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 Apr 24 19:05:05 2013
Date Wed, 24 Apr 2013 20:21:21 GMT
Members present: jan____, benoitc, ryan_ramage, nslater, rnewson, tilgovi, chewbranca, dch

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

1. Preface

2. release calendar 
  a. i have published a release timetable here: http://wiki.apache.org/couchdb/Release_Procedure#Timetable
(nslater, 2)
  b. release calendar to be created that documents planned releases, etc (nslater, 2)
  c. next release window set for 22nd may, bugfix release(s) (nslater, 2)

3. merge procedure

4. merges

5. merge procedure
  a. rnewson and nslater to bring merge procedure proposal to dev@ soon (nslater, 5)

6. nebraska bigcouch

7. rcouch

8. nebraska bigcouch

9. fauxton
  a. chewbranca to invite collaborators to fauxton, dch to get some tweets on that  (dch,
9)

10. windows couch

11. rcouch

12. mochiweb
  a. https://github.com/mochi/mochiweb/commit/56a32a0cc091b69374f35aa0787c36c7776ae816 (dch,
12)
  b. TL;DR we should stick on the v2.4.2 tag which is luckily what I've done, as after that
it drops R14* support, which is a bit draconian. (dch, 12)

13. 1.3.0 post-mortem
  a. benoitc to send mail about roadmap visibility and maintainance (tilgovi, 13)

14. meeting timing / schedule
  a. dch to do the doodle poll (dch, 14)
  b. continue with alternating meeting times til end May and collect feedback in between (dch,
14)
  c. next week to be 13h00 UTC (because northern hemisphere is on DST atm) (dch, 14)

15. gsoc


--------
Actions:
--------
- rnewson and nslater to bring merge procedure proposal to dev@ soon (nslater, 19:15:13)
- chewbranca to invite collaborators to fauxton, dch to get some tweets on that  (dch, 19:28:13)
- benoitc to send mail about roadmap visibility and maintainance (tilgovi, 19:56:08)
- dch to do the doodle poll (dch, 20:12:23)
- next week to be 13h00 UTC (because northern hemisphere is on DST atm) (dch, 20:17:36)

IRC log follows:


# 1. Preface #
19:05:13 [nslater]: rcouch
19:05:13 [nslater]: bigcouch
19:05:15 [nslater]: fauxton
19:05:20 [dch]: welcome!
19:05:20 [nslater]: release calendar
19:05:20 [tilgovi]: meeting time
19:05:20 [nslater]: merge procedure
19:06:05 [dch]: topics noted - times, merges, release calendar, fauxton, bigcouch, rcouch,
windows couch, mochiweb,  I think thats it.
19:06:05 [tilgovi]: did we do a 1.3 postmortem last time?
19:06:13 [dch]: noted
19:06:20 [dch]: topic release calendar 


# 2. release calendar  #
19:07:07 [dch]: nslater: you have the floor
19:07:13 [nslater]: #info i have published a release timetable here: http://wiki.apache.org/couchdb/Release_Procedure#Timetable
19:07:28 [nslater]: this dictates to me the weekly heartbeat of the release procedure
19:07:35 [nslater]: i will be trying this out for a while to see if it works
19:07:50 [nslater]: a release calendar will be one of the things that comes out of that
19:07:58 [nslater]: i.e. a documented plan of when the next releases are, etc, etc
19:08:05 [nslater]: all of this a WIP at the moment
19:08:28 [nslater]: #info release calendar to be created that documents planned releases,
etc
19:08:35 [dch]: nslater: looks pretty sensible all round.
19:08:37 [nslater]: so, for the next release
19:08:51 [nslater]: that process kicks off 1st may
19:09:20 [nslater]: as you can see from the time table, i will be hoping to ramp up to the
actual vote thread by the 22nd may
19:09:28 [nslater]: may is one of those months with 5 tuesdays, so the final step is "Relax."
19:09:58 [nslater]: #info next release window set for 22nd may, bugfix release(s)
19:10:07 [nslater]: we'll want to identify what branches we want to release in a week or so
19:10:20 [nslater]: my current thinking is: anything with changes not yet released
19:10:43 [nslater]: that's it. just a head's up for y'all
19:10:50 [nslater]: next topic should be merges pls


# 3. merge procedure #


# 4. merges #
19:11:36 [nslater]: d'oh


# 5. merge procedure #
19:12:05 [nslater]: i need help with http://wiki.apache.org/couchdb/Merge_Procedure
19:12:22 [nslater]: we need to get this proposal ship shape and then bring it to dev@ for
a decision
19:12:28 [rnewson]: I promised time to help you and we scheduled for tomorrow.
19:12:28 [nslater]: yep, thanks rnewson 
19:12:43 [nslater]: just doing a status run-down. was gonna get to that :P
19:13:05 [nslater]: it would be cool to get this in place before the may release
19:13:20 [nslater]: but my main concern at this stage is getting the 1.4.x branch done and
socialising the workflow changes
19:13:35 [nslater]: it occurred to me that the tl;dr is that release branches as RTC
19:13:44 [tilgovi]: yep
19:13:50 [tilgovi]: that was my big takeaway from the last cycle
19:13:52 [nslater]: so this is an important cultural change we need to pin down as soon as
pos
19:14:20 [nslater]: that's it from me. hopefully should make some good progress with bob tomorrow
19:14:43 [dch]: I'll be around if you need any input too.
19:14:45 [dch]: thanks nslater !
19:14:58 [nslater]: oh wait
19:14:58 [nslater]: let me do an info
19:15:13 [nslater]: #action rnewson and nslater to bring merge procedure proposal to dev@
soon
19:15:20 [nslater]: go ahead


# 6. nebraska bigcouch #
19:16:34 [rnewson]: I'll give an update.
19:17:10 [rnewson]: Since the Nebraska merge was first a topic here, davisp has spent some
time tidying up (rewriting/rebasing) the history of the branch we worked on to make it more
readable and sane.
19:17:28 [rnewson]: but we have not made any further progress than that, we've been tied up
with other Cloudant work.
19:18:25 [nslater]: thanks rnewson! how far away are we, do you think?
19:18:48 [rnewson]: I don't think far, it's just a question of availability. I'll ping davisp
on what's remaining.
19:19:10 [nslater]: cool thx
19:19:25 [dch]: shall I action that?
19:19:40 [nslater]: nah
19:19:48 [nslater]: next topic!! CHOO CHOO
19:19:48 [nslater]: all aboard the topic train


# 7. rcouch #
19:20:03 [nslater]: benoitc: ping
19:20:18 [dch]: I've not seen benoitc about btw, so I think we wll move along in a sec.
19:20:18 [nslater]: i was expecting to work with benoit on the ip clearance stuff
19:20:33 [nslater]: i think benoit was wanting to clean up his code before he prepared the
drop though
19:20:40 [nslater]: so that thing has been stalled for a few weeks
19:20:48 [nslater]: so i dont really know that the status is
19:20:48 [dch]: aight.


# 8. nebraska bigcouch #
19:22:03 [rnewson]: I just wanted to say that, on the merge, we only have some commit history
to tidy up before we can move to IP clearance.
19:22:10 [rnewson]: I thought there was more but I don't think so.
19:22:12 [rnewson]:  /done
19:22:20 [nslater]: ooh cool
19:22:33 [rnewson]: It's playable-with at least.
19:22:48 [rnewson]: and there's a lot there to learn and read for folks that have never seen
bigcouch.
19:22:55 [rnewson]: and a fair bit for those who have.
19:23:03 [dch]: awesome! I think if the final commit  is compleat  we could probably do ip
clearance with a tarball, while the commit tweaking continues.
19:23:10 [rnewson]: Did we clarify that we extracted clustering from cloudant's production
database, not the bigcouch fork?
19:23:33 [dch]: good to know.
19:23:34 [dch]: [next?]
19:23:40 [rnewson]: next


# 9. fauxton #
19:23:50 [benoitc]: i just arrived
19:24:03 [benoitc]: not so much available though
19:24:10 [nslater]: rnewson: interesting....
19:24:20 [dch]: chewbranca: you have the floor mate
19:24:20 [chewbranca]: minor update on fauxton, garren and I are working on adding auth and
reworking some of the routing stuff, current work is in the route-events branch
19:25:49 [chewbranca]: that's all from me on fauxton
19:25:55 [nslater]: chewbranca: cool. i think last time i was in one of these meetings we
talked about inviting collaborators. do you know how far off we are from that? or how far
off we are from being able to merge this in?
19:26:03 [dch]: chewbranca: I am going to be doing a windows build with fauxton soonish, looking
forwards to giving it a spin
19:26:35 [chewbranca]: nslater: we've got open jira tickets and a developer server that makes
it easy to work with now and doesn't require building couchdb
19:26:44 [chewbranca]: collaborators welcome anytime
19:26:58 [nslater]: oh cool. what about sending an invitation to the dev list with details?
19:27:05 [dch]: good idea.
19:27:05 [nslater]: i think there might be a few people interested in collaborating
19:27:28 [chewbranca]: sounds good
19:27:35 [dch]: if you email it, we can get a few of the well-known JS folk to retweet it
too.
19:27:35 [nslater]: sounds like an action item to me!
19:28:13 [dch]: #action chewbranca to invite collaborators to fauxton, dch to get some tweets
on that 
19:28:13 [chewbranca]: nslater: as far as merging things in, fauxton is in master, the route/auth
stuff we're working on is off to the side and doesn't block new people
19:28:35 [nslater]: chewbranca: ah, okay. my mistake
19:28:37 [nslater]: thanks for clarifying
19:28:42 [chewbranca]: dch: great, I'm not familiar with node.js and npm on windows, but happy
to help where I can
19:28:49 [dch]: chewbranca: we will both learn. 
19:29:05 [dch]: I'd love to get a snapshot out for people to play with.
19:29:05 [dch]: anyway


# 10. windows couch #
19:29:35 [chewbranca]: dch yeah, ryan_ramage has http://fauxton.iriscouch.com/fauxton/_design/fauxton/index.html
up and running for people to check out, and you can install it as a couchapp now to any existing
couchdb
19:30:06 [dch]: I'm assuming we will be doing at least 1 or 2 more non-bigcouch releases this
year, so I will start working on a 64-bit windows build
19:30:13 [ryan_ramage]: better link: http://fauxton.iriscouch.com/master/_design/fauxton/index.html
19:30:13 [ryan_ramage]: on master
19:30:35 [dch]: and merge in my glazier toolkit into couchdb itself, as it's pretty stable
now.
19:30:50 [dch]: I don't think it's big enough to warrant ip clearance, and apart from 1 commit
its all mine anyway.
19:30:58 [nslater]: hehe google tells me that page is in norwgeian
19:30:58 [benoitc]: about that i wonder if we could have something like couchdb-sdk
19:31:28 [benoitc]: instead of having tools inside the code, having them in a repo 
19:31:28 [nslater]: dch hmm. is it a significant body of work?
19:31:30 [chewbranca]: ryan_ramage: ahhhh thanks, that would explain why that link I sent
was a bit out of date ;-)
19:31:35 [dch]: I would love wohali to share expertise on MSI packaging (poor man's rpm /
deb for windows)
19:31:50 [benoitc]: idea would be building somekind of toolchain
19:31:53 [dch]: nslater: not in LoC, but if measured in my blood sweat and tears, then you
could fill an ocean.
19:32:05 [benoitc]: somthing like
19:32:43 [dch]: benoitc: do tell
19:32:58 [benoitc]: dch: say you have a toolchain for windows
19:33:05 [benoitc]: but we could have it for other platforms
19:33:08 [benoitc]: like mobiles
19:33:20 [nslater]: dch i'll trust your better judgement on this. but if it is more than just
a large patch (i.e. a body of work developed away from the project lists -- committer or no
committer) then we should probably do clearance for it (which is actually a lightweight process
and takes 3 days for lazy consensus)
19:33:50 [nslater]: can you tell me more about the one commit that is not yours? is that from
a person with an ICLA on file?
19:34:20 [nslater]: or have they indicated clearly that they are happy to contribute the work
to the Apache CouchdB project?
19:34:43 [benoitc]: [dch] reason is https://docs.google.com/file/d/0B2gNkR7hgkIRczU0d3ZqUERfVlE/edit?usp=sharing
19:35:21 [dch]: nslater: it was super small, some doc tweaks and such. I can definitely ask
for an OK to include, but I don't really think its required.
19:35:50 [nslater]: dch like i said, i'll trust your judgement
19:35:53 [dch]: anyway I am getting a bit into the details.
19:35:58 [benoitc]: so idea would be having something more engineered, a set of tool to build
couch on windows but other things
19:35:58 [nslater]: we should be clear on the contribution that wasn't from you though
19:36:05 [dch]: sure
19:36:31 [dch]: benoitc: let's turn that into a separate topic.


# 11. rcouch #
19:36:51 [dch]: benoitc: all yours
19:37:21 [benoitc]: i have been distracted by some  workwork during last month.
19:37:30 [benoitc]: but i'm slowly back on that patch
19:37:43 [benoitc]: since my understanding i sthat ip clearance need the full patch
19:37:50 [benoitc]: if i read well the doc
19:38:05 [benoitc]: it will be finished next week or sooner
19:38:13 [nslater]: benoitc: we just need a copy of the complete code that you're donating
19:38:14 [jan____]: benoitc: you can submit in whichever state thigns are in 
19:38:20 [nslater]: no need to fret the details
19:38:43 [jan____]: benoitc: once the IP clearence is done, we can commit it to a branch,
and you can keep working there befre it goes into arelease
19:38:45 [jan____]: no need to get the code perfect
19:38:50 [benoitc]: nslater: so i am not giving all rcouch for some reasons, i need to extract
the patch completly
19:38:52 [jan____]: …before the clearance
19:38:58 [benoitc]: this patch came along 1 year
19:39:22 [nslater]: okay. i presume it's off-topic to ask why you're not submitting the whole
thing
19:39:28 [nslater]: s,presume,suppose
19:40:13 [benoitc]: becaue theere are a lot more things than just rebar, and some i'm not
sure people want
19:40:23 [benoitc]: also exact.
19:41:50 [benoitc]: we can tal about that later


# 12. mochiweb #
19:42:05 [benoitc]: will make an update on the ml beginning of the next week
19:42:05 [benoitc]: +on
19:42:06 [dch]: cool
19:42:43 [dch]: basically mochiweb is now at v2.6.0 with a couple of interesting fixes for
some tcp problems in R15B* 
19:42:43 [dch]: #link https://github.com/mochi/mochiweb/commit/56a32a0cc091b69374f35aa0787c36c7776ae816
19:42:52 [dch]: which you probably want to be aware of.
19:43:36 [dch]: #info TL;DR we should stick on the v2.4.2 tag which is luckily what I've done,
as after that it drops R14* support, which is a bit draconian.
19:44:06 [dch]: I was going to push the changes to master tonight, I already have them locally,
or do we want to use it tomorrow as a test drive for the merge procedure?
19:44:13 [dch]: nslater: up to you really.
19:45:06 [dch]: ok, next topic


# 13. 1.3.0 post-mortem #
19:45:36 [dch]: tilgovi: I've not been involved in one of these but it sounds like a good
idea. Care to share your thoughts?
19:45:38 [tilgovi]: I'm not sure this a real topic.
19:45:51 [tilgovi]: Because I think we covered a lot of it implicitly.
19:46:06 [tilgovi]: By virtue of our agenda today having topics for the things that we know
are not finalized well enough.
19:46:07 [dch]: so it's more about cleaning up and collecting anything we could improve.
19:46:31 [tilgovi]: So, in the 1.3 cycle I saw things like a merge prodecure in flux and now
we've covered it in depth
19:46:38 [dch]: k
19:46:46 [tilgovi]: Saw constant delays, now we have a calendar and a plan (thanks nslater)
19:47:06 [tilgovi]: so, I guess I put this on here in case there are any good lessons or things
people noticed that we aren't talking about
19:47:28 [tilgovi]: if no one jumps in, we can just move on
19:47:51 [benoitc]: i don't have anything post-mortem
19:47:51 [nslater]: well
19:47:58 [nslater]: delays, that's a tricky one
19:48:06 [nslater]: we had a host of exceptional problems
19:48:06 [nslater]: a bunch of security vulns
19:48:13 [nslater]: and some gnarly bugs
19:48:13 [dch]: well I feel that we tend to work a bit async around release time, but personally
the timetable will help me plan Real Life around Couch Life, hopefully I won't end up borking
binaries again etc etc.
19:48:21 [nslater]: and it happened over the holiday period too
19:48:44 [benoitc]: dch: imo rather than a time tanble we should have a todo
19:49:06 [tilgovi]: hm
19:49:13 [benoitc]: and release when it's about time
19:49:23 [tilgovi]: We are consistentabout linking to open issues by version in JIRA
19:49:28 [tilgovi]: is there something more we should do?
19:49:38 [tilgovi]: Are there todo list items that aren't being captured either by tickets
or actions in these meetings?
19:49:44 [benoitc]: tilgovi: probably tag and assign
19:49:45 [tilgovi]: (and shouldn't be)
19:49:54 [tilgovi]: hm
19:50:21 [tilgovi]: I think we do, no?
19:50:29 [tilgovi]: I'm certainly assigned to some tickets currently
19:50:39 [benoitc]: i don't know about the toodo items you're speaking about :)
19:50:58 [tilgovi]: "dch: imo rather than a time tanble we should have a todo"
19:51:06 [benoitc]: excepty the one for the gsoc or smth
19:51:06 [tilgovi]: I'm trying to figure out what things would be on that
19:51:21 [benoitc]: tilgovi: well the thing devs want to work on
19:51:43 [benoitc]: we should write them somewhere so people know who is working on what
19:51:44 [dch]: I'm happy with jira, and I think we agreed to do time-based releases, if the
branch is usable we ship it, posibly experimental or disabled etc, but we ship on time, every
time, barring security issues and holidays.
19:51:53 [tilgovi]: Are you just talking about signaling what we are working on *now* rather
than what we've assigned ourselves to do (maybe) later?
19:52:06 [dch]: benoitc: perhaps we can do this with some sort of jira ticket query and tags?
19:52:21 [tilgovi]: like, for example, using the time tracking things in JIRA (not sure I
like)
19:52:22 [benoitc]: tilgovi: like i said i don't know about the todo
19:52:29 [benoitc]: also to be clear
19:52:36 [benoitc]: i'm not speaking about issues to solve
19:52:37 [dch]: jira tags are pretty lightweigh
19:52:38 [dch]: jira tags are pretty lightweight
19:53:06 [tilgovi]: ahhhh... benoitc, do you mean roadmap?
19:53:13 [benoitc]: somehow yes
19:53:23 [benoitc]: a roadmap imply a timetable in my head but yes
19:53:44 [tilgovi]: +1
19:53:51 [tilgovi]: we used to have a roadmap somewher
19:53:52 [tilgovi]: I don't know where it is or if it exists
19:54:28 [dch]: tilgovi: see url in #couchdb-dev topic, that perhaps?
19:54:36 [tilgovi]: okay
19:54:52 [tilgovi]: we can turn that into an action item for me if you can think of how to
word it
19:55:06 [tilgovi]: but I'm happy to turn my attention to making the road map more accesible
and getting it more prominently available
19:55:43 [benoitc]: tilgovi: in term of action i will drop a mail on that
19:56:07 [benoitc]: this clearly to let people know who is working on a topic
19:56:08 [tilgovi]: #action benoitc to send mail about roadmap visibility and maintainance
19:58:08 [dch]: k, next topic?


# 14. meeting timing / schedule #
19:58:43 [dch]: ACTION pokes nslater 
19:58:53 [nslater]: hi sorry got distracted with something else
19:59:36 [dch]: heh
19:59:38 [jan____]: ACTION buttons up his shirt
20:00:36 [dch]: well I think we should try and merge or conclude / resolve to consensus the
meeting time / content discussions.
20:00:51 [jan____]: .oø(again)
20:01:28 [dch]: The current time is arbitrary and was based on a time where I could usually
fit it around my commitments, and match up with the New World.
20:02:01 [dch]: I'm still for doing alternating weeks at least for a wee while to see if it
causes grief in practice, and for whom.
20:02:23 [dch]: If it doesn't work for people, we need to hear that concretely and not generic
"might not work out" please.
20:02:54 [jan____]: let’s just do it
20:03:02 [jan____]: I’d love to see JasonSmith in these a little more
20:03:08 [dch]: I'd love to be able to send out a recurring calendar invite once to everybody,
if you know how to do this please let me know.
20:03:15 [jan____]: it worked fine the few times we did them
20:03:32 [dch]: ditto, and also to see if we can encourage e.g. the person in Delhi talking
about couchdb in the bar camp to possibly join in.
20:04:01 [jan____]: +1
20:04:48 [nslater]: ACTION remains silent
20:05:01 [dch]: nslater: benoitc: comments? I'm aware that we likely will not find a perfect
time nor consensus, and I'll happily listen to feedback on how we could improve the meeting.
You hopefully notice I did brief updates this week :D
20:05:24 [benoitc]: if just we can out it outside of the middle of the week. 
20:05:24 [tilgovi]: I have to leave. Later all.
20:05:33 [benoitc]: i can probably be there at any hour
20:05:38 [benoitc]: even 3 am in the morning
20:05:39 [nslater]: cya tilgovi 
20:06:01 [benoitc]: i don't really care about the hour
20:06:01 [dch]: ciao tilgovi!
20:06:31 [jan____]: benoitc: are you saying wednesdays is bad for you?
20:07:18 [benoitc]: yup 
20:07:45 [benoitc]: i'm always side tracked 
20:09:15 [jan____]: benoitc: good to know, what days would be good?
20:09:29 [benoitc]: start or end of the week
20:09:29 [jan____]: I don’t quite care as long as it is not mondays tuesdays or thursdays
:D
20:09:36 [jan____]: and definitely not fridays
20:09:43 [jan____]: damn
20:09:51 [jan____]: I guess I know why we have wed
20:09:58 [benoitc]: end of the week is 3 days for me
20:10:06 [benoitc]: not only friday
20:10:36 [dch]: how about we do a poll with one of those poll-y things to see what works best,
nslater you had an idea about this before I gazumped you and jumped the shark
20:10:44 [jan____]: ah, weekends are unlikely to be successful
20:11:06 [dch]: e..g http://www.doodle.com/bsphe8nehtmx3y8h
20:11:52 [nslater]: we're 11 minutes over time
20:12:06 [benoitc]: anything in the mmiddle has a tendancy to distract me but i'm not the
only one to decide, the doodle is a good idea
20:12:13 [dch]: sure, well let's wrap this up with a couple concrete actions?
20:12:23 [dch]: #action dch to do the doodle poll
20:13:06 [dch]: and I suggest we continue with status quo, for the next monday i.e. til end
may so we have an interim plan.
20:13:06 [dch]: if you're all ok with that?
20:13:36 [dch]: oops ignore monday that was my FAT FINGERS. Wednesday, alternating times.
20:14:00 [nslater]: okay
20:14:13 [nslater]: i would like to propose to benoit that he tries out his own meetings
20:14:21 [nslater]: we dont need to make this meeting all-singing and all-dancing
20:16:06 [dch]: benoitc: or anybody feel free to step in and chair some weeks, I'm no paragon
of meeting protocol.
20:16:36 [dch]: #info continue with alternating meeting times til end May and collect feedback
in between
20:17:36 [dch]: #action next week to be 13h00 UTC (because northern hemisphere is on DST atm)
20:17:36 [dch]: and if we have nothing further, then let's close.
20:17:51 [dch]: oh damn, GSOC update.


# 15. gsoc #
20:18:06 [dch]: TL;DR I will send a note to dev@ and user@ tonight covering where we're at.
20:18:36 [jan____]: dch: 14:00
20:18:36 [dch]: as soon as the email has turned up in the ASF Archives it would be great if
you can re-tweet it, and give helping out creat consideration.
20:18:37 [jan____]: 6 hours earlier
20:19:06 [jan____]: oh
20:19:08 [jan____]: sorry
20:19:14 [jan____]: yeah
20:19:23 [jan____]: I applied the revers UTC has no DST logic
20:19:29 [jan____]: *reverse
20:19:43 [dch]: any thing else folks?
20:20:21 [jan____]: ship it


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