17:00:50 <jzb> #startmeeting
17:00:50 <cs-meeting> Meeting started Wed Jan 2 17:00:50 2013 UTC. The chair is jzb. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:50 <cs-meeting> Useful Commands: #action #agreed #help #info #idea #link #topic.
17:01:19 <jzb> #chair bhaisaab chipc iswc kdamage ke4qqq Noah topcloud u-ichi widodh
17:01:19 <cs-meeting> Current chairs: Noah bhaisaab chipc iswc jzb kdamage ke4qqq topcloud u-ichi widodh
17:02:15 <jzb> OK, let's get started
17:02:33 <jzb> bhaisaab: anything to discuss or report?
17:03:08 <jzb> OK, we'll circle back if bhaisaab turns up
17:03:14 <jzb> chipc: anything to discuss or report?
17:03:25 <chipc> yes
17:03:46 <chipc> first up, I sent out an email asking folks working on 4.1.0 features / improvements to help tidy up Jira
17:04:33 <chipc> there are three things that I'm concerned about right now: (1) items tagged for the release that don't have anyone assigned, (2) items tagged for the release that don't have any progress indicated and (3) items not tagged or even in Jira
17:04:56 <chipc> if people can look and take appropriate actions, that would be good
17:05:11 <jzb> #chair vogxn
17:05:11 <cs-meeting> Current chairs: Noah bhaisaab chipc iswc jzb kdamage ke4qqq topcloud u-ichi vogxn widodh
17:05:12 <vogxn> meeting today?
17:05:15 <jzb> vogxn: yep
17:05:17 <chipc> second item: reminder that the code freeze is scheduled for the end of the month
17:05:52 <chipc> third item: I took the liberty of kicking of a VOTE for project bylaws… but I'm more than happy to cancel if we need more discussion
17:05:57 <chipc> that's it
17:06:30 <jzb> chipc: is that the usual procedure?
17:06:40 <jzb> or is there a usual procedure? (For bylaws)
17:06:59 <chipc> well, I'm frankly not sure!
17:07:10 <chipc> but I don't see why not
17:07:13 <jzb> fair enough. Seems a reasonable approach, then.
17:07:26 <jzb> chipc: thanks!
17:07:35 <jzb> iswc: you're up next
17:07:35 <chipc> if folks have concerns (procedural or content), bring them up on the thread
17:08:09 <jzb> last draft I looked at seemed sane, but I'll be sure to comb over it one more time before voting.
17:08:14 <jzb> iswc: anything to discuss or report?
17:09:15 <jzb> OK, moving on
17:09:22 <jzb> kdamage: anything to discuss or report?
17:09:45 <kdamage> Nope still ramping down from the holidays (smile)
17:10:02 <jzb> kdamage: I hear ya.
17:10:19 <jzb> thanks, kdamage
17:10:27 <jzb> ke4qqq: anything to discuss or report?
17:10:58 <ke4qqq> nothing
17:11:10 <jzb> succinct
17:11:11 <jzb> thanks!
17:11:17 <jzb> Noah: anything to discuss or report?
17:12:30 <jzb> OK
17:12:38 <jzb> #chair serverchief
17:12:38 <cs-meeting> Current chairs: Noah bhaisaab chipc iswc jzb kdamage ke4qqq serverchief topcloud u-ichi vogxn widodh
17:12:46 <jzb> serverchief: welcome aboard - happy 2013 to you
17:12:50 <jzb> you got here just in time.
17:12:58 <jzb> Anything to discuss or report?
17:14:49 <jzb> ok, we'll move on
17:15:02 <jzb> topcloud: your turn - anything to discuss or report?
17:15:07 <topcloud> yup.
17:15:35 <topcloud> i recently found that if we go to the wiki through the following two links, the look is quite different.
17:15:55 <topcloud> was interested in getting some idea from people who are more familiar with confluence on how to fix this.
17:16:04 <topcloud> www.cloudstack.org and click on the cloudstack wiki
17:16:14 <topcloud> cwiki.apache.org and click on cloudstack.
17:16:45 <topcloud> from both of these links, when you arrive at cloudstack wiki, the left side toc frame is not there.
17:16:55 <topcloud> but if you click on edit page and then cancel, it comes back.
17:16:58 <vogxn> think the link is /confluence/display/CLOUDSTACK
17:17:04 <vogxn> and /CLOUDSTACk
17:17:05 <ke4qqq> topcloud: one is the wiki itself (which we aren't supposed to link to)
17:17:14 <ke4qqq> the other is the html export of the wiki
17:17:37 <ke4qqq> https://cwiki.apache.org/CWIKI/
17:17:49 <chipc> topcloud: so that's probably my fault
17:17:50 <jzb> ke4qqq: what is the proper link?
17:17:52 <topcloud> ke4qqq: that part i figured out. but how do we stop people from linking to the html or have html point to the webapp?
17:17:53 <chipc> I'll fix that link
17:18:08 <ke4qqq> topcloud: people should only link to the html
17:18:16 <chipc> ke4qqq: really?
17:18:24 <chipc> but that's not editable, right?
17:18:27 <ke4qqq> read the bullet point in the link above
17:18:41 <ke4qqq> 'The first rule of CWIKI is not to link to CWIKI'
17:18:51 <chipc> ha
17:18:52 <chipc> ok then
17:18:57 <topcloud> crap.
17:18:59 <ke4qqq> There is an 'edit' button on the html
17:19:05 <topcloud> yes there is
17:19:11 <ke4qqq> that will take you to the wiki itself
17:19:14 <topcloud> and that's why if you cancel, it goes back to the wiki.
17:19:27 <topcloud> but the toc frame is not available on the html.
17:19:34 <topcloud> anyway to make that work better?
17:19:50 <ke4qqq> no idea - might be something to ask infra
17:19:53 <topcloud> because just going to the html doesn't give you enough context on what to do next.
17:20:26 <topcloud> ke4qqq: mind taking that up?
17:21:04 <ke4qqq> topcloud: you want the TOC available on the exported html? (just to be clear?)
17:21:14 <topcloud> ke4qqq: yes.
17:21:44 <topcloud> iirc, the page nav is also not available on the html.
17:22:04 <topcloud> so even if i setup a child page nav and put it on the page itself, people still can't see it.
17:22:55 <topcloud> my second issue is this and this may be a citrix only issue but like to get some feedback.
17:23:43 <topcloud> there's quite a few devs who are doing feature development for 4.1 who are not committers. for them, the problem is the contribution process takes a long time.
17:23:58 <topcloud> i can see it today in api refactoring between likitha and rohit.
17:24:18 <ke4qqq> topcloud - showchildren at the bottom enough or not?
17:25:08 <topcloud> ke4qqq: showchildren will be acceptable if no other alternative.
17:25:29 <ke4qqq> ok
17:25:56 <topcloud> so i'm trying to wrap my head around how to help these guys do faster development.
17:26:08 <topcloud> but not sure there's any clear path about this.
17:26:25 <topcloud> i think sateesh also brought this up about doing a branch to do dev on github.
17:26:39 <chipc> topcloud: I thought we had previously discussed how to do this… that significant development happening by a non-committer should probably be done on a public repo somewhere
17:26:39 <jzb> topcloud: part of the "long time" is intentional
17:26:41 <topcloud> any ideas?
17:26:55 <jzb> topcloud: so that other developers have an opportunity to review, comment, contribute, & collaborate
17:27:05 <jzb> speed is not the top priority for an Apache project
17:27:37 <jzb> topcloud: is there a problem with actual getting comments on reviews, etc.?
17:27:49 <topcloud> no i think the problem is this.
17:27:49 <jzb> that has been a problem I think we need to continue to improve on.
17:27:56 <topcloud> they basically take a branch and commit there.
17:28:05 <chipc> topcloud: one option might be for committers to sign up to "mentor" non-committers through the process… by acting as their go-to person (via the list) for getting reviews done
17:28:19 <chipc> they could also help ensure that the development is happening in a collaborative and open way
17:28:22 <topcloud> but every commit requires review process for them and they're really technically competent on cloudstack already.
17:28:57 <topcloud> for example, is there a way to open branch without review process and only review when merge to main?
17:29:22 <ke4qqq> topcloud: no offence
17:30:01 <topcloud> ke4qqq: no offence on anything. like i said, i'm just thinking about it. any feedback welcomed.
17:30:13 <ke4qqq> offense- but being employed to work on cloudstack is not the same as being competent or trusted by the community
17:31:05 <ke4qqq> some folks who are relatively new (like Min) have earned trust rapidly - others aren't participating much regardless of competence and thus have no visibility
17:31:59 <topcloud> ke4qqq: yeah...i'm not looking for a way for all citrix guys to be committers for example.
17:32:07 <ke4qqq> git has no per-branch auth - and they'd still need to go through the committer process (vote, CLA, account creation) even if we were using svn which does have that.
17:32:53 <topcloud> i'm only trying to see how we can help them collaborate among themselves without this elaborate process. i mean one way is private branch but we don't want that.
17:33:11 <ke4qqq> but as jzb - speed isn't the goal - growing the community is - giving someone commit access so they don't have to interact or communicate with the rest of the community seems counterintuitive to the goal.
17:33:27 <chipc> topcloud: why is using a public repo for collaboration not useful?
17:33:37 <topcloud> i think part of growing the community is giving people good efficient process.
17:34:27 <chipc> topcloud: As I said in the email thread, I thought we had discussed that groups of folks collaborating on a feature (who are not committers) should be encouraged to (1) use a public repo like github, and (2) ensure that they share and communicate on the dev list about that repo and progress
17:34:32 <topcloud> chipc: because every commit in a public branch on apache requires this review process. imagine a couple of them collaborating on this and having to do it for every little checkin.
17:34:35 <ke4qqq> topcloud: this same process works for scores of other apache projects, what is so different here?
17:35:18 <chipc> topcloud: but every commit to the asf repo doesn't have to be the same as every commit on a collaboration repo
17:35:37 <chipc> so you spend time working with another developer (talking on the list), using a github fork
17:36:05 <chipc> when you are done, the commits are squashed into logical ones for each developer to send through reviewboard
17:36:25 <topcloud> chipc: that's actually pretty good. the reason that i brought it up is because in one of the threads sateesh was discouraged from putting a branch on github.
17:36:26 <chipc> we want clean commits coming through reviewboard anyway, right?
17:36:36 <kdamage_> That makes sense
17:36:56 <topcloud> i'm happy with github approach.
17:37:30 <chipc> topcloud: I think widodh was making the point about visibility
17:37:40 <chipc> the issue with going off on their own is visibility for the community
17:37:49 <topcloud> ah ok...i see you guys responded on that thread...nevermind
17:38:03 * bhaisaab o/
17:38:18 <jzb> topcloud: any other issues?
17:38:29 <topcloud> that's it for me.
17:38:43 <jzb> topcloud: spiffy. Thanks!
17:39:05 <jzb> u-ichi: you're up - anything to discuss or report?
17:39:23 <chipc> topcloud: we should make sure that the email thread about process gets wrapped up on the thread itself (wink)
17:40:54 <jzb> OK, moving on
17:41:05 <jzb> vogxn: anything to discuss or report?
17:41:09 <vogxn> 2 tthings
17:41:18 <vogxn> cloudstack builds are failing on builds@
17:41:30 <vogxn> there's some problem with the nodes and mvn
17:41:40 <vogxn> i'm looking to fix that
17:41:51 <vogxn> but dunno if we have access to teh jenkins slaves
17:42:13 <vogxn> anybody know if we do?
17:42:31 <jzb> I expect ke4qqq knows
17:42:46 <topcloud> chipc: yup. i should have just said "i like to bring attention to sateesh's email thread"
17:42:49 <ke4qqq> we don't have access to the jenkins slaves
17:42:58 <ke4qqq> but jfarrell has offered to help and has access
17:43:09 <vogxn> ke4qqq: ok. will follow up on infra then?
17:43:25 <ke4qqq> vogxn: yes or just /msg farrell
17:43:30 <vogxn> sure
17:43:33 <vogxn> 2nd thing
17:43:34 <ke4qqq> he's volunteered to help multiple times
17:44:03 <vogxn> still haven't recieved response on the test infra being moved
17:44:18 <vogxn> so - will talk to infra bout that too and keep the lists posted
17:44:41 <ke4qqq> vogxn: I'd follow up directly with gmcdonald - he's in Oz, so keep that in mind for the timezone diff
17:44:53 <vogxn> yeah will do
17:45:40 <vogxn> i'm done
17:45:44 <jzb> vogxn: thanks!
17:45:55 <jzb> widodh: you're up - anything to discuss or report?
17:46:28 <jzb> #action vogxn follow up with gmcdonald / Infra about Jenkins slaves
17:48:08 <jzb> OK
17:48:20 <jzb> Going last
17:48:41 * bhaisaab back
17:48:44 <jzb> so - on my end, I am going to try to ride 4.0.1 to vote by end of week
17:48:51 <jzb> or Monday at latest
17:49:07 * bhaisaab roll back to me at last
17:49:08 <jzb> the problem at this point is the upgrade procedure
17:50:01 <jzb> we also have two bugs that AFAIK have patches that are still waiting on review
17:50:10 <jzb> that's it for me.
17:50:34 <jzb> #action jzb working on 4.0.1 get to vote by Monday
17:50:42 <jzb> bhaisaab: you're up
17:51:03 <bhaisaab> thanks jzb, sorry folks missed the starting of the meeting by 30 mins confusion
17:51:32 <bhaisaab> api_refactoring is coming good, reverted some changes so we've backward compatibility with param processing
17:51:58 <bhaisaab> ke4qqq asked for a list of changes, which I'll post this weekend alongwith merge request2
17:52:11 <bhaisaab> that's it for me
17:52:23 <chipc> bhaisaab: any update on testing to ensure no regressions?
17:52:46 <chipc> (I'm catching up on the dev list now, but wondering what the tl;dr version is)
17:52:53 <bhaisaab> chipc: since the cmd classes have not changed, what we need is integration tests
17:53:04 <chipc> yes, agreed
17:53:06 <bhaisaab> some of them are already there
17:53:24 <bhaisaab> and some of them I'll ask tsp to help me with
17:53:29 <ke4qqq> bhaisaab: maybe look at jclouds - they have some intense CS api tests
17:54:01 <bhaisaab> I'm hoping to finish the tests and some remaining work and submit a merge request for evaluation this weekend
17:54:08 <jzb> #action bhaisaab post list of changes for api_refactoring
17:54:12 <bhaisaab> ke4qqq: good idea, will llook into that
17:55:07 <bhaisaab> any more questions, suggestions?
17:56:00 <chipc> bhaisaab: thanks for the patience dealing with our concerns about the merge
17:56:16 <chipc> but I think they were important points to sort through
17:57:00 <bhaisaab> chipc: thanks to community the last vote and comments were inspiring... seeing a unanimous -1 helped me clear my confusion (smile)
17:57:15 <chipc> (wink)
17:57:18 <chipc> fair enough
17:59:13 <jzb> any other issues, topics, etc/
17:59:14 <jzb> ?
17:59:40 <jzb> if not, I'll go ahead and end the meeting
17:59:45 <jzb> thanks all for joining
17:59:52 <jzb> good start to 2013
18:00:06 <jzb> #endmeeting

  • No labels