14:00:01 <nikhil_k> #startmeeting glance drivers
14:00:01 <openstack> Meeting started Tue Sep  1 14:00:01 2015 UTC and is due to finish in 60 minutes.  The chair is nikhil_k. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:02 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:05 <openstack> The meeting name has been set to 'glance_drivers'
14:00:23 <nikhil_k> #topic Open Discussion
14:00:28 <nikhil_k> There's not ageda for today/
14:00:43 <flaper87> right, I think we've enough to work on till M dev cycle opens
14:00:45 <nikhil_k> If people are around we can discuss l-3 and rc-1 stuff. or anything else
14:00:57 <nikhil_k> flaper87: yeah
14:01:04 <flaper87> There  are still specs to review but I'd delay those in favor of reviews on implementations
14:01:20 <flaper87> that is to say, I'm happy to discuss l-3 rc-1 stuff is other folks are around
14:01:20 <nikhil_k> sounds like a good plan
14:01:21 <flaper87> :)
14:02:09 <rosmaita> o/
14:02:16 <nikhil_k> Reviews to be done for L3 in the order of priority are:
14:02:20 <nikhil_k> #link https://etherpad.openstack.org/p/glance-liberty-3-reviews
14:02:29 <nikhil_k> We may be missing some already
14:03:39 <nikhil_k> I added a note for
14:03:40 <jokke_> I have fill for your 30min if you wish :P
14:03:40 <nikhil_k> https://review.openstack.org/#/c/196240
14:04:18 <nikhil_k> Also, it might be worth doing this outside of the release cycle
14:04:18 <nikhil_k> https://review.openstack.org/#/c/120866
14:04:19 <jokke_> I'd like to get drivers opinion for increasing trend of feature requests masked as bugs
14:04:29 <nikhil_k> jokke_: ok, please go ahead.
14:04:36 <jokke_> ^^
14:04:47 <flaper87> jokke_: go crazy
14:05:31 <jokke_> well ... there seems to be increasing trend and I'm not speaking only Stuart to think that our Images Api v1 and v2 behaving differently would be a bug
14:05:40 <jokke_> and the functionality sync is ok by bugfixes
14:06:46 * sigmavirus24 is here sorry
14:06:58 <jokke_> and that seems to trending on both sides client and API
14:07:06 <jokke_> %s:/API/server/
14:07:42 <nikhil_k> so, is this something you want to discuss?
14:07:43 <nikhil_k> https://bugs.launchpad.net/python-glanceclient/+bug/1489941
14:07:44 <openstack> Launchpad bug 1489941 in Glance "v2 client sorts on name in wrong direction" [Undecided,Triaged]
14:08:42 <jokke_> well that's backwards incompatibility between the APIs, definitely not client bug
14:08:45 <sigmavirus24> jokke_: so the complaint is that the API sorts in a backwards incompatible way in v2?
14:09:35 <jokke_> this is good example of feature request masked as bug https://bugs.launchpad.net/python-glanceclient/+bug/1489534
14:09:37 <openstack> Launchpad bug 1489534 in python-glanceclient "Move metadef commands to sub-commands" [Wishlist,In progress] - Assigned to Niall Bunting (niall-bunting)
14:09:49 <nikhil_k> It would be nice if mfedosin was here. Because, this was a openstack-spec that was implemented by him earlier. Seems worthy enough change to have bigger impact.
14:10:08 <mfedosin> hi :)
14:10:25 <nikhil_k> mfedosin: bug 1489941
14:10:26 <openstack> bug 1489941 in Glance "v2 client sorts on name in wrong direction" [Undecided,Triaged] https://launchpad.net/bugs/1489941
14:11:16 <mfedosin> looking
14:11:19 <nikhil_k> I may need to visit the openstack-spec for this bug
14:11:31 <flaper87> I always tend to assume good faith and I believe cases where features were proposed as bug are just mistakes that can easily be fixed by redirecting folks to the right process
14:12:02 <nikhil_k> flaper87: that's a good way to put it
14:12:15 <jokke_> https://bugs.launchpad.net/python-glanceclient/+bug/1489543 this is another
14:12:16 <openstack> Launchpad bug 1489543 in python-glanceclient "Different fields shown in v2 list than in v1 list" [Wishlist,In progress] - Assigned to SamP (sampath-priyankara)
14:12:24 <flaper87> To be fair, the 1.0 release of the client caused lots of panic and raised many concerns, I can see why that metadef thing was created as a bug and not a spec
14:12:59 <nikhil_k> creating a bug is ok as long as it is visible enough for others too
14:13:01 <flaper87> So, I'd say: If things like this happen, please, lets notify more than 1 folk to spread the case among us and redirect folks to the right process
14:13:38 <nikhil_k> #info I think a spec has capacilty to link bugs so bug accompanied with a spec is the right way for the metadef change
14:13:47 <jokke_> I've noticed during reviews this being raising trend
14:14:17 <flaper87> jokke_: It's great you brought this up because we should all be very careful when reviewing
14:14:19 <flaper87> so, thanks :D
14:14:28 <jokke_> it seems that people assume getting new functionality in easier if there is "Closes-Bug: #XYXYXYXY" in the commit message
14:14:30 <flaper87> I had noticed a couple but I didn't notice the trend
14:14:50 <flaper87> again, I'd assume good faith and probably the process is not clear enough
14:14:52 <flaper87> :)
14:14:58 <nikhil_k> jokke_: this is good awareness!
14:15:02 <flaper87> or the feature did seem a bug to them
14:15:03 <jokke_> so just wanted to hear what's the expectation here from your side :)
14:15:20 <flaper87> jokke_: redirect them to the spec process with nice words and examples
14:15:22 <flaper87> :D
14:16:06 <nikhil_k> It should be case by case basis ideally. However, the best approach is to figure out the impact ratio of openstack-wide/project-only
14:16:21 <nikhil_k> if >1 then spec, if <1 then bug
14:16:24 <jokke_> something nice and polite like "-2; if you want to continue fighting about this, file a spec, there is few examples in the repo"? :P
14:17:08 <nikhil_k> I think API changes, VMT related changes, usability changes come in >1 zone
14:17:10 <flaper87> jokke_: more like: "-2, I think this change is spec worthy. It'd be cool if you could propose one so that we can discuss the impact further"
14:17:24 <jokke_> :)
14:17:25 * flaper87 teaches jokke_ the art of sugarcoating
14:17:48 <nikhil_k> That bring a good point;
14:17:51 <rosmaita> flaper87: i didn't realize you were so diplomatic
14:18:00 <flaper87> rosmaita: only with new devs
14:18:02 <flaper87> :P
14:18:04 <nikhil_k> should we open a new folder called spec-lite ?
14:18:23 <flaper87> rosmaita: but things go south very quickly for me
14:18:25 <jokke_> I was more thinking single file rather than folder
14:18:26 <flaper87> hahahahaha
14:18:39 <jokke_> just list of minor changes people would like to do
14:18:52 <nikhil_k> jokke_: we need to follow documentation semantics
14:18:59 <flaper87> nikhil_k: sounds good
14:19:05 <nikhil_k> so that they show up on specs.openstack.org
14:19:21 <jokke_> nikhil_k: and? even if it's one updating file we can still do that
14:19:24 <flaper87> btw, we had agreed on always using the backlog folder for approved specs and them move them to the milestone
14:19:28 <flaper87> we never got to do that
14:19:29 <nikhil_k> eh
14:19:35 <jokke_> correct leve sub topic with couple of bullet points what and why
14:19:36 <flaper87> Can we start doing it?
14:19:43 <flaper87> I think it's clearer and better
14:19:52 <nikhil_k> jokke_: you are hinting at rel-notes style
14:20:04 <jokke_> nikhil_k: pretty much
14:20:06 <nikhil_k> which might be good option
14:20:12 <nikhil_k> if we define the time period
14:20:33 <jokke_> that would be easy to follow and you would not saturate the spec portal with million links to two line coduments
14:20:38 <nikhil_k> but then I am in this dilemma of defining one for client and store
14:20:53 <jokke_> I mean one file per spec folder
14:21:07 <nikhil_k> I see
14:21:18 <jokke_> so maybe late for liberty but something like mitaka/minor_changes.rst
14:21:34 <flaper87> mitaka ++
14:21:55 <nikhil_k> flaper87: Let's discuss a plan for M in next meeting for backlogs, priorities for next 6 months, pre-approved for M, possible for M ?
14:22:01 <jokke_> nikhil_k: separating server, client and store sounds perfectly reasonable
14:22:05 <flaper87> nikhil_k: sounds perfect
14:22:28 <nikhil_k> jokke_: cool!
14:22:41 <nikhil_k> was thinking that would make a lot of sense
14:22:57 <jokke_> yeii ... now I can go and get coffee and bit fresh air ... thanks guys
14:22:57 <flaper87> sparating the folders but keeping them in glance-specs, right?
14:22:58 <nikhil_k> jokke_: we can avoid a few impact flags in this case
14:23:03 <jokke_> ++
14:23:11 <nikhil_k> flaper87: yeah
14:23:13 <flaper87> ok
14:23:17 <flaper87> yeah, that makes total sense
14:23:57 <jokke_> maybe Thu weekly meeting with wider audience
14:23:58 <jokke_> ?
14:24:30 <flaper87> this sounds drivers specific to me
14:24:34 <flaper87> but I'm ok with whatever
14:24:51 <flaper87> I mean, everyone is welcome to join this meeting
14:25:01 <nikhil_k> jokke_: may be next tuesday and thursday. I think we want to discuss l-3 specific things this week
14:25:04 <flaper87> I meant to say, I would prefer not to use glance's meeting time for this
14:25:10 <nikhil_k> ?
14:25:12 <flaper87> but if there are no other topics, I'm cool
14:25:24 <jokke_> nikhil_k: like the freeze/release panic hour at thu? :P
14:25:40 <nikhil_k> jokke_: haha, you caught the right nerve
14:26:09 <nikhil_k> so that people find a time and spot to show up for things
14:26:21 <jokke_> I'm fine with that ... as long as we do not exclude ... I'm just blunt enough to invade your meeting :P
14:26:37 <nikhil_k> haha, it's a openstack meeting :P
14:26:49 <flaper87> it's still an os meeting
14:26:50 <nikhil_k> just bring awareness for what's coming!
14:26:54 <flaper87> no one should feel excluded
14:27:00 <flaper87> and it's freaking IRC
14:27:02 <nikhil_k> bringing*
14:27:06 <nikhil_k> :)
14:27:16 <nikhil_k> flaper87: that'd be status for today
14:27:22 <flaper87> :D
14:28:36 <nikhil_k> ok guys; if nothing else, have a nie day!
14:28:49 <nikhil_k> Thanks for the good conversation!
14:29:22 <nikhil_k> #endmeeting