21:00:20 <notmyname> #startmeeting swift
21:00:22 <openstack> Meeting started Wed Jun 15 21:00:20 2016 UTC and is due to finish in 60 minutes.  The chair is notmyname. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:00:23 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:00:25 <openstack> The meeting name has been set to 'swift'
21:00:26 <notmyname> who's here for the swift team meeting?
21:00:31 <cschwede> o/
21:00:36 <m_kazuhiro> o/
21:00:36 <jrichli> .
21:00:37 <cutforth> o/
21:00:39 <acoles> here
21:00:39 <dmorita> hi
21:00:50 <tdasilva> hello
21:00:50 <hosanai> o/
21:00:53 <kmARC> o/
21:00:53 <timburke> o/
21:01:00 <mathiasb> o/
21:01:25 <kota_> hello
21:01:28 <notmyname> hello, everyone
21:01:40 <notmyname> not a ton on the agenda today
21:01:42 <notmyname> #link https://wiki.openstack.org/wiki/Meetings/Swift
21:01:51 <notmyname> but let's get started
21:02:08 <torgomatic> .
21:02:13 <notmyname> #topic rolling upgrade / translation bug
21:02:27 <notmyname> cschwede: quick status check on these. any change from last week?
21:02:35 <cschwede> nothing new…
21:02:42 <notmyname> ok
21:02:52 <notmyname> #topic hackathon
21:02:54 <acoles> cschwede: sorry I think I need to review your gettext patch again
21:02:57 <notmyname> #link https://www.eventbrite.com/e/openstack-swift-july-2016-hackathon-tickets-25913773795
21:03:15 <notmyname> please sign up for the hackathon and get a hotel room soon, if you're planning on coming
21:03:18 <cschwede> acoles: no worries, crypto is  more important right now
21:03:58 <ntata> hello
21:04:01 <notmyname> #topic crypto status
21:04:18 <notmyname> crypto-review is up and ready to be reviewed. I've updated http://not.mn/swift/swift_community_dashboard.html to reflect that :-)
21:04:54 <tdasilva> lol
21:04:57 <kota_> lol
21:05:12 <kota_> that's only crypto reviews
21:05:13 <notmyname> soft freeze on master while this is being reviewed. we'll reevaluate the freeze at the end of next week, based on where we are with the reviews
21:05:25 <notmyname> kota_: exactly! that's what you should be doing ;-)
21:05:27 <jrichli> that's great!
21:05:29 <acoles> I pushed a new version of all the patches about an hour ago
21:05:39 <notmyname> acoles: thank you for managing that
21:05:41 <acoles> notmyname: kota_ IS doing it
21:05:49 <notmyname> acoles: I know. he's great!
21:06:01 <jrichli> yes, thanks kota_!
21:06:10 <kota_> ;-)
21:06:25 <acoles> I am trying to leave a reply to every comment, so as to track them, but if I miss one please don't take it personally, just leave the comment again
21:06:33 <notmyname> everyone who reviews the crypto-review branch is great. so if you want to be great too, get to reviewing! ;-)
21:06:40 <acoles> Think I worked through approx 100 comments in last 24 hours!
21:06:50 <notmyname> acoles: are the logistics working for you so far? is it an ok process?
21:07:38 <acoles> notmyname: kindof. I have burnt myself with rebases once or twice :/ but that's me being dumb/tired
21:07:39 <kota_> acoles: you're great
21:07:52 <tdasilva> kota_: +1000
21:08:00 <notmyname> acoles: anything I can do or others can to do help out with that side of things?
21:08:19 <notmyname> (or anything we can stop doing that's causing you pain?)
21:08:20 <jrichli> yes, acoles is awesome
21:08:21 <acoles> there is a risk that I push new patches and miss recent comments on the last version (that happened to torgomatics comments this evening)
21:08:47 <acoles> I have a lag because I am running tests on every patch in chain before pushing
21:09:01 <acoles> so again, please just leave the same comment or point out to me to g olook at an older version
21:09:09 <notmyname> ok
21:09:22 <timburke> ...which, from all those Verified+1 emails just now, seems to be paying off
21:09:38 <notmyname> how's everyone else doing with the crypto reviews? any issues, questions, or otherwise?
21:10:10 <torgomatic> sort of wondering why we have a Putter and a MIMEPutter isntead of just one thing, but I'm working onit
21:10:24 <cschwede> i think the current work is already awesome, just digging into the details right now
21:11:14 <acoles> torgomatic: we definitely don't have a BasePutter ;)
21:11:20 <notmyname> some of the stuff is somewhat opaque to me right now, since I'm not a crypto expert at all (eg I've been looking at the iv/key derivation). hard for me to know if that's good or now
21:11:22 <notmyname> *not
21:13:22 <notmyname> any suggestions on how to tackle the crypto bits? anything better than read the code and think hard?
21:13:37 <acoles> torgomatic: IIRC I think maybe all replicated policy requests *could* just use a MimePutter but I only made those needing footers use a MimePutter ?
21:14:04 <cschwede> i’m neither a crypto expert, but so far all the crypto stuff makes a lot of sense to me - I read a few more papers to lock into details and it looks good to me so far
21:14:14 <jrichli> notmyname: I am open to provide an info I have - just let me know what would be best
21:14:16 <kmARC> jrichli: correct me if I'm wrong but te crypto-related parts have already been reviewed by cca.
21:14:18 <acoles> maybe we can get some of jrichli's colleagues in Zurich to review?
21:14:26 <notmyname> cschwede: oh, so I'll add in "go read academic papers" ;-)
21:14:30 <kmARC> we probably can get him to have another look on it
21:15:02 <jrichli> kmARC: yes, it would be a good idea to have cca take a look now.  some things may have changed a little
21:15:38 <acoles> jrichli: actually did you let cca know that he is named as author on the patch, so he may well want to review it ;)
21:15:41 <kota_> cschwede: oh, if you have recommended papers, they're great to catch up the detail.
21:15:51 <acoles> cca == C Cachin right?
21:16:08 <kota_> (or wrote down into docs at review-6?)
21:16:20 <jrichli> acoles: that's right!  and yes, that's right.
21:16:21 <cschwede> kota_: i’ll have a look and try to provide a few links until Friday
21:16:39 <kmARC> OK. Tomorrow I'll talk to him, hopefully by the time you guys wake up he'd have had a look :)
21:16:39 <kota_> s/wrote/will be written)
21:16:39 <kota_> cschwede: thanks!
21:16:53 <notmyname> kmARC: thanks
21:17:12 <notmyname> anything else from anyone about the crypto-review branch (code or process)?
21:17:16 <acoles> kmARC: thanks
21:18:26 <notmyname> ok, let's move on
21:18:39 <notmyname> #topic docs added to our repo
21:18:55 <notmyname> annegentle: you added this topic about https://review.openstack.org/#/c/312315/
21:19:06 <annegentle> o/ holla
21:19:25 <notmyname> and also https://review.openstack.org/#/c/330070/ was just added
21:19:30 <notmyname> s/added/proposed/
21:19:34 <annegentle> responding to Donaugh's questions on the review now, and wanted to bring this forward at a meeting to be sure I can answer questions
21:19:53 <notmyname> seems that there's been some good conversation in gerrit
21:19:59 <annegentle> I put a few more links in the commit message to provide additional context, good questions
21:20:49 <notmyname> in general, I like the idea of docs about a project being owned/managed by that project
21:21:05 <annegentle> yes, swift has been culturally that way for ages
21:21:24 <notmyname> but some of the logistical stuff is what I'm curious about (where stuff gets published and when, how to manage/update, etc). the stuff asked in gerrit
21:22:03 <annegentle> the job itself will build to developer.openstack/api-ref/object-storage/
21:22:21 <annegentle> then we'll need an .htaccess redirect 301 from http://developer.openstack.org/api-ref-objectstorage-v1.html
21:22:35 <annegentle> other teams are working out kinks before redirecting
21:22:47 <notmyname> ok
21:22:56 <notmyname> I supposed that's similar to the install guide too?
21:23:20 <annegentle> We're also working on how the navigation works for multiple APIs, http://lists.openstack.org/pipermail/openstack-dev/2016-June/096972.html
21:23:51 <annegentle> notmyname: yeah, ideally we don't give a worse document after redirecting.
21:24:26 <annegentle> and ultimately the theme will better match the rest of the docs like https://api.os.gra.ham.ie/compute/ which we are still reviewing
21:24:38 <annegentle> #info REST API docs moving in-tree to project repos
21:25:31 <annegentle> #info Once https://review.openstack.org/#/c/312315/ lands with its dependent build patch, API reference info is published to developer.openstack/api-ref/object-storage/
21:25:59 <annegentle> #info Working on new styling and nav, refer to  https://api.os.gra.ham.ie/compute/ and http://lists.openstack.org/pipermail/openstack-dev/2016-June/096972.html
21:26:03 <notmyname> annegentle: thanks for answering our questions. FYI responses might be slower until we get the crypto stuff merged
21:26:07 <annegentle> I think that covers the questions.
21:26:09 <acoles> is there a plan to add a gate job for the new tox api-ref env? If so can we get a gate job in zuul *first* using a placeholder api-ref env then these patches will pass/fail that in the gate?
21:26:30 <annegentle> acoles: yep, need to update with Depends-on: https://review.openstack.org/#/c/313015/
21:26:57 <annegentle> acoles: based on reviews in https://review.openstack.org/#/c/313015/ I guess they need the content to land first?
21:27:02 <annegentle> acoles: perhaps you can work that out with infra
21:28:01 <notmyname> well, maybe someone can investigate that after crypto. not before
21:28:02 <acoles> annegentle: is that a non-voting job? sorry, yaml i snot my first language.
21:28:14 <acoles> is not*
21:28:26 <annegentle> acoles: I believe these are non-voting, yes.
21:29:56 <notmyname> if non-voting, then if we land it first, we should see it pass with the proposed docs patch
21:29:57 <annegentle> acoles: though I'm not completely clear on whether it can be reviewed on say docs-draft
21:30:05 <notmyname> and the initial failures wouldn't matter
21:30:07 <acoles> annegentle: well if infra are waiting for the swift patch then we can't have DependsOn in the swift patch
21:30:40 <notmyname> annegentle: what's docs-draft?
21:30:42 <annegentle> acoles: Andreas (Ajaeger) said "This is waiting for the swift patch to merge first - and once that is in, this needs another +A."
21:30:56 <annegentle> notmyname: for example http://docs-draft.openstack.org/15/312315/8/check/gate-swift-docs/e90d607//doc/build/html/
21:31:21 <annegentle> notmyname: which I guess is a gate job but teams set it to non-voting if they want it to
21:31:52 <notmyname> since your patch sets up a new doc tree, it's not related to what's created in that docs-daft version
21:32:17 <acoles> annegentle: right, when you said " need to update with Depends-on: https://review.openstack.org/#/c/313015/" I assumed you meant update the swift patch, but that would prevent it merging before 313015 gets a +A (I think?)
21:32:23 <notmyname> so either they could be unified (same tree and/or same conf.py) or a new gate job defined for it
21:32:24 <annegentle> notmyname: sure, what I mean is that there's no "draft" build right now
21:33:13 <notmyname> acoles: you're suggesting swapping the dependency, right? so land the -infra patch frist, then the swift patch. that way we have a gate job before the docs
21:33:23 <annegentle> notmyname: these have new jobs defined in https://review.openstack.org/#/c/313015/ -- what I don't know is if the template contains gate jobs
21:33:35 <annegentle> acoles: I'd like that too
21:33:36 <acoles> notmyname: that's what I'd prefer
21:33:52 <annegentle> acoles: notmyname: I'll follow up with Andreas to see if we can get that.
21:34:00 <annegentle> anything else?
21:34:11 <notmyname> acoles: yeah, that sounds completely reasonable. we already know that what was originally proposed didn't build, so it's good to catch that in the CI job
21:34:15 <notmyname> annegentle: thanks
21:34:26 <acoles> notmyname: exactly ;)
21:34:53 <annegentle> thanks
21:35:02 <notmyname> #topic open discussion
21:35:09 <notmyname> anything else from anyone this week?
21:36:44 <notmyname> ok, great!
21:36:57 <notmyname> thanks everyone for coming. thanks for working on swift
21:37:01 <notmyname> #endmeeting