14:00:13 <dhellmann> #startmeeting oslo
14:00:14 <openstack> Meeting started Fri Oct 25 14:00:13 2013 UTC and is due to finish in 60 minutes.  The chair is dhellmann. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:15 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:17 <openstack> The meeting name has been set to 'oslo'
14:00:25 <dhellmann> #link https://wiki.openstack.org/wiki/Meetings/Oslo
14:00:30 <zyluo> hi
14:00:31 <dhellmann> Show of hands, please?
14:00:33 <dhellmann> o/
14:00:36 <bpokorny> o/
14:00:37 <zyluo> o/
14:00:46 <jswarren> o/
14:00:51 <dstanek> o/
14:00:54 <luisg> o/
14:01:09 <dhellmann> we have 2 topics today
14:01:23 <dhellmann> I was going to start with the one I think will be the shorter discussion
14:01:27 <dhellmann> #topic deprecated decorator (dstanek)
14:01:33 <dhellmann> #link https://review.openstack.org/#/c/50486/
14:01:43 <dhellmann> dstanek, you have the floor :-)
14:01:58 <dstanek> dhellmann: thank you
14:02:36 <dstanek> this review allows a callable to be marked as deprecated and logs a message saying when it was deprecated and when it may be removed
14:02:56 <dstanek> right now it is in keystone and we need it in python-keystoneclient
14:03:18 <dhellmann> it seems like it would be useful in all of the clients
14:03:26 <dhellmann> how is it used within keystone?
14:03:32 <dstanek> a few people outside of keystone have expressed interest in it and i would like to get it into oslo
14:03:55 <dstanek> right now i have a review that deprecates the v2 API
14:04:16 <dstanek> i have to rebase it off of the merged impl, which i'm working on right now
14:04:23 <dhellmann> ok
14:04:44 <dhellmann> this logs the messages, right? so callers of the API wouldn't see them?
14:04:49 <dstanek> i believe that we'll be using it shortly for other functionality that is marked as deprecated only in code comments
14:05:23 <dhellmann> makes sense
14:05:42 <dhellmann> I don't have any objections to bringing it in -- does anyone have any comments?
14:05:49 <dstanek> what are the steps for me to get it in?  is it as simple as code review to debate or is there more to it?
14:06:21 <dhellmann> the usual path is to add it as-is from the upstream repo, then we fiddle with the api and implementation until everyone is happy
14:06:39 <dhellmann> you'll need a better module name than "utils" in this case
14:06:44 <dhellmann> :-)
14:06:55 <dstanek> yes :-)
14:07:09 <dstanek> what do you suggest? deprecated.py?
14:07:39 <dhellmann> we already have a deprecated subdirectory with code we do not want used any more, but that we're not ready to remove
14:07:41 <dstanek> there is other stuff in our utils module that definitely wouldn't go into oslo
14:07:54 <dhellmann> what about adding it to the versionutils module?
14:07:54 <zyluo> it could be utils in the "deprecated" package?
14:08:13 <dhellmann> zyluo: well, we don't want code from that package copied out of the incubator any more
14:08:22 <zyluo> dhellmann, ic
14:08:23 <dhellmann> we kept it around in case we have security bugs or something that we need to fix
14:08:52 <zyluo> putting it in versionutils.py sounds good
14:08:58 <dstanek> i'm ok with versionutils
14:09:42 <dhellmann> ok, be prepared in case someone finds a better name when the code review goes up, but that seems like a good place to start
14:09:48 <mrodden> i thought we had something similar in log.py...
14:10:34 <dhellmann> it looks like ContextAdapter has a deprecated() log call
14:10:47 <dstanek> mrodden: there is a deprecated method...that's what i use to actually do the logging
14:10:49 <dhellmann> maybe this new decorator should use that
14:10:54 <dhellmann> well, there we go :-)
14:11:12 <dstanek> the only thing i don't like about it is the exception it raises in fatal more
14:11:33 <dstanek> it's related to deprecating config options or something like that
14:11:51 <mrodden> its controlled by a falg
14:11:52 <mrodden> flag*
14:11:57 <dhellmann> maybe we can adjust that wording
14:12:20 <dstanek> yeah, it's a DeprecatedConfig exception
14:12:26 <mrodden> yeah
14:12:29 <mrodden> just noticed that
14:12:46 <dhellmann> ok, those are details we can work out
14:12:48 <markmc> gah, sorry I'm late
14:12:52 <dstanek> i wouldn't mind renaming that to DeprecatedError and adjusting the message
14:12:58 <dhellmann> hi, markmc
14:13:10 <dhellmann> dstanek: sounds good, maybe as a follow-up changeset?
14:13:20 <dstanek> sure thing
14:13:24 <dhellmann> cool
14:13:40 <dhellmann> anything else on the new deprecated decorator?
14:14:12 <dhellmann> right, moving on then
14:14:13 <dhellmann> #topic Solidify detailed approach to lazy translation, based on 10/21 IRC chat (bpokorny)
14:14:17 <dhellmann> #link https://blueprints.launchpad.net/oslo/+spec/i18n-messages
14:14:23 <dhellmann> What details are left to resolve?
14:14:39 <dhellmann> bpokorny, you're up
14:14:47 <bpokorny> Thanks, dhellmann.
14:15:27 <bpokorny> I think we have a good understanding of the details on where to go with this one now, based on a chat with a few of you.
14:15:41 <bpokorny> So mrodden added the details we talked about to the blueprint.
14:16:05 <bpokorny> My intention with putting it on the agenda was just to see if anyone has any comments on those details at this point.
14:16:37 <bpokorny> I think mrodden captured the things we discussed previously.
14:16:43 <jswarren> The item about __repr__... is that just for debugging purposes?
14:16:44 * beekneemech shows up 15 minutes late
14:16:50 <dhellmann> that list looks accurate to me
14:16:58 <dhellmann> jswarren: yes, I think so
14:17:26 <jswarren> OK.
14:18:04 <bpokorny> If no concerns at this point on them, we'll get an initial code patch sent up for review.
14:18:26 <dhellmann> I feel pretty good about this approach. It will be interesting to see what happens when str() and unicode() "break", though.
14:18:44 <dhellmann> but let's see, instead of trying to guess :-)
14:18:56 <bpokorny> Yes :)  There could be some interesting things to work through as we work the patch.
14:19:26 <jswarren> I wonder which project should be the first victim...er, I mean candidate.
14:19:39 <dhellmann> probably something small
14:20:06 <dhellmann> I'm not sure how much we're using translation in ceilometer, but that code base is fairly small so it might be a good candidate
14:20:38 <dhellmann> then we'll want a project using the old wsgi framework, too
14:21:11 <jswarren> glance is probably the worst-case scenario.
14:21:14 <bpokorny> Maybe Ceilometer and Cinder to start with?
14:21:24 <dhellmann> jswarren: why is glance "worst-case"?
14:21:24 <luisg> heat had some interesting challenges when doing it the 1st time around
14:21:51 <jswarren> Adding lazy translation to it has been the most disruptive of all the projects, from what I can tell.
14:21:59 <dhellmann> hmm
14:22:13 <jswarren> There's wsgi and json concerns.
14:23:17 <jswarren> heat might be just as bad...dunno
14:23:24 <dhellmann> we might want to tackle glance fairly early, if it will take a while, but we should start with something easier
14:23:44 <dhellmann> I like bpokorny
14:23:46 <dhellmann> oops
14:23:53 <dhellmann> I like bpokorny's suggestion :-)
14:23:56 <luisg> :)
14:24:06 * bpokorny blushes
14:24:08 <bpokorny> Haha
14:24:10 <jswarren> He's quite likable.
14:24:13 <luisg> haha
14:24:33 <jswarren> likeable
14:24:49 <jswarren> Works for me, FWIW
14:24:59 <bpokorny> Sounds good.  We'll plan to start with Ceilometer and Cinder once we get the oslo patch up.
14:25:02 * dhellmann blames all of today's typos on cold fingers
14:25:10 <bpokorny> :)
14:25:42 <dhellmann> bpokorny: after we have some code in the incubator, we should give those other teams a heads-up so the patches in their repo won't be a surprise
14:26:05 <bpokorny> Yep, I can do that.
14:26:19 <dhellmann> good
14:26:20 <mrodden> wonder if we need BPs for other projects?
14:26:23 <bpokorny> Would the best forum to be just post it on IRC, or get it on the meeting agenda?
14:26:29 <jd__> o/
14:26:39 <dhellmann> I think starting with the mailing list is ok
14:26:45 <bpokorny> ok
14:26:47 <dhellmann> and we can reference the blueprint we have for oslo
14:27:01 <dhellmann> unless the ptls of those projects want us to create separate blueprints for their own tracking
14:27:33 <dhellmann> does anyone have anything else on this topic?
14:28:10 <dhellmann> #topic open discussion
14:28:58 <dhellmann> I've emailed all of the maintainers asking for status about "their" parts of the incubator, especially whether code is ready to graduate
14:29:09 <dhellmann> I haven't heard back from anyone yet :-(
14:30:19 * zyluo will update it next week
14:30:24 <dhellmann> if anyone else has opinions on that, drop me a note so I can include your comments in the wrap-up summit session
14:30:27 <dhellmann> oh, great, thanks zyluo
14:30:32 <dhellmann> :-)
14:31:10 <mrodden> i noticed there is an i18n session now too
14:31:39 <dhellmann> #link http://summit.openstack.org/cfp/details/186
14:31:40 <dhellmann> yes
14:32:33 <dhellmann> I have one or two more to clarify and then I will be publishing the first version of the schedule today
14:32:38 * dhellmann crosses fingers
14:32:51 <jd__> dhellmann: I saw that many people wrote something on the etherpad (me included)
14:33:04 <dhellmann> jd__: ah, good
14:33:05 <jd__> for some value of many
14:33:43 <dhellmann> haha
14:33:46 <dhellmann> #link https://etherpad.openstack.org/p/icehouse-oslo-status
14:34:15 <dhellmann> oh, yes, there's lots more there than last time I looked
14:35:05 <dhellmann> if no one has anything else, we can wrap up a little early
14:36:03 <dhellmann> ok, I guess that's it
14:36:05 <dhellmann> thanks everyone!
14:36:13 <mrodden> thank!
14:36:15 <bpokorny> Thanks all!
14:36:16 <mrodden> thanks!*
14:36:26 <dhellmann> #endmeeting