All times shown according to UTC.

Time Nick Message
00:12 brson joined #buildbot
00:22 bb-github joined #buildbot
00:22 bb-github [13buildbot] 15vadz opened pull request #1391: Miscellaneous fixes for try subcommand. (06master...06try-fixes) 02
00:22 bb-github left #buildbot
00:30 bb-github joined #buildbot
00:30 bb-github [13buildbot] 15vadz comment on issue #1391: Sorry, I had somehow branched off a wrong place initially. Rebased on the latest master now. 02
00:30 bb-github left #buildbot
00:35 bb-trac [trac] #3070/support-request (new) updated by VZ (Submitted [ PR].    I use Cygwin daily ...)
00:44 ewong left #buildbot
00:45 ewong joined #buildbot
00:46 KeiKun joined #buildbot
01:11 bb-github joined #buildbot
01:11 bb-github [13buildbot] 15djmitche comment on pull request #1391 14f5aaaac: Won't this be problematic on Windows if SSH is at, say: `C:\Program Files (x86)\WinSSH\ssh.exe`? 02
01:11 bb-github left #buildbot
01:12 bb-github joined #buildbot
01:12 bb-github [13buildbot] 15djmitche comment on issue #1391: Looks good.  Needs the Travis failures fixed, though, and a release note added. 02
01:12 bb-github left #buildbot
02:01 bb-trac [trac] #3002/task (accepted) updated by verm (Replying to [comment:15 dustin]:  > I've done split horizon DNS before ("views" in ...)
02:26 ckang joined #buildbot
02:50 bb-trac [trac] #3002/task (accepted) updated by verm (I setup NSD with our zone, converted to the abbreviated format.    I disabled the ...)
03:41 booo joined #buildbot
03:43 skelly sa2ajj_brb: created as #21
03:43 bb-github joined #buildbot
03:43 bb-github [13buildbot-infra] 15seankelly opened pull request #21: [DO NOT MERGE] Jail role (06master...06jail-role) 02
03:43 bb-github left #buildbot
03:59 bb-github joined #buildbot
03:59 bb-github [13buildbot-infra] 15verm comment on pull request #21 14f3c4237: OPTIONS_UNSET=DOCS EXAMPLES... 02
03:59 bb-github left #buildbot
04:07 bb-trac [trac] #3002/task (accepted) updated by dustin (Mikhail already committed some Ansible code for this - ...)
04:11 bb-github joined #buildbot
04:11 bb-github [13buildbot-infra] 15seankelly comment on pull request #21 1479ae8dc: Ah yes, OPTIONS. I've seen stuff about that in UPDATING, but I forgot about it otherwise. Thanks for noting what needed to be changed. 02
04:11 bb-github left #buildbot
04:15 DOS622 joined #buildbot
04:16 bb-trac [trac] #3002/task (accepted) updated by verm (Replying to [comment:30 dustin]:  > Mikhail already committed some Ansible code for ...)
04:44 KeiKun_ joined #buildbot
06:22 bb-github joined #buildbot
06:22 bb-github [13buildbot] 15sa2ajj comment on issue #1378: @tardyp you also need to rebase the branch: it's currently unmergeable. 02
06:22 bb-github left #buildbot
06:31 jaredgrubb joined #buildbot
06:46 ewong I'm using an old version of buildbot (0.82 I believe).   there's this win32 job that's been pending for the past five hours... I do have Win32 idle..  can someone point out what I need to do to unhork that?  I've looked at the buildrequests table and it is unclaimed.
06:49 err 0.8.4
06:54 ah nvm..  just needed to set the complete=1 to previous horked builds
07:06 bb-github joined #buildbot
07:06 bb-github [13buildbot] 15sa2ajj closed pull request #1386: Fix BuildProperty row name (set singular) (06master...06patch-3) 02
07:06 bb-github left #buildbot
07:12 ewong nice..  totally unhorked that pending job
07:38 lissyx joined #buildbot
08:01 bb-github joined #buildbot
08:01 bb-github [13buildbot] 15benallard comment on issue #1387: Rebased + doc fixed. 02
08:01 bb-github left #buildbot
08:04 lissyx joined #buildbot
08:07 bb-trac [trac] #2851/task (accepted) updated by sa2ajj (empty comment)
08:07 bb-github joined #buildbot
08:07 bb-github [13buildbot] 15benallard comment on pull request #1391 14f5aaaac: **If** that is a concern, python has the ``shlex.split`` function. ([…]/shlex.html#shlex.split) 02
08:07 bb-github left #buildbot
08:08 bb-trac [trac] #3011/task (new) updated by sa2ajj (Amar, do you have the installation procedure anywhere documented/logged?)
08:09 bb-github joined #buildbot
08:09 bb-github [13buildbot] 15tardyp comment on issue #1378: rebased 02
08:09 bb-github left #buildbot
08:11 bb-trac [trac] #3069/task (new) updated by sa2ajj (I downgrade the priority for this bridge to cross.)
08:12 bb-github joined #buildbot
08:12 bb-github [13buildbot] 15tardyp comment on pull request #1387 14ce0c8f5: I dont understand the XXX, not sure if you want to keep it. 02
08:12 bb-github left #buildbot
08:15 bb-github joined #buildbot
08:15 bb-github [13buildbot] 15benallard comment on pull request #1387 14ce0c8f5: Oh, sure ! This aims at pointing out that this data_properies stuff cross references db_builds (instead of db_properties which doesn't exists), just to prevent surprises ... 02
08:15 bb-github left #buildbot
08:26 bb-github joined #buildbot
08:26 bb-github [13buildbot] 15benallard comment on pull request #1387 147127ff5: Removed. 02
08:26 bb-github left #buildbot
08:33 mattiasr joined #buildbot
08:41 bb-github joined #buildbot
08:41 bb-github [13buildbot] 15vincentxb comment on pull request #1378 1452aa15d: The hierarchy would rather be:... 02
08:41 bb-github left #buildbot
08:45 mjw joined #buildbot
08:48 bb-trac [trac] #3070/support-request (new) updated by Ben (Back in the time where I had a lot to do with Windows, I switched from cygwin to ...)
08:48 KeiKun joined #buildbot
08:49 1JTAA406V joined #buildbot
08:49 1JTAA406V [13buildbot] 15sa2ajj pushed 15 new commits to 06master: 02
08:49 13buildbot/06master 1410336f9 15Pierre Tardy: introduce CustomService...
08:49 13buildbot/06master 14a849151 15Pierre Tardy: CustomService config tests...
08:49 13buildbot/06master 14b306679 15Pierre Tardy: attempt for an integration test...
08:49 1JTAA406V left #buildbot
08:49 7JTABLG6Q joined #buildbot
08:49 7JTABLG6Q [13buildbot] 15sa2ajj closed pull request #1378: Services (06master...06services) 02
08:49 7JTABLG6Q left #buildbot
08:52 lissyx joined #buildbot
09:02 xavierd joined #buildbot
09:02 delanne morning
09:03 sa2ajj morning
09:03 delanne: fyi, i'm working on
09:04 hopefully, we'll get an instance today/tomorrow
09:05 and it will be automatically updated
09:06 delanne sa2ajj, the url changes ?
09:06 it was:
09:07 sa2ajj, thanks for doing this
09:07 sa2ajj it will be
09:07 delanne ok
09:07 misterj joined #buildbot
09:07 misterj hi all
09:09 sa2ajj hi :)
09:10 misterj Is this IRC chan the right place to ask questions about config problem one might have?
09:10 I'm a bit confused by the welcome message ^^
09:10 sa2ajj yes, it is the right place
09:11 misterj thanks
09:11 sa2ajj there're not so many people who read welcome message :)
09:11 misterj I do :p
09:11 I am trying to handle multiple repositories, and on two of these, I want to build on every commit to master and another branch
09:12 So what I am doing now is having multiple schedulers (one for each repo/branch, so with 2 repos and 2 branches that makes 4) and I am passing to those schedulers codebases with the right branches
09:13 but I guess we are not supposed to do so since I am not successful doing that... I might not understand really well the codebase concept I think
09:13 sa2ajj can you show the builder definition where you do the actual build?
09:14 misterj yes for sure I will paste relevant information to a pastebin
09:14 bb-github joined #buildbot
09:14 bb-github [13buildbot] 15sa2ajj closed pull request #1387: Build properties data (06master...06build_properties_data) 02
09:14 bb-github left #buildbot
09:20 misterj here it is
09:20 infobob[…]lHpfWLvLx1gc0BAT/ (repasted for misterj)
09:23 sa2ajj i did not look too deep, however at line 10 i'd recommend using .copy()
09:23 lines 11-12 essentially update allrepocodebase
09:23 and that could be the problem
09:23 misterj ugh python...
09:23 sa2ajj (you can put a debug print of allrepocodebase after line 35
09:23 misterj I'm still not really familiar with those copy things :/
09:24 But I thought I did and it was ok... Will try anyways :)
09:24 sa2ajj i thought i was, but recently i made a similar mistake :)
09:24 do add a debug print just to be sure
09:26 treeStableTimer probably should be None (or just omitted), this way you really get a build for each commit
09:27 * sa2ajj is afk for ~20mins
09:29 misterj ok you were right about the copy thing... I really need to practice python to avoid such mistakes -_-
09:29 ok, thank you very much :)
09:36 I needed to use the "deepcopy" method since there was reference in reference
09:36 but it is now working
09:41 dirichlet joined #buildbot
09:50 sa2ajj good to hear :)
11:02 misterj joined #buildbot
11:14 allegary joined #buildbot
12:24 KeiKun_ joined #buildbot
12:58 bb-trac [trac] #2851/task (accepted) updated by sysadmins (empty comment)
13:10 bb-github joined #buildbot
13:10 bb-github [13buildbot] 15benallard comment on issue #1380: Just pushed the last part of this sequence here. 02
13:10 bb-github left #buildbot
13:11 KeiKun joined #buildbot
13:19 bb-trac [trac] #2851/task (accepted) updated by Ben (Eat your own dog food, eh !)
13:21 [trac] #2851/task (accepted) updated by sa2ajj (Let's see how tasty it is :))
13:26 [trac] #2956/enhancement (assigned) updated by Ben (Can this now be closed ?)
13:32 [trac] #2956/enhancement (accepted) updated by sa2ajj (See above my comment about environment variables.  I have a branch, I just need to ...)
13:32 [trac] #3066/defect (new) updated by Ben (Unfortunately, those are not the only ones to be flaky:    {{{  [ERROR]  Traceback ...)
13:32 [trac] #2956/enhancement (accepted) updated by sa2ajj (But first I want to take care of #2851.)
14:00 tychoish joined #buildbot
14:01 tychoish joined #buildbot
14:01 tychoish joined #buildbot
14:03 tychoish joined #buildbot
14:03 wouter joined #buildbot
14:05 wouter hi -- is it possible for a triggered build (which is started with waitForFinish) to set a property for the build that fired it off?
14:07 sa2ajj i think it's other way around: when you trigger a build, you can provide additional propoerties
14:07 wouter yes, I know that, and I've been using that, but that's not what I need for this particular case
14:07 bb-trac [trac] #3066/defect (new) updated by dustin (Is that occurring after disabling the master tests (comment 4)?    The ...)
14:07 wouter let me explain
14:07 sa2ajj can yo uelaborate then please?
14:07 :)
14:08 wouter the triggering build is fired off by the change source, so has the git revision number
14:08 but it never does a "git clone" or something; the triggered build does that, and it does "make distcheck"
14:08 the tiggering build downloads the tarball generated by make distcheck in the triggered build, but the version number is generated by "git describe", which is information that buildbot doesn't have
14:09 so I would like the result of "git describe" to move from the triggered build to the triggering build, somehow
14:09 otherwise the triggering build doesn't know which file to download
14:09 if that makes sense
14:11 sa2ajj hmm
14:11 skelly sa2ajj: did you see my PR?
14:12 sa2ajj skelly: yes
14:12 looks good, i'll comment it
14:12 wouter: what you want to do makes sense, i am not sure about the approach though
14:12 wouter sa2ajj: at least I'm not crazy then ;-)
14:13 sa2ajj: I'm just thinking that maybe the triggered build could generate a file "<git revision SHA1>.desc" that would contain the output of "git describe" for that particular revision
14:14 sa2ajj if you have more than one slave, you'd need to make sure that a) it's available in a common place (somewhere @ master) b) that there's no way the file is "overwritten"
14:14 wouter but then of course I would need to be able to use the contents of a file that only exists after a particular triggered build has finished in my build factory
14:14 sa2ajj djmitche: if you are there, maybe you could comment?  (you definitely have a bit more experience than I :))
14:14 wouter oh, it would upload that file along with the tar.gz (with FileUpload)
14:15 but I'd still need to be able to use the contents of that file in an Interpolate() statement somehow
14:16 bb-trac [trac] #3066/defect (new) updated by Ben (It is indeed, when not running the disabled ones.)
14:17 * djmitche reads scrollback
14:18 wouter ah, I could use a properties renderer
14:19 djmitche yeah, you should be able to pass a property from triggered to triggering, assuming waitForFinish
14:22 sa2ajj: btw tomorrow is the start of a holiday weekend for me, with family around, so I'll likely not be available much
14:22 well, this afternoon
14:23 sa2ajj no worries :)
14:23 have a good one :)
14:23 djmitche great idea re: trac user btw
14:23 thanks :)
14:25 sa2ajj :)
14:27 wouter: btw, did you try to check the properties of the triggering build after the triggered one is finished?
14:27 (my work experience is limited to just fire triggered builds and not care about when they finish)
14:28 by checking i mean "check in the ui" (which is the simplest way)
14:31 wouter sa2ajj: that's the first thing I tried :-)
14:31 sa2ajj and it does not have any properties set by the triggered build?
14:32 wouter nope
14:32 sa2ajj and your triggered build sets some properties, does it not?
14:32 wouter but it looks like the webinterface only contains the properties of the triggering build which are set at the start
14:32 sa2ajj that could not be
14:32 wouter um. not sure about that part, actually ;-)
14:32 let me check
14:33 sa2ajj after a build is finished, it would show all the properties that were set along the way
14:33 wouter hm, no, it doesn't set properties
14:33 that would be... a good thing to test :-)
14:34 sa2ajj well, got_revision is one that is set by a 'git' step...
14:34 wouter ah, right, true
14:35 and there is no got_revision in that list of properties
14:36 * sa2ajj hmms
14:37 sa2ajj what i wonder is: if you triggered build does so little, why this can't be performed as part of the triggering build?
14:37 wouter because there are something like 20 triggering builds
14:37 sa2ajj oh
14:37 wouter and the first thing the triggered build does is check "does my output already exist", and if so, return immediately (successfully)
14:38 sa2ajj well, i'd first try to set some "custom" property: not the one from the standard list
14:38 and check if that propagates (i do not have a high hope for that though)
14:38 wouter I am trying to figure out how at the moment...
14:38 sa2ajj but djmitche above said that props should propagate...
14:38 you are using 0.8.9, right?
14:38 wouter checking...
14:38 yup
14:39 sa2ajj ok
14:39 beacuse if you were running master, that'd be a totally different story :)
14:39 wouter ah, but hang on
14:39 tromey joined #buildbot
14:39 wouter and the first thing the triggered build does is check "does my output already exist", and if so, return immediately (successfully)
14:39 ^^ that means we don't get at git if the output already exists
14:39 which means you don't get got_reivision
14:40 sa2ajj true
14:40 wouter which means, well...
14:40 sa2ajj so try to set a custom property then :)
14:40 wouter how do you do that?
14:40[…]g-properties.html doesn't say (unless I'm missing something)
14:42 sa2ajj[…]etting-properties
14:43 wouter ah, wonderful
14:43 thanks
14:46 * sa2ajj is afk for ~30 mins
14:46 wouter right
14:46 you've pointed me in the right direction, I think I'll be able to figure it out from here
14:47 thanks!
14:50 djmitche huh, tests are unstable for me now too :(
14:50 * djmitche bisects
14:53 wouter sa2ajj: nope, properties do not propagate to triggering builds
14:53 sa2ajj: but that's fine, I'm already doing FileUpload() on the output of the triggered build, I can add another file
15:11 sa2ajj right
15:13 KeiKun_ joined #buildbot
15:15 sa2ajj djmitche: if you have a problem and you decided to use 'git bisect', you have two problems :)
15:15 djmitche haha
15:15 wait'll I pull out 'git trisect'
15:16 huh, I wonder if Buildbot just doesn't work under 'trial -u' anymore
15:18 mids joined #buildbot
15:19 mids I am trying to show only certain projects in my waterfall view by adding the ?project=name argument, but filters out build events; but still shows the builder columns for other projects. is there a way to also hide them?
15:20 sa2ajj i'm afraid not. iirc correctly, waterfall page shows _all_ builders
15:20 djmitche: and it works w/o -u?
15:20 that'd be strange
15:23 bb-github joined #buildbot
15:23 bb-github [13buildbot] 15benallard opened pull request #1392: Show the Properties on the web interface (06master...06properties_web) 02
15:23 bb-github left #buildbot
15:23 bb-github joined #buildbot
15:23 bb-github [13buildbot] 15benallard comment on issue #1392: @tardyp: Any opinion about changing the property data type ? Any advice on how to start ? 02
15:23 bb-github left #buildbot
15:27 bb-trac [trac] #3071/defect (v:master) created by Ben (/builds/:id/changes returns error)
15:31 [trac] #3071/defect (new) updated by sa2ajj (empty comment)
15:34 mids sa2ajj: thx
15:35 sa2ajj sorry about that
15:46 jaredgrubb joined #buildbot
15:54 sa2ajj tardyp: what is the requirement for node?
15:55 tardyp sa2ajj: a recentish version.
15:55 sa2ajj can i believe what i see in package.json?
15:55 tardyp yes
15:55 however it is not tested
15:55 sa2ajj ok
15:56 let's see :)
15:56 tardyp it is the version that I had on my laptop last time I updated the package.json
15:56 what I recommend is to just install the latest version
15:56 and dont use the one of the distro
15:56 node is moving too fast for distros
15:56 sa2ajj i'd like to try w/ the freebsd package first
15:57 wouter tardyp: that very much depends on your requirements
15:57 tardyp I'm not sure that this freebsd thing is a good idea
15:57 wouter tardyp: if your requirements are "need to run most recent code", then "use distro package" is almost always the wrong idea
15:58 tardyp I would actually prefer to do builds on a linux vm with docker containers
15:58 wouter if your requirement is "need to use something which will work for the next five years", then well
15:58 tardyp for now, buildbot nine is unreleased, so we dont freeze
15:58 sa2ajj tardyp: freebsd jail is what i currently have
15:58 tardyp and dont test for released version of node.
15:59 at some time we will have to deal with supporting versions of node, indeed
15:59 sa2ajj: I know, and would have disagreed if asked
16:00 freebsd jail is not something everybody knows about
16:01 tromey that's where openbsd developers end up
16:01 bb-github joined #buildbot
16:01 bb-github [13buildbot] 15benallard comment on issue #1305: The karma test is failing since this:... 02
16:01 bb-github left #buildbot
16:02 tardyp The thing is everybody is at least fluent with one linux distro
16:02 even bsd guys
16:02 bb-github joined #buildbot
16:02 bb-github [13buildbot] 15benallard comment on issue #1305: Just opened TRAC-3071 with an error when visiting the new data endpoint. (also posting it here to prevent it for being unseen.) 02
16:02 bb-github left #buildbot
16:03 skelly FreeBSD packages has 0.10.33 in ports
16:03 sa2ajj tardyp: i'll note your preference
16:03 tardyp which oposite is wrong. I have no idea about freebsd, and if I'm given a jail for nine showcasing purposes, I will be annoyed
16:03 skelly er, ports, not packages
16:04 tardyp So I would have liked that the new infra at least support some CoreOS VMS you would give for developers, so that we can run whatever distro we are fluent with in a container
16:04 * sa2ajj nods
16:05 sa2ajj we're in process of _deploying_ the infrastructure
16:05 and the main goal is to make sure that existing services can be easily re-deployed should it be necessary
16:05 they all run under freebsd atm
16:05 tardyp yes, services, no problem
16:06 sa2ajj the only reason why i try to deploy nine under freebsd is because we do not have anything else, though there're plans
16:06 skelly there's vm1 that will surely have a number of Linux VMs
16:06 tardyp but in the initial plan, there was a VM machine
16:06 sa2ajj yes
16:06 ^^
16:06 tardyp this is the one, I was expecting to deploy the nine demo.
16:07 I think we need a nine metabuildbot
16:07 which is updated from time to time
16:07 and a nine demo bot for demo
16:07 that we could autoupgrade in cronjob or something
16:08 * sa2ajj totally agrees
16:09 tardyp My wish is that we could afford to give one VM for every developer so that they can showcase their UI changes, and have a discussion in the PR
16:10 we could optionally share the DB between those UI only show case VMs
16:10 * tardyp not complaining.. You guys are doing a great job
16:11 bb-trac [trac] #3072/task (v:) created by sa2ajj (setup vm1)
16:11 skelly tardyp: what sort of specs would you want for the dev VMs?
16:12 tardyp Well I'm not expecting a lot of concurrent active VMs
16:13 skelly right, I'm thinking more about disk and memory resources
16:13 tardyp buildbot is mostly mono threaded, so nothing more than one core VM, with 1GB ram, and 1GB HD
16:14 then if there is an external mysql server that would make things easier
16:15 each dev would be given a VM plus a database access. one db with rw, and ro on the metabuildbot db
16:16 my master dockerfiles are ~512MB worth
16:16 bb-github joined #buildbot
16:16 bb-github [13buildbot] 15jaredgrubb comment on issue #1387: LGTM! 02
16:16 bb-github left #buildbot
16:16 tardyp maybe 1GB HD is a bit low
16:17 skelly I don't have access to vm1, but it does have 4x100 GB SSDs
16:24 bb-github joined #buildbot
16:24 bb-github [13buildbot] 15tardyp comment on issue #1392: @benallard , you probably want to do that in the data api.... 02
16:24 bb-github left #buildbot
16:33 bb-github joined #buildbot
16:33 bb-github [13buildbot] 15benallard comment on issue #1387: @jaredgrubb, thanks, but that one was merged this morning already ;... 02
16:33 bb-github left #buildbot
16:42 bb-github joined #buildbot
16:42 bb-github [13buildbot] 15benallard comment on issue #1392: I tried to do that, but I always end up getting '[null]' in my properties value. I even modified the types.SourcesProperties class, and this didn't brought me any steps further. The unittest wasn't happy because of other check I wasn't able to find ... That's when I decided to push it as-is.... 02
16:42 bb-github left #buildbot
17:02 bb-github joined #buildbot
17:02 bb-github [13buildbot] 15tardyp comment on issue #1392: hum. I see.. indeed this is more work, you'll have to change the unit tests spec 02
17:02 bb-github left #buildbot
17:24 jorgev joined #buildbot
17:30 lissyx joined #buildbot
17:35 bb-trac [trac] #3072/task (new) updated by sa2ajj (vm1 is also a !FreeBSD machine.)
17:39 pocek joined #buildbot
18:12 lissyx joined #buildbot
18:17 * djmitche runs the bootstrap script on vm1 :)
18:21 sa2ajj djmitche: success?
18:22 bb-github joined #buildbot
18:22 bb-github [13buildbot-infra] 15djmitche opened pull request #22: Add vm1 (06master...06add-vm1) 02
18:22 bb-github left #buildbot
18:22 bb-trac [trac] #3072/task (new) updated by dustin (I have a login.  I just ran[…]bot-infra/pull/22  ...)
18:22 skelly now need mac1!
18:23 sa2ajj _actually_ use ssh control session, right :)
18:23 bb-github joined #buildbot
18:23 bb-github [13buildbot-infra] 15sa2ajj closed pull request #22: Add vm1 (06master...06add-vm1) 02
18:23 bb-github left #buildbot
18:24 bb-trac [trac] #2851/task (accepted) updated by dustin (There as some discussion on ...)
18:25 djmitche ISTR we haven't actually bought the mac yet
18:26 skelly that's unfortunate considering the Mac Mini changes
18:26 djmitche changes?
18:27 skelly there was a "server" edition before for two hard drives and there isn't now as the limit is one
18:27 djmitche ah
18:27 skelly also, no quad core processor, just dual; memory is also soldered on, but that might not be as big of a deal if buying direct from Apple anyway
18:27 djmitche otoh, apple hardware falls out of date quickly
18:28 so not buying it until we're ready to throw it into production is good
18:28 and I may be wrong about not buying it
18:28 skelly there is hope it's just temporary due to Intel's segmenting
18:29 djmitche yeah, who knows
18:29 ok, off to make some pies n stuff
18:29 sa2ajj djmitche: i'll remove from the zone?
18:29 djmitche and sh9ovel
18:29 sa2ajj: yep, good call
18:29 oh
18:29 what happened with
18:29 bind? nsd?
18:30 sa2ajj it should bind
18:31 verm__ yesterday created the jail, but we did not populate it yet
18:31 djmitche I think he put nsd on there
18:32 see bug updates
18:32 bb-github joined #buildbot
18:32 bb-github [13buildbot-infra] 15sa2ajj opened pull request #23: add to zone (06master...06add-nine-to-dns) 02
18:32 bb-github left #buildbot
18:33 sa2ajj yes, he did
18:33 djmitche looks like BIND is fine for now
18:34 sa2ajj he put some other secondaries
18:34 * sa2ajj is confused
18:36 sa2ajj a persimon might help with the confusion
18:39 bb-github joined #buildbot
18:39 bb-github [13buildbot-infra] 15sa2ajj pushed 2 new commits to 06master: 02
18:39 13buildbot-infra/06master 14cf6ae29 15Mikhail Sobolev: add to zone...
18:39 13buildbot-infra/06master 1433a14e2 15Mikhail Sobolev: Merge pull request #23 from sa2ajj/add-nine-to-dns...
18:39 bb-github left #buildbot
18:43 KeiKun joined #buildbot
18:44 sboudria_ joined #buildbot
18:58 bb-github joined #buildbot
18:58 bb-github [13buildbot-infra] 15sa2ajj opened pull request #24: add a list of host fingerprints (06master...06add-fingerprints) 02
18:58 bb-github left #buildbot
19:06 djmitche posted to sysadmins
19:08 sa2ajj i posted this pr with fingerprints, but i am now confused :(
19:09 those are fingerprints for .pub keys found in /etc/ssh/
19:09 however ssh-keyscan shows a smaller number of keys...
19:09 skelly does it skip some algorithms?
19:10 sa2ajj yes, i think the version is not the latest
19:10 djmitche mine is a DSA key
19:10 oh
19:10 host keys
19:10 * djmitche shuts up
19:11 sa2ajj seems to be working
19:11 but i had troubles installing pysqlite using pip in the virtual environment
19:12 (it'd require some --global-options or something)
19:12 but the above result depends on some manual work :(
19:22 sboudria_ joined #buildbot
19:26 djmitche we should probably use mysql or postgres
19:27 bstack joined #buildbot
19:28 sa2ajj yes, we should use some "real database", as bellanard usually says :)
19:31 djmitche heh, yeah
19:31 if it's easier to install especially
19:31 pg8000 is pure-python
19:31 as is python-mysql
19:31 er, pymysql
19:31 sa2ajj do we want to create a virtualenv for master?
19:32 or i could resort to 'pin install ... --user'?
19:32 the reason i ask is that in the latter case we can satisfy all dependencies from the system
19:32 which might be good or not
19:35 djmitche I'd lean toward a virtualenv
19:36 sa2ajj ok, i'll keep trying then :)
19:37 re pg8000 -- i'd rather use a C binding of sort
19:39 djmitche yeah, as long as it doesn't require libmysql or sometihng system-level
19:41 sa2ajj :)
19:41 why you do not like system-level things?
19:44 djmitche it's just a little harder to coordinate
19:44 you're at the whim of system upgrades
19:44 etc.
19:45 sa2ajj hmm... ok.
19:50 djmitche it's not a hard preference
19:50 esp since we're installing the whole system here
19:53 sa2ajj i somehow prefer system wide things... :)
19:55 but let's have something first
19:57 djmitche I hate every systemwide packaging system with a blinding passion
19:58 so that might have something to do with it
19:58 so far FreeBSD hasn't earned that hatred, but I'm sure it will :)
19:58 sa2ajj :)
19:58 djmitche I'm happy to learn better too :)
20:00 sa2ajj they all suck, do not worry :)
20:01 the main reason i like system wide stuff (i.e. packages of sort) == reproducibility (if i spelled it right)
20:01 with them i had success, w/o them -- never
20:06 djmitche that might be better with FreeBSD
20:07 with e.g., Ubuntu, 'apt-get install foo' can end up with something totally differente every time
20:07 since packages get updated so quickly
20:07 maybe we end up running our own pkgng mirror, dunno..
20:07 I don't really even want to think about that right now
20:08 sa2ajj then don't: holiday is around the corner :)
20:09 djmitche exactly
20:19 skelly if you hate portage, than you will probably hate ports
20:51 jaredgrubb joined #buildbot
21:14 jorgev joined #buildbot
21:47 DOS622 joined #buildbot
22:26 bb-github joined #buildbot
22:26 bb-github [13buildbot] 15vadz comment on pull request #1391 14f5aaaac: Unfortunately I think it _is_ a concern (personally I never use paths with spaces as it's just a recipe for disaster, so not for me, but I'm pretty sure this would bite others) and `shlex` doesn't really help because it treats backslashes as escape characters, meaning that it totally mangles Windows paths.... 02
22:26 bb-github left #buildbot
22:36 jaredgrubb joined #buildbot
22:50 KeiKun_ joined #buildbot
23:06 bb-github joined #buildbot
23:06 bb-github [13buildbot] 15vadz comment on issue #1391: Adding release note is not a problem (it should go into `docs/relnotes/index.rst`, right?), but I'm stumped by Travis failures. I don't understand why do they happen nor why do they happen in only the two builds using `TWISTED=latest`. And I have trouble reproducing them locally because I don't seem to be able to run the tests at all:... 02
23:06 bb-github left #buildbot

← Previous day | Index | Server Index | Channel Index | Today | Next day → | Atom Feed | Search | Google Search | Plain-Text | plain, newest first