couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF IRC Services" <asf...@wilderness.apache.org>
Subject Summary of IRC meeting in #couchdb-meeting, Wed Mar 6 19:59:20 2013
Date Wed, 06 Mar 2013 20:59:50 GMT
Members present: emjayess, sbp, eckoit, davisp, nslater, rnewson, chewbranca, benoitc

----------------
Meeting summary:
----------------

1. Preface

2. porting CouchDB to LLVM, allowing use of Emscripten etc.
  a. sbp to post a proposl on porting couch to llvm to dev list (nslater, 2)

3. 1.3 release status

4. bigcouch status
  a. waiting on couchdb-pmc.git for 1.3.0 release (nslater, 4)
  b. will call lazy consensus for new release procedure at the same time as the release (nslater,
4)
  c. davisp and rnewson are working on pulling all of the code out of cloudant's private forks
 (nslater, 4)
  d. goal is to have a bigcouch branch by the end of next week for IP clearance procedure
(nslater, 4)
  e. rnewson lists to some weird music (davisp, 4)

5. fauxton status

6. work has started on fauxton test suite

7. fauxton
  a. spring cleaning on fauxton branch (nslater, 7)
  b. work has started on fauxton test suite (nslater, 7)

8. rcouch status
  a. rcouch branch is now configuring icu, spidermonkey, etc (nslater, 8)
  b. benoit still figuring out how to run tests, etc (nslater, 8)
  c. still some questions about how to manage mochiweb, details to follow (nslater, 8)
  d. ip clearance still pending, blocking asf branch creation (nslater, 8)

9. project activity visibility
  a. nslater to come back to community guidelines proposal after 1.3. release (nslater, 9)
  b. please please please respond to "[DISCUSS] Goals for 2013" thread - everyone! (nslater,
9)

10. 1.4 roadmap
  a. benoit to gather list of wishes for next major release (nslater, 10)
  b. https://confluence.atlassian.com/display/JIRA/Browsing+a+Project's+Road+Map (nslater,
10)
  c. still not decided on how to maintain a roadmap. benoit to investigate (nslater, 10)


--------
Actions:
--------
- sbp to post a proposl on porting couch to llvm to dev list (nslater, 20:05:28)
- benoit to gather list of wishes for next major release (nslater, 20:54:41)

IRC log follows:


# 1. Preface #
19:59:30 [nslater]: please list topics. i will wait for 5 mins
20:00:13 [nslater]: from the ML, benoitc brings up: project activity visibility, 1.4 roadmap,
1.3 release status
20:00:28 [nslater]: i will add bigcouch status, fauxton status, and... lemme check
20:00:43 [nslater]: rcouch status
20:00:50 [sbp]: * porting CouchDB to LLVM, allowing use of Emscripten etc.
20:01:05 [nslater]: thank you for your contribution sbp
20:01:07 [nslater]: please note this is a logged channel during meetings
20:01:20 [nslater]: okay thats it for topics from me. anyone else?
20:01:36 [nslater]: i will wait for another few minutes
20:03:30 [nslater]: okay lets start


# 2. porting CouchDB to LLVM, allowing use of Emscripten etc. #
20:03:43 [chewbranca]: o/
20:03:50 [nslater]: sbp the floor is yours
20:03:58 [sbp]: thanks. so I was thinking that we could probably use http://erllvm.softlab.ntua.gr/
for this
20:04:20 [nslater]: wait... is this a serious suggestion?
20:04:28 [sbp]: basically this would allow us to compile CouchDB down to LLVM level, which
would mean we get native LLVM-level compiler optimisations
20:04:35 [nslater]: ACTION rolls about on the floor
20:04:35 [sbp]: and then from there we can rebuild it using different toolchains for other
environments
20:04:50 [sbp]: Emscripten being an obvious example, but I'm sure there are others
20:04:52 [nslater]: sounds good sbp
20:04:59 [nslater]: id suggest you post a proposal to dev@couchdb.a.o
20:05:05 [sbp]: with TypedArrays in JavaScript and so on, it could probably be made viable
within a year
20:05:16 [davisp]: sbp: Sounds like a good plan. When you've got a third proof-of-concept
backend for ErlLLVM I'd be interested in seeing it
20:05:28 [nslater]: #action sbp to post a proposl on porting couch to llvm to dev list
20:05:58 [nslater]: moving swiftly on... ;)


# 3. 1.3 release status #
20:06:28 [nslater]: okay!
20:06:28 [nslater]: if you've been following along with http://wiki.apache.org/couchdb/Release_Procedure
20:06:35 [nslater]: ... you will have seen a lot of changes happening
20:06:45 [nslater]: i am still waiting on infra for http://git-wip-us.apache.org/repos/asf/couchdb-pmc.git
to be set up
20:06:58 [nslater]: its going to be for utility scripts, proj mgmt stuff, email templates
etc
20:07:05 [davisp]: sounds good
20:07:06 [nslater]: a lot of the release procedure is now automated because of this
20:07:20 [nslater]: as soon as the repos is created, i am going to cut 1.3.0-rc.1
20:07:35 [nslater]: at the same time i will start a lazy consensus thread for the proposed
release procedure
20:07:43 [nslater]: i dont expect any objections. and hopefully both will pass on the same
day
20:08:07 [nslater]: i still have a few finishing touches on this wiki - but ive been flat
out on this for a few weeks now
20:08:22 [nslater]: end result is: should be much easier, and must faster to cut monthly releases
20:08:35 [nslater]: one critical step was removing the tests from the candidate build cycle
20:08:43 [nslater]: the test suite is now exclusively run while testing the actual candidate
20:08:58 [nslater]: this was the primary reason releases took so long
20:09:05 [nslater]: any questions about this?
20:09:28 [nslater]: okay


# 4. bigcouch status #
20:09:43 [nslater]: davisp: over to you
20:09:43 [davisp]: hooray me
20:10:05 [davisp]: Me and rnewson are working on pulling all of the code out of cloudant's
private forks 
20:10:21 [davisp]: so far we've finished chttpd which is the clustered HTTP layer
20:10:21 [nslater]: #info waiting on couchdb-pmc.git for 1.3.0 release
20:10:23 [nslater]: #info will call lazy consensus for new release procedure at the same time
as the release
20:10:28 [nslater]: (sorry)
20:10:35 [davisp]: we're current working on couch_replicator
20:10:43 [nslater]: cool
20:10:43 [davisp]: the hard part will be some changes to the actual couch code itself
20:10:58 [benoitc]: like ?
20:10:58 [benoitc]: the ddoc part?
20:11:00 [davisp]: benoitc: ddoc part?
20:11:05 [benoitc]: or there are other ?
20:11:13 [rnewson]: can we let davisp speak his piece before asking questions?
20:11:20 [benoitc]: davisp: well last time i checked you had a change to ollect ddocs from
the cluster
20:11:20 [davisp]: rnewson: shush
20:11:22 [benoitc]: for the validate_doc_update
20:11:35 [davisp]: benoitc: yes, a number of things like that
20:11:37 [benoitc]: ok
20:11:45 [davisp]: benoitc: we have a change to the by_seq tree to store #full_doc_info{}
records
20:11:52 [benoitc]: ok
20:11:58 [davisp]: whichi speeds up replication and compaction
20:12:13 [davisp]: There's a new compactor that I think will have conflicts
20:12:21 [davisp]: We remove one of the main db updater pids
20:12:22 [nslater]: #info davisp and rnewson are working on pulling all of the code out of
cloudant's private forks 
20:12:35 [benoitc]: davisp: ok mostly like in rcouch then
20:12:44 [davisp]: Not at all sure
20:12:58 [davisp]: benoitc: I think the one thing I know about in rcouch code related is the
view update_seq thing yeah?
20:13:13 [benoitc]: davisp: there is also an adaptation from your patch
20:13:20 [davisp]: The ref counting thing is also different in bigcouch. That'll probably
take some time
20:13:20 [davisp]: which patch?
20:13:20 [benoitc]: removing refcounting and other things
20:13:28 [benoitc]: in your repo
20:13:28 [davisp]: ahh cool
20:13:35 [davisp]: yeah, that was prep work for this
20:13:43 [davisp]: so it should make the rcouch integration easier in the end
20:13:50 [benoitc]: ok i also did changes on that
20:13:51 [benoitc]: ok cool!
20:14:05 [davisp]: benoitc: hopefully once I've gone through that we can take a look at rcouch
as well
20:14:05 [nslater]: any other #info items to add?
20:14:13 [davisp]: nslater: #info all the things
20:14:28 [nslater]: can yo info anything else we need to highlight here?
20:14:28 [benoitc]: davisp: cool
20:14:43 [davisp]: nslater: benoitc: Goal is to have a branch by the end of next week that'll
be a candidate for shuffling through the IP clearance procedure
20:15:13 [nslater]: #info goal is to have a bigcouch branch by the end of next week for IP
clearance procedure
20:15:20 [nslater]: sweeeeet
20:15:20 [davisp]: #info rnewson lists to some weird music
20:15:20 [benoitc]: nice
20:15:28 [nslater]: okay lets move on shall we?
20:15:29 [davisp]: moving along


# 5. fauxton status #
20:15:50 [nslater]: chewbranca drsm79 you have the floor
20:16:05 [chewbranca]: fauxton is coming along, we're doing some spring cleaning right now
to make sure the things that are in place work as expected
20:16:35 [chewbranca]: work has also started on getting a test suite in place
20:16:58 [chewbranca]: primary goals are make sure all the things that are there now work
as expected and continue to work as expected
20:17:28 [nslater]: cool
20:17:29 [benoitc]: what are all the thing you want ?


# 6. work has started on fauxton test suite #
20:18:13 [benoitc]: chewbranca: ----------->°---
20:18:15 [chewbranca]: benoitc: mainly just going through and fixing little things that don't
work right, and making sure placeholder stuff isn't in the way
20:18:29 [chewbranca]: basically we're doing QA
20:19:00 [benoitc]: hrmm i mean n term of feature, what do you want before to push in the
repo ?
20:19:13 [nslater]: oh shit why did i topic that haha


# 7. fauxton #
20:19:28 [nslater]: #info spring cleaning on fauxton branch
20:19:29 [nslater]: #info work has started on fauxton test suite
20:19:31 [chewbranca]: benoitc: not following
20:20:43 [benoitc]: ACTION just wanted to know what were the "things" 
20:20:58 [eckoit]: I have some things I will be bringing into fauxton this week. I will be
sending PR's through chewbranca
20:21:06 [chewbranca]: ie click a link on a random page and it doesn't do what you expect
20:21:28 [chewbranca]: there's random bugs that need to be fixed, I'm working on that the
rest of this week to make things more polished and functional
20:21:59 [chewbranca]: yesterday I brought in the initial view editor updates, will push that
back out to the asf repo with the other updates this week
20:22:07 [benoitc]: okaye
20:22:20 [nslater]: eckoit: would you find a commit bit useful?
20:22:44 [eckoit]: nslater: yes for sure...would like to commit to the asf repo direct
20:22:50 [chewbranca]: just chugging along, drsm79 is out at a conference this week so not
at full steam ahead
20:22:50 [rnewson]: eckoit +1
20:22:50 [nslater]: eckoit: okay. noted
20:23:28 [nslater]: chewbranca: cool. thanks. anything else to #info?
20:23:43 [chewbranca]: nope, not today
20:23:50 [nslater]: thanks!
20:23:58 [nslater]: benoitc: all yours
20:26:35 [nslater]: Humbedooh: hmm?
20:26:35 [nslater]: oops


# 8. rcouch status #
20:27:13 [benoitc]: ok
20:27:13 [nslater]: i guess benoitc is busy. we'll loop back around
20:27:13 [nslater]: oh.. okay
20:27:13 [nslater]: hehe
20:27:43 [benoitc]: so i will have a branch of rcouch up tomorrow
20:27:50 [benoitc]: quite finished but was side tracked today
20:27:58 [nslater]: 0-|-<
20:27:58 [nslater]: 0->-<
20:28:05 [benoitc]: what is working for now: detect icu on the disk
20:28:13 [benoitc]: or build a static version
20:28:28 [benoitc]: detect spidermonkey 1.8.5 on disk
20:28:35 [benoitc]: or specify a path for shared lib
20:28:43 [benoitc]: or build it statically
20:28:50 [benoitc]: what i have to figure yet is how to run tests
20:29:20 [benoitc]: but il will try some code in the morning
20:29:35 [benoitc]: so far it creates a full couchdb release
20:29:43 [benoitc]: that you can relocate
20:30:28 [benoitc]: there are still some questions like how to manage mochiweb and co
20:30:43 [benoitc]: i'm done i think except i some have questions
20:30:58 [nslater]: benoitc: that is excellent news!
20:31:22 [nslater]: #info rcouch branch is now configuring icu, spidermonkey, etc
20:31:35 [nslater]: #info benoit still figuring out how to run tests, etc
20:31:45 [nslater]: benoitc: will you be posting your questions to the ml?
20:32:07 [nslater]: #info still some questions about how to manage mochiweb, details to follow
20:33:14 [nslater]: benoitc: ?
20:33:28 [benoitc]: nslater: yes sorry
20:33:29 [nslater]: cool
20:33:35 [benoitc]: i will do a status tonight
20:33:35 [nslater]: anything more to add or to #info?
20:33:43 [benoitc]: oh
20:33:44 [nslater]: benoitc: wait... hmm...
20:33:53 [nslater]: we need to do ip clearence before you can do the branch, right?
20:34:00 [benoitc]: we need to launch ip clearances i guess$
20:34:05 [benoitc]: yes
20:34:06 [benoitc]: sorry it's lagging here
20:34:28 [nslater]: when are you free. i can knock that out of the park with you if we have
a time to sit down and get the details sorted
20:34:28 [benoitc]: nslater: so yes
20:34:52 [nslater]: let's take this off-line
20:34:58 [benoitc]: ok
20:35:07 [nslater]: #info ip clearance still pending, blocking asf branch creation


# 9. project activity visibility #
20:35:22 [nslater]: benoitc: you want to lead with this one?
20:35:35 [nslater]: i sent out an email to the dev list which sort of covers this... maybe
a little bit
20:35:43 [nslater]: goals for 2013
20:38:06 [nslater]: benoitc: i would loop back to this but the next topic is also one of yours
20:38:28 [benoitc]: mmm
20:38:31 [nslater]: anybody else have any thoughts here?
20:38:36 [nslater]: or thoughts about the email i sent?
20:39:00 [benoitc]: i don't see any email yet
20:39:06 [nslater]: i think one of the things we have to get better at doing is posting summaries
of all the important things that are going on to the dev mailing list
20:39:08 [benoitc]: but i'm like on a 9600 baud/s
20:39:28 [emjayess]: project activity visibility? is that ~ the community mgmt mail? or something
else
20:39:31 [benoitc]: so that and i woild like also to see more mails like the last from jason
20:39:35 [nslater]: this is especially important for folks like davisp, rnewson, benoitc,
drsm79, chewbranca etc - who are all working on branches from the mainline
20:39:50 [nslater]: emjayess: its about making sure that we're all spamming the dev list with
important news! :D
20:39:52 [benoitc]: ie pushing project status tries and co
20:40:43 [nslater]: i am not sure if there are any action items here
20:40:51 [nslater]: i have a todo item about creating a community guidelines doc
20:40:58 [nslater]: which i guess could cover this
20:41:13 [nslater]: benoitc: subject is "[DISCUSS] Goals for 2013"
20:41:35 [nslater]: id like to come back to the community guidelines proposal after the 1.3.
release
20:41:43 [nslater]: i am too swamped to spare any additional cycles
20:41:58 [benoitc]: i don't have it yet
20:41:58 [nslater]: #info nslater to come back to community guidelines proposal after 1.3.
release
20:42:00 [benoitc]: odd
20:42:05 [nslater]: does anyone else?
20:42:21 [eckoit]: just, check yes. I have it
20:42:28 [nslater]: cool
20:42:36 [nslater]: please, please... please. everyone! please post your ideas there
20:42:43 [benoitc]: that's probably just my connection
20:42:50 [nslater]: i think its crucial we identify some goals that we can reasonably acheive
20:42:53 [nslater]: and then rally the community around it for the whole year
20:43:01 [nslater]: but i need some help coming up with metrics
20:43:35 [nslater]: benoitc: is there anything else you'd like to bring up during the meeting?
20:43:43 [nslater]: (on this topic)
20:44:05 [nslater]: #info please please please respond to "[DISCUSS] Goals for 2013" thread
- everyone!
20:44:35 [benoitc]: no
20:44:43 [nslater]: okay lets move on


# 10. 1.4 roadmap #
20:45:29 [nslater]: benoitc: this is your topic again. wanna take the floor with this?
20:46:15 [benoitc]: nslater: i think it's a little latee
20:46:28 [benoitc]: but as an action i woish we could sertup a roadmap
20:46:28 [benoitc]: asap
20:46:35 [benoitc]: what we want really for this one
20:46:43 [nslater]: benoitc: you have edit perms on the wiki and jira, right?
20:46:52 [benoitc]: also now that the 1.3 is a little delayed, what is the deadline
20:46:57 [benoitc]: nslater: probabluy
20:47:21 [benoitc]: i think i hav eit yes
20:47:27 [nslater]: well, id say, when we release this - well just do the 3 month thing from
that point
20:47:28 [benoitc]: at leats i can go on plugins
20:47:33 [nslater]: which should be soon
20:47:41 [nslater]: how do you think we should do the roadmap?
20:47:43 [nslater]: can we figure out a way to use jira?
20:47:48 [nslater]: perhaps set up a report that can be shared?
20:47:56 [nslater]: and then use ticket statuses etc
20:48:28 [benoitc]: hmm my knowledge of jira is preety much 0
20:48:34 [benoitc]: biut I can have a look
20:48:54 [benoitc]: maybe it can also start with a review of what is the goal of each devs
on the ml
20:48:54 [nslater]: cool. please do and let us know!
20:49:01 [benoitc]: if some want tome features
20:49:01 [nslater]: my hopes are that we can just leverage jira
20:49:09 [benoitc]: for me maybe rcouch branch
20:49:15 [nslater]: benoitc: maybe you wanna kick off that thread too
20:49:15 [benoitc]: and also websockets
20:49:17 [benoitc]: + rewrier
20:49:38 [nslater]: remember: there's no central management here. if you think this is something
we should do: do it :)
20:50:00 [nslater]: (i know i dont need to remind you! just saying... run with it! sounds
like a great idea!)
20:50:23 [emjayess]: question: are the big inbound merges going to warrant a bump in major
version? or just like ho-hum here's 1.4 btw we merged bigcouch
20:50:23 [nslater]: i have thought about it in the past. looked at jira even
20:50:45 [nslater]: emjayess: bigcouch + couchdb will be couchdb 2.0...
20:50:47 [benoitc]: nslater: that's not for me , more for the users
20:50:53 [nslater]: emjayess: there's a chance it wont even be called couchdb
20:50:53 [benoitc]: trying to figure what's going on
20:51:09 [nslater]: benoitc: yeah, sure
20:51:23 [nslater]: i just mean, people have been talking about having a roadmap for ages
20:51:23 [nslater]: but nobody has produced one
20:51:30 [nslater]: all it takes is one person to run with the idea
20:51:31 [emjayess]: alrgiht then .thx
20:51:53 [nslater]: emjayess: the naming thing is contentious. but i will be calling a [DISCUSS]
thread about rebranding
20:52:08 [nslater]: emjayess: this isn't the meeting to talk about that though :)
20:52:31 [nslater]: benoitc: would you like to generate any action items for yourself wrt
the 1.4 roadmap?
20:52:38 [emjayess]: *bigtables
20:52:38 [emjayess]: :P
20:53:11 [nslater]: emjayess: :-) i am going to refuse to even say my ideas until we have
agreed as a project that its a good idea. bikeshed hell, otherwise
20:53:23 [benoitc]: nslater: i will think more
20:53:38 [benoitc]: i mostly wanted to draw a draft of what could be put on this release
20:53:40 [nslater]: benoitc: no action items for now? how about that you will post some ideas
about it this week?
20:53:46 [benoitc]: depending on the wish of everyone ...
20:53:56 [nslater]: is that something you want to gather?
20:54:11 [benoitc]: yes but can be  a post oon the ml
20:54:18 [benoitc]: let see it as an action 
20:54:41 [nslater]: #action benoit to gather list of wishes for next major release
20:54:49 [nslater]: emjayess has a point that this might not be 1.4...
20:54:58 [nslater]: there's a good chance it will be 2.0 or whatever
20:55:05 [nslater]: i mean, it's going to be a very important release for us
20:55:18 [nslater]: in terms of marketing, etc
20:56:03 [emjayess]: mebbe #link https://confluence.atlassian.com/display/JIRA/Browsing+a+Project's+Road+Map
20:56:11 [emjayess]: oh that's confluence, sorry
20:56:33 [emjayess]: or it's jira? i dunno, i'z confuzed 
20:57:33 [nslater]: that still looks interesting
20:57:33 [nslater]: it's a confluence page ABOUT jira
20:57:41 [nslater]: #link https://confluence.atlassian.com/display/JIRA/Browsing+a+Project's+Road+Map
20:57:56 [nslater]: #info still not decided on how to maintain a roadmap. benoit to investigate
20:57:58 [nslater]: okay, i think that's it guys
20:58:03 [nslater]: as usual, almost exactly an hour
20:58:35 [benoitc]: links is interresting
20:58:41 [benoitc]: ok 
20:58:41 [nslater]: okay, thanks guys
20:58:43 [benoitc]: so i'm off a little
20:58:48 [nslater]: cya!


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message