14:00:58 <nikhil_k> #startmeeting glance_artifacts_sub_team
14:00:59 <openstack> Meeting started Mon Sep 14 14:00:58 2015 UTC and is due to finish in 60 minutes.  The chair is nikhil_k. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:01:00 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:01:03 <openstack> The meeting name has been set to 'glance_artifacts_sub_team'
14:01:08 <ativelkov> o/
14:01:20 <dshakhray> o/
14:01:46 <nikhil_k> So, we do not have a set agenda for today.
14:02:07 <nikhil_k> I wanted to quickly sync for some highlights that we need to prep for the summit
14:02:17 <nikhil_k> you guys okay with that/
14:02:18 <nikhil_k> ?
14:02:23 <ativelkov> nikhil_k: sure
14:02:32 <nikhil_k> #topic summit prep items
14:02:48 <nikhil_k> SO, we are still in EXPERIMENTAL API phase
14:02:48 <mfedosin> o/
14:03:00 <nikhil_k> and we are aware that there's lot of work ativelkov is doing to fix that
14:03:23 <ativelkov> we have 3 ppl here working with that
14:03:39 <nikhil_k> however, with the dis-grumblement from people on GLance
14:03:45 <mfedosin> yep, we're discussing it a lot these days
14:03:59 <sigmavirus24> o/
14:04:00 <nikhil_k> I wanted to see how better can we communicate and make this an interactive project for the community
14:04:29 <nikhil_k> THere's a accepted  session on this at Tokyo summit
14:04:35 <nikhil_k> can we do something prior to that?
14:04:42 <nikhil_k> May be start with a blog? ativelkov ?
14:05:12 <ativelkov> nikhil_k: sounds nice. So, you need a blog on basic concepts of v3-to-be?
14:05:13 <nikhil_k> Also, would documenting use cases and the reasoning behind the API , would that make sense?
14:05:33 <mfedosin> I hope this week we'll write a bunch of docs and proposals for v3
14:05:40 <nikhil_k> ativelkov: yes, I think what we have in etherpad isn't very reachable across domains
14:06:06 <nikhil_k> may be start with v3-to-be starting with the already documented stuff
14:06:07 <sigmavirus24> mfedosin: how's the API spec coming?
14:06:10 <nikhil_k> in a blog
14:06:23 <sigmavirus24> speaking of docs and proposals for v3 ;)
14:06:49 <nikhil_k> that's the reason for the blog-ask sigmavirus24
14:07:05 <nikhil_k> so that openstackers non-glancers will find it convenient and easy read
14:07:22 <ativelkov> nikhil_k: where should that be published?
14:07:24 <nikhil_k> and make effort to provide feedback on already existing use cases
14:07:27 <ativelkov> openstack wiki?
14:07:42 <nikhil_k> ativelkov: that sounds good
14:07:49 <ativelkov> We can try making it to mirantis blog, but the corporate approvals will take time
14:07:59 <mfedosin> sigmavirus24, I'm in two minds - to abandon the current spec and write new from scratch or to update it with new PS to Mitaka
14:08:13 <nikhil_k> I think we should leave some links incl. spec, an etherpad for discussion etc for people to provide feedback on that
14:08:29 <ativelkov> mfedosin: abandoning is bad idea, as it has some feedback from reviewers
14:08:41 <nikhil_k> ativelkov: openstack wiki sounds simpler to me too
14:08:43 <mfedosin> many things is going to be changed there
14:08:44 <ativelkov> and we are trying to address that feedback, not to hide it
14:08:57 <nikhil_k> ativelkov: so that it's easier to update and add a watch on the page
14:09:23 <ativelkov> its fine if the new patchset changes 100% of lines of code, but please leave all that comments in place
14:09:27 <ativelkov> nikhil_k: got it. Will do
14:09:35 <nikhil_k> Thank you.
14:09:37 <sigmavirus24> mfedosin: new patchsets are fine for that purpose
14:09:53 <sigmavirus24> I just want a spec describing the API that's somewhere even if it isn't technically approved for L
14:09:57 <mfedosin> ativelkov, sigmavirus24, okay. I got you :)
14:10:32 <nikhil_k> One more thing that keeps bothering me.
14:10:51 <nikhil_k> We currently do not have any driver in glance for artifacts
14:11:10 <nikhil_k> even though, it's a big project and priority for glance as we picked for L
14:11:44 <nikhil_k> I wanted to do one more round of rotation this week and propose one of you for this reason
14:12:10 <ativelkov> So, this is technically a core-reviewer for artifact specs?
14:12:11 <nikhil_k> there's been enough work on this topic, good interaction on corres specs and discussion on images too
14:12:20 <nikhil_k> ativelkov: for all glance-specs
14:12:42 <mfedosin> I agree :)
14:12:51 <nikhil_k> I don't think any glance spec can be ignored on the impact it will have on artifacts for non-artifact proposal (too)
14:12:58 <ativelkov> Agree
14:13:01 <nikhil_k> and visa versa
14:13:19 <ativelkov> yes, it makes sense - especially given that we'll wrap v2 under the v3 hood
14:13:38 <mfedosin> consider me as a volunteer
14:13:56 <ativelkov> +1 on mfedosin nomination :)
14:14:06 <mfedosin> btw, will anyone use this wrap?
14:14:30 <mfedosin> I mean it's easier to use v2 directly, isn't it?
14:14:34 <nikhil_k> so, I was hoping to add ativelkov just on the basis that he's been around for the longest period of time and driven artifacts for so long but I can add mfedosin if he has more b/w and on the same page as previously discussed approch for artifacts as well as interactive updates to the same
14:14:57 <mfedosin> add both ;)
14:15:00 <nikhil_k> just wanted to throw that out there
14:15:04 <ativelkov> nikhil_k: Mike has more time available: I'm at least 50% Murano right now
14:15:12 <nikhil_k> so that there's good justification for the this nomination
14:15:19 <nikhil_k> ativelkov: gotcha.
14:15:23 <nikhil_k> mfedosin: thanks for the ack
14:15:29 <ativelkov> I'll keep an yey on the specs as well, but cannot make any signifcant time commitments
14:15:51 <nikhil_k> sounds good
14:16:12 <mfedosin> I'll try my best to review all the specs
14:16:15 <nikhil_k> #action nikhil_k to propose mfedosin for glance-specs core nomination
14:16:33 <ativelkov> mfedosin: if some artifact type needs a (dynamic) reference to an image-as-artifact, it will use v3
14:17:01 <nikhil_k> mfedosin: I think what we need more now is cross team collaboration and some feedback +1/-1 directions to specs, attending drivers meeting is big plus!
14:17:40 <nikhil_k> whatever you can help with on those fronts would be helpful towards building a stronger community
14:18:03 <mfedosin> It's a little bit tricky, because I have downstream meeting at the same time
14:18:37 <nikhil_k> mfedosin: no worries, we can reschedule if needed or pick stuff up on thursday at allocated time slot
14:18:37 <mfedosin> but I think I can handle it
14:18:44 <nikhil_k> cool
14:18:53 <nikhil_k> That was it from my end for today.
14:18:58 <nikhil_k> #open discussion
14:19:02 <nikhil_k> #topic open discussion
14:19:15 <mfedosin> Let's talk about commits on review
14:19:22 <ativelkov> nikhil_k: are you aware that we are co-presenting a talk on Glance Api evolution of the main summit? :)
14:19:35 <mfedosin> we made a list of Murano blockers
14:19:54 <nikhil_k> ativelkov: haha, that's what I meant earlier in this meeting today. the approved summit talk on the glance v3 api :)
14:19:58 <mfedosin> https://etherpad.openstack.org/p/fastTrackPatches
14:20:10 <ativelkov> nikhil_k: ah, got it.
14:20:20 <nikhil_k> ativelkov: thanks though!
14:20:42 <ativelkov> Thanks for the link, mfedosin
14:20:53 <ativelkov> So, there are 6 fastTrack patches which are blockers for Murano
14:20:54 <nikhil_k> mfedosin: can we categorize them based on bugs and API, config etc impact that would not be proper for Liberty RC period?
14:21:26 <nikhil_k> I am looking if all are bugs or not..
14:21:39 <mfedosin> nikhil_k, yes, sure. in etherpad?
14:21:46 <nikhil_k> mfedosin: yes, please
14:21:46 <ativelkov> nikhil_k: all of them are bugs in the experimental part of the API
14:22:00 <nikhil_k> ativelkov: I see, that should be fine. mostly
14:22:22 <ativelkov> Some still has the APIImpact flags though, but that are actually bugs, not the "new features" or changes
14:22:58 <nikhil_k> excellent. thank you.
14:24:06 <mfedosin> I'm not sure that we have to merge all of them, but it's better to close as much bugs as we can
14:26:36 <nikhil_k> mfedosin: it would be nice to know if murano team is going to be affected by those blockers severely
14:26:55 <nikhil_k> it would be bad to break things for them when we could have avoided!
14:27:14 <nikhil_k> so, please let me know if there are specific patches to look out for in RC
14:27:25 <nikhil_k> and I will try to review and get reviews
14:27:31 <nikhil_k> ativelkov: ^
14:28:06 <ativelkov> nikhil_k: the first 6 patches are realu blockers for Murano
14:28:11 <ativelkov> really*
14:28:47 * sigmavirus24 wonders the value of using "ApiImpact" for those changes
14:28:59 <mfedosin> and we reviewed them carefully
14:29:31 <ativelkov> sigmavirus24: that's what I was asked to do even for experimental part of the API
14:29:33 <nikhil_k> sigmavirus24: how bad is it?
14:29:48 <nikhil_k> sigmavirus24: may be not even put that flag there?
14:29:49 <sigmavirus24> ativelkov: wasn't saying you were doing anything wrong
14:30:03 <nikhil_k> I am trying to think who all are interested in that flag
14:30:14 <sigmavirus24> nikhil_k: so that tends to indicate an impact on a stable API first of all, second of all, that is supposed to attract API WG member attention
14:30:30 <nikhil_k> I see
14:30:31 <sigmavirus24> Third, when combined with a DocImpact flag that files a bug on the api-site repo I think
14:30:47 <nikhil_k> to me it was a tool for grepping through commits and figure out how API evolved
14:30:51 <sigmavirus24> And if that has documentation for artifacts I'd be a little surprised given how volatile the API is
14:31:08 <ativelkov> well, for example, the first patch puts the output of list artifacts from just a json array into a proper object
14:31:10 <sigmavirus24> nikhil_k: maybe a different tag then that's specific to that case then?
14:31:27 <nikhil_k> sigmavirus24: I see, that's a good point.
14:31:34 <sigmavirus24> ativelkov: sure, I'm not saying this is wrong, just that perhaps we should use a different flag because we could be spamming other teams in openstack
14:31:35 <nikhil_k> we should chat on -glance
14:31:43 <nikhil_k> we are out of time here
14:31:50 <nikhil_k> thanks all for joining!
14:31:54 <ativelkov> Thanks!
14:31:54 <mfedosin> thanks!
14:31:55 <nikhil_k> #endmeeting