On 08 Oct 2014, at 21:48 , ASF IRC Bot <[email protected]> wrote:
> Summary of IRC Meeting in #couchdb-meeting at Wed Oct 8 19:12:36 2014: > > Attendees: Wohali, dch, Kxepal, sagelywizard, jan____, chewbranca > > - Preface > - CouchDB 2.0 > - Fauxton > - open floor > - Action: jan____ to cut a 1.x.x branch > - Action: kxepal to propose a 1.7 release that includes deprecations for > what is already in master (and expected to hit 2.0) > - Action: kxepal to take lead on fixing broken eunit stuff > - Action: sagelywizard to get closure on his thread and commit view sequence > work as experimental > - Action: wohali to send email with suggested last steps before we move > towards a 2.0 release > - Action: jan____ to help with s/futon/fauxton + Action: jan___ to move one-click-cluster setup discussion forward. > > > IRC log follows: > > ## Preface ## > [Wed Oct 8 19:12:38 2014] <Wohali>: i missed that > [Wed Oct 8 19:12:38 2014] <Wohali>: doh > ## CouchDB 2.0 ## > [Wed Oct 8 19:12:45 2014] <Wohali>: so things are pretty settled on master > at this point > [Wed Oct 8 19:12:48 2014] <Wohali>: aside from the view changes code that > sagelywizard__ has been working on > [Wed Oct 8 19:12:51 2014] <Wohali>: (which I am inclined to have committed > now, as experimental) > [Wed Oct 8 19:13:01 2014] <Wohali>: we should start moving towards a release > IMO > [Wed Oct 8 19:13:05 2014] <Wohali>: and working out the last few things that > need fixing > [Wed Oct 8 19:13:08 2014] <Wohali>: to my mind that is: > [Wed Oct 8 19:13:19 2014] <Wohali>: * Cluster setup, it's ugly, we need a > better solution > [Wed Oct 8 19:13:29 2014] <Wohali>: * s/futon/Fauxton and remove futon from > the build > [Wed Oct 8 19:13:33 2014] <Wohali>: * docs > [Wed Oct 8 19:13:47 2014] <Wohali>: comments? > [Wed Oct 8 19:13:57 2014] <Kxepal>: I'd like to add some cleanup in at least > config values > [Wed Oct 8 19:14:16 2014] <Wohali>: is there a ticket/short description you > can provide? > [Wed Oct 8 19:14:25 2014] <jan____>: I’m happy to lead on the cluster setup > thing, as per my proposal recently > [Wed Oct 8 19:14:26 2014] <Kxepal>: + fixing eunit and doing more testing > [Wed Oct 8 19:14:38 2014] <Wohali>: i thought testing was working > [Wed Oct 8 19:14:43 2014] <Wohali>: if it's broken obviously not > [Wed Oct 8 19:14:54 2014] <Wohali>: jan____: good, i knoew rnewson wanted to > talk about that with you > [Wed Oct 8 19:14:55 2014] <Kxepal>: there are some suites are disabled for > now > [Wed Oct 8 19:15:06 2014] <Wohali>: Kxepal: do you feel that is a release > blocker? > [Wed Oct 8 19:15:20 2014] <Kxepal>: Wohali: I do since we aren't sure in the > behaviour. > [Wed Oct 8 19:15:31 2014] <Wohali>: okay. are you volunteering? :) > [Wed Oct 8 19:15:34 2014] <Kxepal>: yes > [Wed Oct 8 19:15:43 2014] <Wohali>: hooray > [Wed Oct 8 19:15:54 2014] <Wohali>: are there any Fauxtoneers here? > [Wed Oct 8 19:15:58 2014] <Wohali>: i know robertkowalski is busy today > [Wed Oct 8 19:16:11 2014] <Wohali>: not sure if garren is free > [Wed Oct 8 19:16:39 2014] <Wohali>: but putting fauxton default in the build > and ripping out futon is long overdue > [Wed Oct 8 19:16:59 2014] <Wohali>: and i think we solved the where does > jquery.couch.js live problem too > [Wed Oct 8 19:17:13 2014] <Kxepal>: yes, we did > [Wed Oct 8 19:17:17 2014] <Wohali>: good > [Wed Oct 8 19:17:31 2014] <Wohali>: any other thoughts from people present? > [Wed Oct 8 19:17:54 2014] <jan____>: I think I agree with the plans :) > [Wed Oct 8 19:18:11 2014] <Wohali>: ok > [Wed Oct 8 19:18:12 2014] <jan____>: happy to lend a hand, especially with > ripping out jquery.couch.js > [Wed Oct 8 19:18:35 2014] <Kxepal>: it's in safe place already P: > [Wed Oct 8 19:18:49 2014] <jan____>: :) > [Wed Oct 8 19:19:16 2014] <Wohali>: ok > [Wed Oct 8 19:19:37 2014] <Wohali>: so let's move ahead, I can get a summary > email on this to dev@ > [Wed Oct 8 19:20:06 2014] <Wohali>: there is alsot his little thing called > ApacheCon EU. it sure would be nice to have a beta or release by then ;) ;) > [Wed Oct 8 19:20:33 2014] <Wohali>: next topic > [Wed Oct 8 19:20:40 2014] <Wohali>: we don't seem to have any fauxtoneers > here i'm afraid > [Wed Oct 8 19:20:51 2014] <Wohali>: but i can say that there is lots of good > work being done > [Wed Oct 8 19:20:51 2014] <dch>: in the dark, so to speak. > [Wed Oct 8 19:21:08 2014] <Wohali>: lots of new commits and new people > joining up to help every day > [Wed Oct 8 19:21:17 2014] <Wohali>: even pouch guys submitted a PR > [Wed Oct 8 19:21:20 2014] <Wohali>: so that's very nice. > [Wed Oct 8 19:21:41 2014] <jan____>: +1 really happy to see that > [Wed Oct 8 19:21:55 2014] <Kxepal>: +1 > [Wed Oct 8 19:22:11 2014] <Kxepal>: hope to see the same activity for other > couchdb-* repos one day (: > [Wed Oct 8 19:22:13 2014] <jan____>: (the takeaway is that breaking out > things into smaller repos really helps with contributions, especially the > non-erlang bits > [Wed Oct 8 19:22:17 2014] <jan____>: ( > [Wed Oct 8 19:22:20 2014] <Wohali>: (totally!) > [Wed Oct 8 19:22:21 2014] <jan____>: ) > [Wed Oct 8 19:22:26 2014] <Kxepal>: ) > [Wed Oct 8 19:22:32 2014] <Wohali>: is anyone around who wants to talk about > view sequences? > [Wed Oct 8 19:22:40 2014] <Wohali>: that would be benoitc sagelywizard > [Wed Oct 8 19:22:50 2014] <Wohali>: am guessing no > [Wed Oct 8 19:23:13 2014] <Wohali>: oops > [Wed Oct 8 19:23:14 2014] <chewbranca>: late comment, but it would be great > to get as many of the eunit test suites fixed as possible > ## Fauxton ## > [Wed Oct 8 19:23:19 2014] <jan____>: the thing to note then is that this is > an ongoing discussion on dev@ > [Wed Oct 8 19:24:38 2014] <Wohali>: view sequences is, though the discussion > is not getting any input on how that interface will be used. i am inclined to > move that we commit it as experimental and get more input as we start to > build out things that use it like chained mapreduce. > [Wed Oct 8 19:24:53 2014] <Wohali>: chewbranca: *nod* looks like kxepal > volunteered :) > [Wed Oct 8 19:25:27 2014] <chewbranca>: Wohali: a few of them are pretty > tricky > [Wed Oct 8 19:25:29 2014] <jan____>: Wohali: is it relevant to the > “replicate from views” feature? > [Wed Oct 8 19:25:38 2014] <jan____>: releaventrelated > [Wed Oct 8 19:25:59 2014] <chewbranca>: timing bugs in compaction tests and > fun things like that > [Wed Oct 8 19:26:31 2014] <Wohali>: jan____: Yeah, anything that requires an > view sequence index. _changes on views, etc. the rcoucn implementation, near > as I can tell is missing two key components: deleted rows, and duplicate keys > [Wed Oct 8 19:26:55 2014] <Wohali>: chewbranca: ugh, fun. and presumably > inserting artificial delays isn't sufficient > [Wed Oct 8 19:27:30 2014] <Wohali>: jan____: in either case, I don't think > biulding "replicating from views" or 'chained map reduce" is going to hit in > time for a 2.0 release, so I'd rather get the functionality in as > experimental, then revisit what changes are required to build something > useful on top of it in a 2.1 timeframe. > [Wed Oct 8 19:27:34 2014] <chewbranca>: nope, that rarely ever works well, > but I tried it and it didn't work for everything > [Wed Oct 8 19:27:34 2014] <jan____>: Wohali: curious, so replication should > not work, but it is clearly advertised as such (and I believed used in the > wild) > [Wed Oct 8 19:27:54 2014] <jan____>: Wohali: sounds like a plan! > [Wed Oct 8 19:27:55 2014] <Wohali>: jan____: rnewson has more detail here, > and I'm afraid I can't address it > [Wed Oct 8 19:28:13 2014] <jan____>: Wohali: alright, let’s leave it at dev@ > [Wed Oct 8 19:28:19 2014] <Wohali>: distributed systems are hard, let's go > shopping. > [Wed Oct 8 19:29:07 2014] <Wohali>: hrm > ## open floor ## > [Wed Oct 8 19:29:16 2014] <Wohali>: not sure what others have here? > [Wed Oct 8 19:29:38 2014] <jan____>: ACTION clumsily falls through the open > floor > [Wed Oct 8 19:30:48 2014] <Wohali>: TL;DR version of minutes: lots of > change. time to test and release. > [Wed Oct 8 19:31:09 2014] <Wohali>: anyone else? i'll close in 60s if not > [Wed Oct 8 19:31:24 2014] <chewbranca>: quick clarification on view changes > [Wed Oct 8 19:31:38 2014] <dch>: how much related docs are needed to ship > this? > [Wed Oct 8 19:32:37 2014] <chewbranca>: jan____: from what I understand from > talking to rnewson, is that the underlying problem with view changes is that > negative updates are not propagated, so for instance if you have a doc that > emits 10 rows and then update the doc resulting in it only emitting 3 rows, > then there isn't a mechanism to make "delete" changes for those rows > [Wed Oct 8 19:33:30 2014] <jan____>: chewbranca: as implemented in rcouch? > [Wed Oct 8 19:34:06 2014] <chewbranca>: jan____: rcouch and the couchdb > implementation from what I understand, although rnewson, sagelywizard, and > benoitc could comment more accurately > [Wed Oct 8 19:34:54 2014] <jan____>: chewbranca: where does “the couchdb > implemetnation” come from? is that BC heritage, rcouch heritage, merge or > clean room? > [Wed Oct 8 19:35:27 2014] <Wohali>: my understanding is that it started with > the rcouch implementation and evolved from there. > [Wed Oct 8 19:35:29 2014] <chewbranca>: jan____: I believe sagelywizard took > the version in rcouch and updated it > [Wed Oct 8 19:35:33 2014] <Wohali>: ^^^ > [Wed Oct 8 19:35:46 2014] <jan____>: aye > [Wed Oct 8 19:35:51 2014] <Wohali>: updated for fabric changes, etc. > [Wed Oct 8 19:36:08 2014] <jan____>: ok, I’ll do some research then, this > seems like an odd omission > [Wed Oct 8 19:36:09 2014] <chewbranca>: > https://github.com/apache/couchdb-couch-mrview/pull/2 > [Wed Oct 8 19:36:48 2014] <Wohali>: actually, it looks like benoitc did a +1 > here so we should just merge it > [Wed Oct 8 19:37:00 2014] <sagelywizard>: If you change a doc to emit fewer > rows, that update will propagate. > [Wed Oct 8 19:37:10 2014] <Wohali>: there's the man > [Wed Oct 8 19:37:29 2014] <chewbranca>: outdated +1 fwiw > [Wed Oct 8 19:37:41 2014] <Wohali>: is it? ok > [Wed Oct 8 19:37:42 2014] <Kxepal>: Wohali: oh..I recall one topic > [Wed Oct 8 19:37:42 2014] <chewbranca>: sagelywizard: that's recent though > right? > [Wed Oct 8 19:38:00 2014] <Wohali>: Kxepal: k, let's close this out > [Wed Oct 8 19:38:03 2014] <Kxepal>: but may be ML is better place for it > [Wed Oct 8 19:38:15 2014] <Kxepal>: 1.7 and API deprecation (: > [Wed Oct 8 19:38:32 2014] <Wohali>: Kxepal: I didn't see a list of APis that > were actually deprecated but yes, I think we agreed > [Wed Oct 8 19:38:52 2014] <chewbranca>: sagelywizard: you should sync up > with rnewson ;-) > [Wed Oct 8 19:38:54 2014] <Kxepal>: Wohali: I had already sent differences I > found + may be something had already changed. > [Wed Oct 8 19:39:14 2014] <Kxepal>: + there are some plans to change moar > bits > [Wed Oct 8 19:39:22 2014] <Kxepal>: while there is time for > [Wed Oct 8 19:39:24 2014] <Wohali>: are you ready for a PR against the 1.6.x > branch to make a 1.7? we obviusly can't cut 1.7 from master now. :) > [Wed Oct 8 19:39:31 2014] <Wohali>: obviously* > [Wed Oct 8 19:39:46 2014] <Kxepal>: Wohali: to PR what? > [Wed Oct 8 19:39:47 2014] <Wohali>: Kxepal: "change more bits?" watch out, > would those also hit master? > [Wed Oct 8 19:39:47 2014] <sagelywizard>: chewbranca: There was a bug in the > _view_changes UI code which Bob found. Been fixed since then. Maybe that's > what you're thinking of? > [Wed Oct 8 19:40:00 2014] <Wohali>: Kxepal: A PR for a 1.7 release on > deprecation documentation. > [Wed Oct 8 19:40:00 2014] <jan____>: ah yes, on that note, I suggest we make > a 1.x.x branch, so we can do 1.7.x and forther from it > [Wed Oct 8 19:40:10 2014] <Wohali>: jan____: agree > [Wed Oct 8 19:40:10 2014] <Kxepal>: Wohali: change more bits in api for 2.0 > [Wed Oct 8 19:40:11 2014] <jan____>: *further > [Wed Oct 8 19:40:28 2014] <Wohali>: Kxepal: I suggest we don't, at this > point. > [Wed Oct 8 19:40:39 2014] <sagelywizard>: chewbranca: And yeah. Bob has been > busy though. > [Wed Oct 8 19:40:43 2014] <Kxepal>: Wohali: sure + eunit I have to do > [Wed Oct 8 19:40:49 2014] <Wohali>: exactly > [Wed Oct 8 19:40:55 2014] <Wohali>: we need to get past 2.0 > [Wed Oct 8 19:41:06 2014] <Wohali>: then doing tweaks for 2.1/3.0 aren't > coupled with zOMG BIG CHANGE > [Wed Oct 8 19:41:17 2014] <jan____>: this > [Wed Oct 8 19:41:30 2014] <Kxepal>: ok, let's discuss the detailed plan in ML > [Wed Oct 8 19:41:33 2014] <Wohali>: great. > [Wed Oct 8 19:41:35 2014] <dch>: might be better leaving significant > deprecation for after 2.0 > [Wed Oct 8 19:41:50 2014] <dch>: already a significant amt of changes > [Wed Oct 8 19:41:50 2014] <Wohali>: dch: yeah, I think kxepal is just > talking about anything that is already on master that resulted in a > deprecation > [Wed Oct 8 19:42:00 2014] <dch>: gotcha > [Wed Oct 8 19:42:08 2014] <Kxepal>: yes > [Wed Oct 8 19:42:17 2014] <Wohali>: #action jan____ to cut a 1.x.x branch > [Wed Oct 8 19:42:23 2014] <jan____>: nah, let’s remove all _ fields from doc > json *hids* > [Wed Oct 8 19:42:33 2014] <Wohali>: #action kxepal to propose a 1.7 release > that includes deprecations for what is already in master (and expected to hit > 2.0) > [Wed Oct 8 19:42:41 2014] <Kxepal>: jan____: +1 (: > [Wed Oct 8 19:42:53 2014] <Wohali>: #action kxepal to take lead on fixing > broken eunit stuff > [Wed Oct 8 19:43:08 2014] <Wohali>: #action sagelywizard to get closure on > his thread and commit view sequence work as experimental > [Wed Oct 8 19:43:23 2014] <Wohali>: #action wohali to send email with > suggested last steps before we move towards a 2.0 release > [Wed Oct 8 19:43:26 2014] <Wohali>: did i capture everything? > [Wed Oct 8 19:43:59 2014] <jan____>: #action jan____ to help with > s/futon/fauxton > [Wed Oct 8 19:44:09 2014] <Wohali>: ah yes thanks > [Wed Oct 8 19:44:11 2014] <Kxepal>: good point! > [Wed Oct 8 19:44:15 2014] <Wohali>: tap garren for help if you see it > [Wed Oct 8 19:44:19 2014] <Wohali>: that was going to be on my email > [Wed Oct 8 19:45:34 2014] <Wohali>: anyone else? > [Wed Oct 8 19:45:43 2014] <dch>: Wohali: I mentioned “what about docs" > [Wed Oct 8 19:46:00 2014] <Wohali>: that's in my email but > [Wed Oct 8 19:46:03 2014] <dch>: basically what do we *need* to ship with. > [Wed Oct 8 19:46:06 2014] <Wohali>: if you have somethign to contribute, you > have the floor :) > [Wed Oct 8 19:46:07 2014] <dch>: all good then. > [Wed Oct 8 19:46:54 2014] <Kxepal>: dch: something we could pick up from > http://bigcouch.cloudant.com/ but mostly all need to be revised in the fact > to match 2.0 behavior > [Wed Oct 8 19:46:57 2014] <Wohali>: i'll be asking for volunteers to run all > the docs against a 2.0 release and make sure they're accurate > [Wed Oct 8 19:47:02 2014] <Kxepal>: especially imported stuff from the guide > [Wed Oct 8 19:47:09 2014] <dch>: mmmm > [Wed Oct 8 19:47:16 2014] <Wohali>: the guide is going to need a rework..... > [Wed Oct 8 19:47:45 2014] <Wohali>: that's upt o nslater and jan____ > [Wed Oct 8 19:47:49 2014] <Wohali>: ok everyone > [Wed Oct 8 19:47:51 2014] <Wohali>: thanks for coming! > [Wed Oct 8 19:47:55 2014] <Wohali>: ASFBot: meeting end > > > Meeting ended at Wed Oct 8 19:47:55 2014
signature.asc
Description: Message signed with OpenPGP using GPGMail
