16:30:06 <djmitche> #startmeeting weekly
16:30:06 <bb-supy`> Meeting started Tue Jan 26 16:30:06 2016 UTC and is due to finish in 60 minutes.  The chair is djmitche. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:30:06 <bb-supy`> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:30:06 <bb-supy`> The meeting name has been set to 'weekly'
16:30:14 <djmitche> #topic Introduction
16:30:26 <djmitche> #link https://titanpad.com/buildbot-agenda
16:30:27 <infobob> http://paste.pound-python.org/show/ElavjDMKHWGJriJQlkib/ (repasted for djmitche)
16:30:31 <djmitche> thanks infobob :
16:30:51 <djmitche> #topic Old buildbot-devel mailing list
16:30:59 <djmitche> sa2ajj: ^^ this may be of interest to you
16:31:18 <djmitche> #info We've gotten a few subscription attempts to the old list, as it's still in the Google search results
16:31:33 <djmitche> I'd like to delete the list, so Google will remove it, and so nobody gets confused
16:31:44 <djmitche> I emailed the two who tried to subscribe, but neither has re-tried on the new mailing lists, that I've seen
16:32:18 <djmitche> Any opinions pro/con?
16:32:29 <bdbaddog> can you migrate the contents to the new mailing list? (archives?)
16:32:53 <djmitche> I think we already did that (checking)
16:33:21 <djmitche> hm, maybe not
16:33:39 <djmitche> oh, right, gmane and all of those have just re-subscribed
16:33:47 <djmitche> the SF archives are useless
16:35:14 <djmitche> e.g., http://blog.gmane.org/gmane.comp.python.buildbot.devel/month=20150701
16:36:01 <djmitche> #agree will delete the old list to avoid confusion and remove it from the google search results
16:36:10 * djmitche feeling lonely here
16:36:22 <djmitche> #topic new worker on VM infrastructure
16:36:53 <djmitche> about a half-hour ago, verm__ was mentioning he's working on setting up a debian install on vm1.buildbot.net, via bhyve
16:37:05 <bdbaddog> My bad on this. I meant to fire up the docker slaves but life got in the way..
16:37:14 <djmitche> oh, that's OK
16:37:20 <rutsky> hi all!
16:37:43 <djmitche> I'm mostly keen to get it running on BB infra so that we can be confident it will keep going
16:37:51 <bdbaddog> sure gotcha.
16:37:54 <djmitche> not that your infra is necessarily flaky, just in general :)
16:38:00 * djmitche waves to rutsky
16:38:37 <djmitche> #info running postgres in Docker on FreeBSD turned out to be a dead end (for now - missing syscall emulation)
16:38:42 <djmitche> #info setting up a Debian VM instead
16:38:56 <djmitche> #topic cli tool to talk to API - demo
16:39:15 <djmitche> I think Amar is busy today, so no demo -- but hopefully we'll hear something about this project soon!
16:39:18 <rutsky> where is the sources for that tool?
16:39:26 <djmitche> I don't think it's been open-sourced yet
16:39:33 <rutsky> because I've changed API in this weekend in my branch...
16:39:49 <rutsky> ok
16:40:17 <djmitche> From Jan 5:
16:40:18 <djmitche> 16:52:07 <verm__> i'm going to finish the minimal layout to get info like builders / slave info etc then put it into a repo
16:40:32 <djmitche> https://supybot.buildbot.net/meetings/buildbot/2016/buildbot.2016-01-05-16.32.log.html#l-54
16:40:35 <rutsky> will it be python cmdline util?
16:40:40 <djmitche> I think so, yes
16:40:58 <djmitche> but without verm__ I think that's all we know :)
16:41:04 <rutsky> ok, then I don't think there will be problems with porting to renamed API
16:41:16 <djmitche> no, it should be fine
16:41:29 <djmitche> and to my knowledge, not released yet, so no compatibility issues in just search/replacing
16:41:29 <djmitche> #topic Bug 2340
16:41:33 <djmitche> speaking of which :)
16:41:43 <rutsky> a lot of work is done :)
16:41:44 <djmitche> rutsky: so what's new?
16:41:46 <djmitche> yeah :)
16:41:54 <rutsky> and not many remaining
16:42:10 <rutsky> I've updated most of public API during this weekend
16:42:15 <rutsky> created migrations
16:42:21 <rutsky> ported Data API
16:42:27 <rutsky> updated web UI
16:42:32 <verm__> no time to work on it :(  i got stuck on the issue of being able to listen to builds realtime i'll have more time next week
16:42:36 <rutsky> (both base and md_base)
16:42:46 <djmitche> verm__: ok, cool, we'll keep checking in
16:42:53 <verm__> i'm working on finishing the base work once that's done i'll put it up
16:42:58 <djmitche> cool
16:43:01 <verm__> eg, commandline processing, layout, --help etc
16:43:05 <rutsky> I need to update docs + a lot of minor renamings (mostly in tests)
16:43:17 <djmitche> rutsky: what do the migrations change? table names or just column names?
16:43:25 <rutsky> Review of my changes can be started
16:43:35 <rutsky> tables, column names, indices
16:43:46 <rutsky> it's in the docs
16:43:58 <rutsky> https://github.com/rutsky/buildbot/blob/worker_API_rename_POC2/master/docs/manual/worker-transition.rst#database
16:44:07 <rutsky> comlete list of changed DB entities
16:45:06 <rutsky> while renaming is not finished for all stuff, most of API (like 98%) is ported --- PR review can be started by reviewing high-level API changes
16:45:25 <rutsky> I tries to document most of significant renamings in that RST doc file
16:45:38 <rutsky> so I'm asking for it's review
16:45:43 <djmitche> you are incredibly thorough :)
16:45:53 <rutsky> thanks)
16:46:15 <rutsky> this is the reason of 500+ commits and almost month of work)
16:46:28 <djmitche> #action everyone review https://github.com/rutsky/buildbot/blob/worker_API_rename_POC2/master/docs/manual/worker-transition.rst which summarizes changes
16:46:39 <rutsky> plus a lot of tricky issues arised in master code...
16:46:52 <djmitche> #info DB API, Data API, Web UI largely ported
16:46:52 <tardyp> the thing is that github will even not allow to review such big PR :)
16:46:54 <djmitche> I bet
16:47:03 <rutsky> tardyp: yeah(
16:47:11 <djmitche> tardyp: yeah, we may have to do it commit-by-commit
16:47:14 <djmitche> that's OK though
16:47:20 <tardyp> can you upload a PR with just the doc update?
16:47:30 <djmitche> ++ good idea
16:47:33 <tardyp> I dont like too much reviewing inline in commits
16:47:34 <rutsky> tardyp: you mean BB documentation?
16:47:34 <djmitche> then we can comment in the PR
16:47:54 <djmitche> rutsky: a PR with just the worker-transition.rst file
16:47:57 <rutsky> tardyp: I summarized changes in that singe worker-transition.rst file
16:48:43 <rutsky> My plan is to get review for high-level API changes (described in worker-transition.rst), then prepare docs of how fallback is implemented
16:48:54 <djmitche> I think he's looking for a PR that Github will allow comments on, so one with just the .rst file -- then we can comment on it in that PR
16:48:59 <rutsky> then provide docs how these changes are tested
16:49:06 <rutsky> oh
16:49:15 <rutsky> yes, I think comments should go on whole PR
16:49:32 <rutsky> I thing you can comment line by line that rst file...
16:50:12 <rutsky> or not? Github doesn't seem to show rst file in "sources" mode, as it does for source files...
16:50:25 <tardyp> you have an option for that
16:50:36 <tardyp> at least it work when you do a PR
16:50:54 <tardyp> what we ask is that you create a new branch with just the change of the doc
16:51:04 <tardyp> you merge this branch first to master
16:51:16 <rutsky> tardyp: can do
16:51:22 <djmitche> yeah I don't see the .rst file in https://github.com/buildbot/buildbot/pull/1944 (it's in there, github just doesn't show it)
16:51:24 <djmitche> ok :)
16:51:43 <tardyp> 1944 is so big that github refuses to show all files
16:51:54 <djmitche> #action rutsky will provide a PR with just the worker-transition.rst file for easier review (since github doesn't show the whole diff of the big PR)
16:52:03 <rutsky> so decided --- I'll create separate PR with only worker-transition.rst (plus links on it from other rst, to make sphinx happy)
16:52:07 <djmitche> also hangs my browser for a few second :)
16:52:09 <djmitche> cool
16:52:24 <djmitche> anything else for that project?
16:52:34 <rutsky> any ideas of how you would like to review this PR?
16:52:39 <djmitche> #action djmitche to contact SFC again regarding MOSS status
16:53:07 <rutsky> probably side projects in which you would like to test these changes?
16:53:10 <bdbaddog> Should we also ping Gerv about their side of the MOSS grant?
16:53:13 <rutsky> tardyp: e.g. buildbot-travis?
16:54:21 <tardyp> rutsky: yes, My plan is to take you code, and run the diff manually to make myself confortable with the code
16:54:39 <tardyp> this will include some manual tests with my usual test environment
16:54:40 <djmitche> bdbaddog: not a bad idea -- do you want to do that or should I?
16:54:46 <rutsky> djmitche, tardyp, maybe setup web ui somewhere to see UI changes in actions?
16:54:56 <djmitche> ++ good idea
16:55:00 <bdbaddog> sure. I'll follow up. we've not heard from Gerv in a bit.
16:55:33 <rutsky> bdbaddog, djmitche, who is Gerv?..
16:56:07 <bdbaddog> Gerv is the guy at Mozilla running the MOSS grant project.
16:56:25 <bdbaddog> Though it seems like they decided to announce doing it, then do it, then figure out if and how they can do it.
16:56:33 <bdbaddog> (in that order)..
16:56:46 <rutsky> heh)
16:56:52 <tardyp> mmhh
16:57:11 <bdbaddog> thus the whole snafu on them funding the grants they've awarded. They're checking with their accounting firm (I think)
16:57:30 <djmitche> #action bdbaddog to check in with Gerv at Mozilla
16:57:39 <bdbaddog> (done.. ;)
16:57:44 <djmitche> rock on :
16:57:50 <djmitche> and, yeah, that's pretty typical
16:58:03 <djmitche> remember when Moz and EFF announced Let's Encrypt?  18 months ago?
16:58:07 <rutsky> regarding MOSS grant, will there be contract or something like that?
16:58:10 <djmitche> and then 16 months later they actually built it?/
16:58:23 <djmitche> rutsky: there will be one between Mozilla and Conservancy
16:58:38 <djmitche> and once that's in place and the funds have been transferred, we can sign one with you
16:59:08 <rutsky> oh, ok. Meanwhile I will do some really useful work :)
16:59:25 <bdbaddog> +1
16:59:32 <djmitche> yeah, I'm personally committed to making good on the bounty, but there's still a chance for everything to fall apart
16:59:39 <djmitche> ok
16:59:45 <djmitche> #topic Development week-in-review
16:59:55 <djmitche> #info we released 0.9.beta6!!
16:59:58 <rutsky> djmitche, tardyp: what about setting up server with UI for renaming changes?
17:00:31 <djmitche> we could deploy it to nine.buildbot.net?
17:01:05 <rutsky> Can we do so? --- Data API has been changed
17:01:09 <djmitche> we could just switch that to apply from the bug2340 branch instead
17:01:31 <djmitche> I think we'd still have to run the DB migrations by hand
17:01:38 <rutsky> is nine.buildbot.net is completely separate from buildbot.buildbot.net?
17:01:41 <djmitche> yes
17:02:00 <rutsky> oh, ok, then whole nine.buildbot.net can be migrated to that branch
17:02:12 <rutsky> whould we create branch at main repo?
17:02:16 <djmitche> yes
17:02:21 <djmitche> I think you have permissions to do that?
17:02:28 <rutsky> I think I'm not...
17:02:39 <rutsky> I'm not in Buildbot GitHub organisation
17:02:48 <tardyp> rutsky: did you change data_module in the UI?
17:02:53 <rutsky> would love to be in it, btw :)
17:02:59 <rutsky> tardyp: yes
17:03:11 <tardyp> then it is a bit problematic
17:03:26 <tardyp> as we have to publish the data module in order for nine.bb.net to work
17:03:27 <rutsky> tardyp: data_module, base, md_base (plugins, such as waterfall_view doesn't use "slave" stuff)
17:03:27 <djmitche> rutsky: you're added now to that repo
17:03:48 <rutsky> djmitche: thanks! :)
17:04:07 <tardyp> can we make first a data module version which would work for both before/after
17:04:26 <tardyp> I think it is just about adding one endpoint (instead of replacing one)
17:04:52 <rutsky> hm...
17:04:59 <rutsky> also object names were changed
17:05:07 <djmitche> #action djmitche to change nine.bb.net to pull from the bug2340 branch, perform DB migration, once that is ready
17:05:10 <rutsky> slavename -> workername in object
17:05:28 <rutsky> and keys that points to collections in result of REST API calls
17:05:47 <tardyp> mhh. actually, we can make a major version change to make sure that b6 installs wont pull this version
17:06:08 <bb-trac> [trac] #3420/undecided (assigned) updated by dustin (empty comment) http://trac.buildbot.net/ticket/3420
17:06:14 <bb-trac> [trac] #3419/undecided (assigned) updated by dustin (empty comment) http://trac.buildbot.net/ticket/3419
17:06:26 <rutsky> tardyp: v2 -> v3?
17:06:32 <djmitche> ++!
17:06:38 <djmitche> yay semver
17:06:51 <rutsky> tardyp: how that will help to run both version simultaneously?
17:06:52 <tardyp> I was not talking about v3 in the endpoint
17:07:22 <tardyp> I was talking about version of the datamodule
17:07:24 * tardyp afk
17:08:15 <rutsky> tardyp: good point, djmitche, I think action for me to bump data_module version is needed :)
17:08:43 <djmitche> #action rutsky to bump data_module major version so that it can be published to npm without earlier betas automatically updating to it
17:09:07 <rutsky> however I still don't understand what we will do with supporting old/new version currently at nine.bb.net?
17:09:29 <rutsky> tardyp: or we don't want that?
17:09:55 <rutsky> djmitche: I will name branch "bug2340"?
17:10:00 <djmitche> We can update that service to the bug2340 branch, which will have data_module: "^3.0.0" in its package.json
17:10:02 <djmitche> sure
17:10:42 <rutsky> so current PR will go into bug2340 branch, and worker-transition.rst will go in separate PR against master?
17:10:47 <rutsky> (only for review)
17:12:15 <rutsky> djmitche ?
17:13:33 <djmitche> sounds good
17:13:45 <djmitche> nobody depends on nine.bb.net, so we can mess around until we get it right :)
17:13:54 <rutsky> ok
17:14:02 <djmitche> so! development week in review :)
17:14:21 <djmitche> I see two bugs about DB compatibility -- do you think those are specific to those versions of teh DB, or a regression in teh code?
17:14:49 <rutsky> or Docker misconfiguration... :)
17:15:07 <djmitche> yeah
17:15:10 <djmitche> I'll look at them
17:15:21 * rutsky haven't investigated issues, just reported what have seen during migration testing
17:15:24 <djmitche> tardyp: I feel like the only major blocker to shipping 0.9.0 is the sidebar not collapsing
17:15:37 <djmitche> (#3418)
17:15:46 <rutsky> minor: issue http://trac.buildbot.net/ticket/3356 was kind of reopened
17:15:51 <bdbaddog> what about force scheduler?
17:16:08 <bdbaddog> for many of my clients using force scheduler is part of their workflow.
17:16:13 <rutsky> +1 for forcescheduler issue...
17:16:17 <tardyp> djmitche: yes, I've seen this issue
17:16:37 <tardyp> its more a problem of forcheduler ui, which should work with pinned sidebar
17:16:39 <bb-trac> [trac] #3356/defect (reopened) updated by dustin (empty comment) http://trac.buildbot.net/ticket/3356
17:16:40 <rutsky> btw "force" button works, so it's enough for testing ;)
17:17:04 <djmitche> bdbaddog: yes, #3416 is important too
17:17:13 <tardyp> there is this bug also with force about the force scheduler throwing errors
17:17:28 <tardyp> it works, but there is a console error I should investigate
17:17:36 <bb-trac> [trac] #3416/defect (assigned) updated by dustin (empty comment) http://trac.buildbot.net/ticket/3416
17:17:42 <tardyp> because probably the UI is not correct
17:17:53 <djmitche> bdbaddog: were there other issues you're seeing?
17:17:55 <rutsky> #3356 requires minor change that I asked to do in PR: @jhuold: https://github.com/buildbot/buildbot/pull/1957
17:18:11 <bdbaddog> I couldn't get very far with the broken ui..
17:18:28 <bdbaddog> I was trying to do a trial ugprade from scons.buildbot.org
17:18:45 <bdbaddog> with an eye toward upgrading the production.. but alas.. no can do at this point.
17:19:04 <bdbaddog> Is there any selenium or equivalent testing on the new ui?
17:19:56 <bb-trac> [trac] #3356/defect (assigned) updated by dustin (@rutsky, it sounds like there's a little more needed here?  debug.glade can be ...) http://trac.buildbot.net/ticket/3356
17:20:09 <djmitche> no selenium, no
17:20:17 <tardyp> bdbaddog: I think this is a good gsoc project
17:20:44 <bdbaddog> so the ui has no current tests?
17:20:51 <bb-trac> [trac] #3356/defect (assigned) updated by dustin (Oh I see, work is in progress on PR 1957.) http://trac.buildbot.net/ticket/3356
17:21:17 <tardyp> bdbaddog: the ui has unit tests
17:21:29 <tardyp> using karma
17:21:43 <bb-trac> [trac] #3421/undecided (assigned) updated by dustin (empty comment) http://trac.buildbot.net/ticket/3421
17:21:43 <tardyp> but there are a lot of things very hard to test without selenium
17:21:55 <tardyp> and e2e tests are quite complicated to setup
17:22:02 <djmitche> indeed :)
17:22:11 <bdbaddog> o.k.
17:22:23 <rutsky> djmitche: why is http://trac.buildbot.net/ticket/3423 is not in weekly tickets report?
17:22:31 <djmitche> #info still some issues around UI and force scheduler
17:22:57 <djmitche> rutsky: good question..
17:22:57 <bdbaddog> Is there any wiki page to help a dev ramp up on the ui code for nine?
17:23:08 <djmitche> bdbaddog: the docs are pretty good
17:23:20 <djmitche> http://buildbot.readthedocs.org/en/latest/developer/www.html
17:23:27 <bdbaddog> I'd look at that pinning issue, but I was drinking from the wrong end of the firehose just jumping in.
17:23:34 <rutsky> djmitche: README in repo also quite useful
17:23:38 <djmitche> haha
17:24:09 <djmitche> #info we are currently in a code freeze for #2340, too
17:24:27 <djmitche> #info probably one more beta after that, then 0.9.0 (no promises!)
17:24:40 <djmitche> ok, I need to get going -- anything else for the meeting notes?/
17:25:14 <rutsky> from my side none
17:25:16 <tardyp> djmitche: this is not the correct link
17:25:26 <bdbaddog> that's it for me.  I forwarded to botherders response from Gerv.
17:25:31 <bdbaddog> basically still waiting.
17:25:38 <bdbaddog> for paperwork issues
17:25:47 <djmitche> rutsky: I think that ticket was missed because of https://github.com/buildbot/buildbot-infra/blob/master/files/weekly-summary.py#L163
17:25:49 <tardyp> http://docs.buildbot.net/latest/developer/www-server.html
17:25:54 <djmitche> bdbaddog: ok, thanks
17:26:05 <tardyp> I've split the doc into several, but readthedocs wasn't updated
17:26:11 <djmitche> yes, that's the better link - thanks
17:26:14 <rutsky> djmitche: oh)
17:27:00 <bb-trac> [trac] #3424/enhancement (v:) created by dustin (weekly summary should include tickets from today, too) http://trac.buildbot.net/ticket/3424
17:27:08 <djmitche> well, I ended this meeting with a lot more bugs than I started :)
17:27:12 <djmitche> #endmeeting