/p/gitlab-migration-problems
17:04:22 <gaba> anybody else for the gitlab meeting?
17:04:46 <ahf> i've added some *comments* on top of the list items from my friday experiments
17:05:16 <ahf> i gave up on the [[Image(filename, style)]] one, so images are only possible as attachments for now i think
17:05:26 <ahf> i could spend more time on it, but i am not sure it is worth getting a perfect score here
17:06:20 <gaba> yes, we can have a list of things that we are not doing now
17:06:29 <gaba> and see if anybody have a big objection on those
17:07:27 <ahf> yes. the markdown conversion problems are easy it seems, which is good. on friday my goal was to do all the metadata items we have
17:07:41 <ahf> like point conversion and the issue with spaces vs comma in keywords
17:07:50 <ahf> and i found even a semi-colon somewhere on friday
17:08:19 <gaba> ok
17:09:27 * ahf is just going over his notes from friday to see if he missed anything
17:09:28 <pili> hi
17:09:38 <pili> sorry, I spaced out working on websites :)
17:09:47 <gaba> so the list that is in the pad marked as 'no possible' are things we are leaving behind
17:10:22 <ahf> gaba: it is the things that i think wont be possible to do in GL, yeah
17:10:26 <ahf> pili: no worries
17:10:50 <ahf> ok, the bottom items are all metadata stuff, that is fine
17:10:50 <gaba> ahf: are you ok if I reorganize the list in the pad to have the no possible things in one place?
17:11:02 <ahf> gaba: yes! feel free to reorder things
17:11:19 <ahf> i think you have a better sense of order than i do
17:12:12 <gaba> ok
17:12:55 <ahf> i added some *solved* *impossible*, etc. on top of some of the todo items
17:13:06 <ahf> those i would like to keep around (but maybe in a different place) so i can group things
17:14:58 <gaba> not sure what is the stuff from line 24 to line 53
17:15:38 <ahf> looking
17:16:51 <ahf> the things to look for and things not to look for can be removed. i'm gonna do that
17:16:56 <ahf> that was when the pad was a survey
17:17:08 <gaba> ok
17:17:32 <gaba> pili: can you take a look at the things that we are not going to do and see if there is anything there you care about?
17:17:41 <pili> let me see...
17:17:46 <ahf> yep
17:19:01 <pili> "ISSUES NOT RESOLVED" right?
17:19:02 <pili> I can't see anything there that I care strongly about :)
17:19:16 <ahf> oki, a lot of it cosmetic things, fortunately
17:19:27 <pili> yup
17:19:46 <pili> are we going to keep a static (archived?) copy of trac?
17:19:56 <pili> or is it just going to disappear at some point?
17:20:09 <anarcat> wait, we're having a gitlab meeting now?
17:20:12 <anarcat> totally missed that
17:20:12 <pili> e.g because we need to decommission the machines that run trac currently
17:20:14 <anarcat> hi :)
17:20:22 <pili> perfect timing anarcat  ;)
17:20:22 <gaba> hi :)
17:20:29 <ahf> pili: that part i don't know the answer to, but i think anarcat might have an opinion about that
17:20:34 <anarcat> yes
17:20:36 <anarcat> i want to kill trac
17:20:38 <ahf> i think he did some archiving before of tpo stuff
17:20:44 <gaba> anarcat just exactly gets in in the controversial topic :P
17:20:54 <anarcat> o/
17:21:01 * pili wonders which of the words she mentioned anarcat has highlights for ;)
17:21:02 * anarcat puts on his BOFH hat
17:21:10 <anarcat> hehe
17:21:14 <gaba> "archive trac" pili
17:21:18 <pili> :D
17:21:20 <anarcat> no highlight, i was just clearing my activity list in irssi :)
17:21:30 <pili> yeah yeah... ;)
17:21:33 <anarcat> this is the plan for trac https://trac.torproject.org/projects/tor/ticket/31690
17:21:54 <pili> ok, I will read that, thank you :)
17:22:09 <gaba> nice
17:22:09 <anarcat> we have a crawl of trac up until #30856 into the internet archive, that's about ~3-4k tickets missing
17:22:16 <anarcat> we could do another run of this
17:22:18 <ahf> sweet
17:22:36 <ahf> the archive is on the internet achive then, so browsing it will be via their site?
17:22:46 <anarcat> ahf: there are a few proposals
17:22:59 <ahf> my trac is loading
17:23:02 <gaba> i would love to leave trac archived somewhere...
17:23:31 <anarcat> the tickets mentions 4 approaches: "1. the golden redirect", "2. readonly trac", "3. fossilization", and "4. destruction"
17:23:45 <ahf> (2) and (3) are the same, but (3) allows us to turn off trac, right?
17:23:50 <anarcat> one option i have *not* put there is option "0. keep trac around forever" which i strongly object to
17:23:55 <ahf> like for the user it is the same
17:24:00 <gaba> fossilization is converting to static only i guess
17:24:03 <anarcat> ahf: effectively yes
17:24:11 <ahf> ack
17:24:16 <anarcat> i would also object strongly to option 2
17:24:26 <gaba> yes, i like 4 :)
17:24:29 <gaba> i mean 3
17:24:41 <anarcat> well option 4 is a little intense, we should probably do at least 1
17:24:42 <ahf> i like 3 for a period and then maybe 4
17:24:52 <ahf> but 4 is not a priority for me :-S
17:24:56 <anarcat> heh
17:24:56 <gaba> yes, but first we need to get to the other side that is this migration
17:25:06 <gaba> next steps
17:25:06 <anarcat> well
17:25:17 <anarcat> i would really hate to be the sucker stuck with maintaining trac at the end of this :p
17:25:35 <gaba> yes, that is why we need to get to the other side of a migration...
17:26:00 <gaba> ahf: is there any other test that needs to be done or we are ready to have a date for migration if we agree on a plan?
17:26:07 <anarcat> but it also means that the migration needs to do certain things for us to keep our options open for trac
17:26:12 <anarcat> e.g. we need to migrate the wiki somehow
17:26:24 <gaba> yes, ahf has been working on that
17:26:26 <ahf> gaba: tests?
17:26:34 <anarcat> gaba: awesome, thanks
17:26:34 <gaba> anarcat: https://pad.riseup.net/p/gitlab-migration-problems
17:26:41 <ahf> we need to go over the list of items when all itmes are marked as done, and see if it's true
17:26:42 <anarcat> yeah, i still need to catchup with my tabs :)
17:26:46 <ahf> or if i have missed something there
17:26:53 <gaba> ahf: you have been writing the migration script and running it
17:27:02 <pili> anarcat: gitlab projects have wikis so project maintainers should migrate the content if we can't do it automatically (I guess...)
17:27:03 <pili> without having followed the plan too closely :S
17:27:10 <ahf> gaba: ah, yep
17:27:24 <ahf> gaba: i think the list there is we need to carefully go over the list and see if something is actually not fixed
17:27:30 <gaba> ok
17:27:40 <gaba> you did a last migration, where is the link?
17:27:42 <ahf> and i think we hsould do that once the metadata changes are in, because they are where we have information loss
17:27:45 <anarcat> pili: there is "only one wiki" in trac, so my hope is that we migrate at least that into the legacy project
17:27:54 <pili> ah, ok
17:28:11 <gaba> yes, we need to migrate everything into legacy project and then from there to the last place in gitlab
17:28:24 <anarcat> then of course, like the legacy project issues, teams can move stuff out of there, either by forking the wiki (it's a git repo!) or by copying content by hand
17:28:48 <ahf> last run i have is on https://gitlab.torproject.org/ahf-admin/legacy-20200320 - i plan to start a run with the metadata changes over the upcoming weekend so we can walk over those
17:29:17 <gaba> ahf: so we should wait for your next run then?
17:29:24 <anarcat> anyways, sorry if i hijacked your meeting :)
17:29:26 <ahf> i think so, yes. right now the things are only cosmetic
17:29:29 <ahf> anarcat: you did not!
17:29:29 <anarcat> i hope i answered the question
17:29:56 <gaba> so for what I understand next steps would be:
17:29:59 <gaba> 0. ahf does one more run of the migration into a legacy test project.
17:29:59 <gaba> 1. check that all issues marked as resolved are fine in the legacy test project.
17:30:02 <gaba> 2. check with tor-project@ that everybody is ok with stuff that is not resolved.
17:30:05 <gaba> 3. review plan https://nc.torproject.net/s/3MpFApQ7cwfrPZE
17:30:07 <gaba> 4. migrate
17:30:10 <gaba> 5. decide on trac's next life https://trac.torproject.org/projects/tor/ticket/31690
17:30:26 <ahf> yes, (1) have an "if there are issue, goto (0) and repeat
17:30:29 <ahf> "
17:30:50 <anarcat> gaba: 3 and 5 seem backwards to me, i feel they should be before 1 :)
17:31:01 <anarcat> ie. make sure our plan is sound, and include trac in there
17:31:07 <gaba> yes
17:31:15 <anarcat> that would also expand point 4 considerably, i think :)
17:31:18 <gaba> ok. I'm ok to whatever is decided on trac's after life
17:31:42 <gaba> ahf: yes, 0-1 will repeat until we are fine
17:31:55 <ahf> yes
17:31:57 <gaba> for 1 I would guess is you and me and maybe pili looking at it?
17:32:07 <ahf> the more eyes the better to be honest
17:32:16 <ahf> once we are happy i would really like to ask dcf to look too
17:32:22 <gaba> and that could be next week then. You tell  us when you have it ready
17:32:23 <pili> ok
17:32:24 <ahf> he has spotted things i would have missed many times
17:32:27 <anarcat> 1.b) if all is good, go to 2; 1.c) goto 0
17:32:34 <anarcat> this feels like BASIC ;)
17:32:36 <gaba> :)
17:32:39 <ahf> gaba: yep, if i start a sync friday evening my time it should have a lot of tickets like monday evening
17:32:49 <gaba> sounds good
17:33:00 <gaba> can we review 3 in this meeting to be sure we are ok?
17:33:09 <ahf> should we start having a weekly status in here at this time each week ideally again?
17:33:13 <ahf> yep
17:33:13 <anarcat> gaba: i'm game
17:33:16 <anarcat> i just lost the game
17:33:50 <gaba> weekly status seems good to me
17:34:34 <pili> ok
17:34:43 <ahf> ool
17:34:46 <ahf> cool*
17:35:55 <gaba> the nc file has at the bottom something that says process
17:36:08 <gaba> I went through the file and I updated it based on conversations we had before
17:36:47 <ahf> cool
17:37:30 <anarcat> so something that's missing in that process is the redirect
17:37:38 <anarcat> i mean
17:37:52 <anarcat> like when we put trac readonly
17:37:58 <ahf> and the irc bot is also missing
17:38:00 <anarcat> because we will necessarily have to do that at some point
17:38:01 <anarcat> yeah
17:38:50 <gaba> anarcat, ahf: can you add those?
17:39:38 <ahf> added to the list on the pad now
17:39:39 <anarcat> not sure what's going on with that list numbering
17:39:41 <gaba> if we are migrating project by project from trac we will have to archive the project in trac right after we do the migration into legacy
17:40:34 <anarcat> yeah, that makes sense
17:40:52 <ahf> i think the redirect is easier for http than the irc bot is. because of the legacy project
17:40:53 <ahf> yeah
17:41:30 <anarcat> the bot only needs to be extended to support foo/bar#N to talk to gitlab
17:41:47 <anarcat> with #N being a special case with tpo/legacy#N being implicit
17:41:49 <anarcat> "only"
17:41:53 <anarcat> not sure where that fits in the doc
17:42:41 <gaba> is this something after we do the migration? im guessing the bot will be down a little bit
17:42:52 <anarcat> we should do it before, IMHO
17:42:58 <anarcat> because we can
17:43:04 <pili> the bot is really useful
17:43:06 <ahf> we can live without the bot a little while, but i think we should do it before if possible. i think the bot is a supybot?
17:43:30 <gaba> ok
17:43:41 <anarcat> ahf: i have *no* idea - maybe?
17:43:53 <anarcat> one problem is it's a multi-stakeholder bot
17:44:00 <anarcat> iirc, it's also used by the debian project?
17:44:01 <gaba> who could extend the bot?
17:44:08 <anarcat> i am not sure
17:44:21 <anarcat> we're talking about zwiebelbot right?
17:44:26 <ahf> asked weasel now
17:44:27 <ahf> yeah
17:44:27 <gaba> yes
17:44:39 <anarcat> 13:45:23 [OFTC] -!-  ircname  : zwiebelbot run by weasel (@debian.org)
17:44:59 <anarcat> 13:46:02 <zwiebelbot> The current (running) version of this Limnoria is 2019.02.23, running on Python 3.7.3 (default, Dec 20 2019, 18:57:59)  [GCC 8.3.0].  The newest versions available
17:45:15 <gaba> ok, we shoudl check with weasel to see if he can do this
17:45:35 <ahf> oh
17:45:36 <anarcat> ahf: it is supybot, but the question really is more (1) where is it running and (2) who has access to that and (3) who wants to hack it
17:45:54 <ahf> weasel once showed me how it was doing the bug sthing and it looked very easy
17:46:02 <ahf> like it can do debian#123 now
17:46:05 <ahf> i think
17:46:08 <ahf> maybe not
17:46:18 <ahf> looks like not
17:46:18 <anarcat> it can
17:46:23 <anarcat> i suspect it's channel-specific
17:46:28 <ahf> ah, maybe
17:46:34 <anarcat> because in debian channels, #N points to bugs.debian.org/N
17:46:40 <anarcat> and it's the same zwiebelbot
17:46:54 <anarcat> 13:47:36 <+weasel> anarcat: it's mine.  it runs on my stuff
17:47:04 <anarcat> 13:47:48 <+weasel> ahf: https://github.com/weaselp/ticketbot/ is the module
17:47:41 <ahf> perfect
17:48:08 <anarcat> okay, so (3)?
17:48:09 <anarcat> :)
17:49:13 <ahf> sounds like it
17:49:52 <anarcat> who wants to hack at this
17:50:26 <anarcat> and when
17:50:29 <anarcat> and where do we fit that in the plan
17:51:10 <anarcat> *crickets*
17:51:12 <ahf> i want to, but i want to be done with the other stuff first
17:51:14 <anarcat> such enthousiasm
17:51:29 <anarcat> i usually like playing with irc bots, but in this case i'm a little hesitant
17:51:29 <ahf> it is not the highest priority right now i think, so we don't need to decide on that now
17:51:41 <gaba> yes, not the highest priority
17:51:49 <gaba> let's check to see if weasel wants to do it
17:52:36 * gaba brb
17:53:18 <ahf> okay, is this something we can look at over the next week and then finish up next week?
17:53:21 <ahf> we are near the hour
17:54:05 <anarcat> i can talk with weasel i guess
17:54:10 <anarcat> i am talking with weasel, even
17:54:51 <ahf> cool
17:54:59 <ahf> any objections to me ending the meeting or?
17:55:28 <pili> nope :)
17:55:35 <anarcat> k
17:55:42 <anarcat> less talk more rock
17:55:43 <anarcat> ;)
17:56:10 <gaba> sounds good
17:57:33 <gaba> thanks!
17:57:43 <ahf> #endmeeting