couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Benoit Chesneau <>
Subject Re: Summary of IRC meeting in #couchdb-meeting, Wed Apr 10 19:01:05 2013
Date Thu, 11 Apr 2013 07:51:20 GMT
Of course you can ignore anything that is asked during a meeting. But
what is a point of a meeting then? Note that "obligatory" was under
quotes in my original sentence as well.

A meeting if needed should be constructive. My point when *I*
suggested to have a meeting was to have an informal meetup online from
time to time to understand and know what's going on on the project,
who is doing what with who and where we can find the information on
it. Instead of living in the dark like it was before. Ie share the
project flux between each others, and *just* getting the heartbeat of
the project.,

Not something that was transformed on a procedural meeting with
actions like we have today.  Like in legacy organizations. Something I
don't really support neither I see the interest. People are busy and
what we saw during latest meeting - people saying the same thing and
self-congratulating, mostly - wasn't that useful. For myself I
discovered most of the heart of fauxton during the couchconf and and a
discussionat Erlang Factory than during the meetings. And more
importantly I know where to see and where I can hack or help. When in
the meeting at some point I never asked because of the way it's done.
Not saying that more and more people don't come .

In short I really want to see the meetings as meetups where devs but
also advanced users can come and share what they are working on, what
they want/are trying to do. To initiate the connections between some
if needed. I'm tired to have to watch the twitter timeline for that.
That thing could be set once a month, people come if they can. Mostly
like an offline meetup in a cafe but online. on irc.

But I definitely dislike having a weeklymeeting where one the goal is
to dispatch actions. The mailing list is enough for that and should be
the good relay for that sort of things. But if needed  I am also in
favor of having "group"/"project meeting announced on the mailing
list. For a release meeting having a mail with " Release meeting on
irc: ...."  subject and the body describing the topics. Same for
others features . lt would be more constructive and could happen more

I would be interested by what others think about it.

- benoit

On Wed, Apr 10, 2013 at 10:30 PM, Noah Slater <> wrote:
> The weekly meeting can be short. If there's nothing to report, there's
> nothing to report. No need to waffle on each week with repetitive status
> updates. But I actually don't think we've suffered that at all. Every
> meeting we've had has been useful from my perspective. And it establishes a
> heartbeat for the project, which has so far been very beneficial.
> I will be continuing the weekly meetings for the time being, and I am happy
> to chair them if other people are not interested. This is something I don't
> need consensus for. If you don't want to show up, don't show up. ;)
> Similarly, if anyone else has an idea for a meeting, whether it's a regular
> meeting, or a one-off for a feature, just do it. No need to ask anybody's
> permission. Just like you don't ask permission to send an email to dev@! :)
> Also, monthly "obligatory" meeting won't work because:
>  * Doing it once a month would render the meetings useless for week to week
> work, like releases.
>  * More than likely to miss one or more key people in a single monthly
> meeting.
>  * There is no such thing as "obligatory" on this project... ;)
> On 10 April 2013 21:07, ASF IRC Services <>wrote:
>> Members present: benoitc, wendall911, ryan_ramage, chewbranca, dch
>> ----------------
>> Meeting summary:
>> ----------------
>> 1. Preface
>> 2. meeting frequency
>>   a. benoitc to send a proposal to the mailing list to have a once a month
>> obligatory meeting (ryan_ramage, 2)
>> 3. 1.3.0 release congrats
>>   a. thanks for all the hard work to everyone for getting 1.3.0 out the
>> door (ryan_ramage, 3)
>> 4. javascript test update. View engine update/replacement. Removal of
>> _restart
>>   a. couch 2.0 query server proposal
>>   b. wendall911 to start discussion on dev and add a page in the wiki to
>> start gathering ideas for a formal proposal  (ryan_ramage, 4)
>> 5. erlang r13b04
>>   a. dch to update COUCHDB-1696 and prod the list on it (dch, 5)
>> 6. fauxton
>> 7. GsoC
>>   a. everybody to review the GSoC tickets and filter out or prioritise ->
>> dch to send email (ryan_ramage, 7)
>>   b. dch to mail dev@ and user@ (dch, 7)
>>   c. ryan_ramage to check fauxton builds/deploys (ryan_ramage, 7)
>>   d. offers to mentor are available for GSoC (ryan_ramage, 7)
>> 8. open discussion
>> --------
>> Actions:
>> --------
>> - benoitc to send a proposal to the mailing list to have a once a month
>> obligatory meeting (ryan_ramage, 19:17:28)
>> - wendall911 to start discussion on dev and add a page in the wiki to
>> start gathering ideas for a formal proposal  (ryan_ramage, 19:25:35)
>> - dch to update COUCHDB-1696 and prod the list on it (dch, 19:38:28)
>> - everybody to review the GSoC tickets and filter out or prioritise -> dch
>> to send email (ryan_ramage, 19:55:58)
>> - dch to mail dev@ and user@ (dch, 19:56:00)
>> - ryan_ramage to check fauxton builds/deploys (ryan_ramage, 19:57:37)
>> IRC log follows:
>> # 1. Preface #
>> 19:01:20 [ryan_ramage]: sweet
>> 19:01:58 [ryan_ramage]: hey all, dch can't stick around so has put me in
>> charge of this loose group of awesome
>> 19:02:13 [ryan_ramage]: so send me your topics and we will start in 5 min
>> 19:03:58 [ryan_ramage]: poke chewbranca, davisp garren JasonSmith kocolosk
>> nslater rnewson alice|wl bartender benoitc bigbluehat dpk drsm79
>> 19:04:30 [ryan_ramage]: poke Humbedooh jan____ jeffmjack jeffmjack Kxepal
>> mikewallace sbisbee sbp svnlto vmx Yakulu
>> 19:05:20 [benoitc]: i'bve no other topic except that we should relax the
>> weekly meeting to one/month
>> 19:06:15 [benoitc]: which is a topic.
>> 19:06:58 [ryan_ramage]: ok. will add meeting frequency as a topic
>> 19:07:43 [ryan_ramage]: wow, this is quiet.
>> 19:08:16 [wendall911]: who is MC today?
>> 19:08:35 [wendall911]: who's running this show!
>> 19:08:35 [ryan_ramage]: ok one more minute. I'll add a topic for the 1.3
>> release post congrats
>> 19:08:35 [wendall911]: :)
>> 19:08:36 [ryan_ramage]: wendall911: I am
>> 19:08:50 [ryan_ramage]: got any topics, send 'em my way
>> 19:08:50 [wendall911]: cool, I'd like to add a topic
>> 19:09:28 [wendall911]: Discuss javascript test update. View engine
>> update/replacement. Removal of _restart
>> 19:09:36 [ryan_ramage]: ok cool. noted.
>> 19:09:38 [wendall911]: they are all related for my current issues
>> 19:09:58 [ryan_ramage]: ok, lets start
>> # 2. meeting frequency #
>> 19:10:07 [dch]: topic um r13b04, has anybody tested it lately (another
>> topic)
>> 19:10:35 [ryan_ramage]: benoitc: you would like to suggest from once a
>> week to once a month?
>> 19:10:43 [benoitc]: yes
>> 19:10:58 [ryan_ramage]: ok, sway the crowd....
>> 19:11:20 [benoitc]: cause i personnaly find that there is less interrest
>> tduring these last 3 meetings
>> 19:11:37 [benoitc]: all can bet more energizing with real content
>> 19:11:50 [benoitc]: s/bet/be
>> 19:12:13 [wendall911]: I think this would be fine. I'd like to have the
>> ability to plan special meetings. Where we can discuss specific issues on
>> the side that are outside of what the entire group may want to contribute
>> to.
>> 19:12:35 [benoitc]: i have even heitate to be be online for this one if it
>> was to hear the same thing as l;ast week whoich was the same from last last
>> week
>> 19:13:06 [ryan_ramage]: yes, its usually like many orbs, people are busy
>> and updates start to sound the same
>> 19:13:20 [ryan_ramage]: s/orbs/orgs
>> 19:13:28 [benoitc]: yes
>> 19:13:35 [wendall911]: benoitc: I only showed up today because I wanted
>> some thoughts on where to get started on view engine. List is probably the
>> proper place though.
>> 19:13:50 [wendall911]: Essentially, not necessarily something to meet
>> about.
>> 19:13:58 [dch]: I'm easy but I like the momentum -- it keeps me moving.
>> 19:14:20 [benoitc]: wendall911: and that's perfectly fine. i think my idea
>> would be havin aone "obligatory" meeting /month
>> 19:14:20 [dch]: and there's definite value in being in the same place/time
>> occasionally, but it doesn't have to be weekly. We just started with that.
>> 19:14:29 [ryan_ramage]: maybe the format and frequency are related. How
>> topics are kept, tracked and followed up.
>> 19:14:38 [benoitc]: and meeting orginezed by informal group or when we
>> need an instant discussion
>> 19:15:13 [wendall911]: +1 on benoitc's suggestion
>> 19:15:50 [wendall911]: benoitc: It might be nice for these side
>> discussions to be announced on the list. It would give anyone a chance to
>> attend if they wanted.
>> 19:15:58 [benoitc]: since it need aa concensus ,i propose to propose that
>> on the ml
>> 19:16:23 [benoitc]: wendall911: yes I keep saying that since awhile, also
>> then the discussion can start off
>> 19:16:43 [ryan_ramage]: yes so benoitc can you send a ml the proposal.
>> 19:16:43 [benoitc]: and other things related (like thinking on what we can
>> add to a topic)
>> 19:16:45 [ryan_ramage]: I would back it
>> 19:16:50 [benoitc]: ryan_ramage: yes
>> 19:17:15 [benoitc]: will do tomorrow
>> 19:17:21 [benoitc]: put it as an action
>> 19:17:28 [ryan_ramage]: #action benoitc to send a proposal to the mailing
>> list to have a once a month obligatory meeting
>> 19:17:52 [ryan_ramage]: ok, shall we move on?
>> 19:18:05 [benoitc]:  i'm done myself
>> # 3. 1.3.0 release congrats #
>> 19:18:35 [ryan_ramage]: awesome work everyone
>> 19:18:50 [ryan_ramage]: a very big release, to be proud of
>> 19:19:43 [ryan_ramage]: #info thanks for all the hard work to everyone for
>> getting 1.3.0 out the door
>> 19:20:13 [ryan_ramage]: ok, prolly nothing actionible hereā€¦.
>> # 4. javascript test update. View engine update/replacement. Removal of
>> _restart #
>> 19:20:51 [ryan_ramage]: go ahead wendall911
>> 19:20:59 [wendall911]: I think I should move this to the list and propose
>> a side meeting
>> 19:21:30 [wendall911]: It's a little complicated, and I've been combing
>> the lists and gathering specific comments about each issue.
>> 19:21:59 [wendall911]: There are several ideas floating around right now,
>> but what I want to do is arrive at some kind of a consensus so I can start
>> coding on this
>> 19:22:20 [benoitc]: wendall911: just to add some  food fot thoughts ,
>> there ois someone that posted a proposal on g+
>> 19:22:28 [benoitc]: for the view engine protocol
>> 19:22:35 [wendall911]: benoitc: ok, didn't see that
>> 19:22:45 [wendall911]: benoitc: that's the exact reason this should be
>> centered around a discussion
>> 19:22:58 [ryan_ramage]: benoitc: I am going to put on my apache nslater
>> hat and say send the link/discussion to the ml
>> 19:23:07 [benoitc]:
>> 19:23:29 [benoitc]: yeah not time for that these days sorry
>> 19:23:58 [benoitc]: wendall911: i would say ml + wiki
>> 19:24:05 [wendall911]: Nice, I think I'm mostly concerned at this point
>> with the underlying libs used for the view engine. Nodejs, V8 or
>> spidermonkey
>> 19:24:13 [benoitc]: imo would be good if we come with a full proposal
>> 19:24:20 [wendall911]: agreed
>> 19:24:20 [benoitc]: i've also spome doc i've zritten sometime ago
>> 19:24:35 [ryan_ramage]: #info couch 2.0 query server proposal
>> 19:24:50 [benoitc]: i think we should first discuss about th eprotocol
>> 19:24:50 [wendall911]: benoitc: ok, I'll start a discussion on dev and add
>> a page in the wiki to start gathering ideas for a formal proposal
>> 19:24:58 [benoitc]: then we could have alternative implementation
>> 19:24:59 [benoitc]: s
>> 19:25:20 [benoitc]: also relate questions:
>> 19:25:28 [dch]: yeah maybe ask Samuel if he minds moving this to the wiki
>> for moar discussion
>> 19:25:29 [benoitc]: how the couchapp engine enter in that ?
>> 19:25:35 [ryan_ramage]: #action wendall911 to start discussion on dev and
>> add a page in the wiki to start gathering ideas for a formal proposal
>> 19:25:37 [benoitc]: couldn't we simpliify th ething ...
>> 19:25:50 [benoitc]: chewbranca: has also started a thing or two
>> 19:25:58 [benoitc]: dch: he did
>> 19:26:07 [benoitc]: there is a post on the ml if I remember
>> 19:26:28 [wendall911]: I'll dig through the archive a bit more and add to
>> my email and wiki
>> 19:26:30 [chewbranca]: hey, just getting here
>> 19:26:43 [wendall911]: I have about 10 threads so far with comments around
>> this issue that add value
>> 19:26:43 [ryan_ramage]: welcome chewbranca
>> 19:27:05 [ryan_ramage]: wendall911: yeah would be good to gather on one
>> place on the wiki
>> 19:27:14 [chewbranca]:  ACTION waves at ryan_ramage
>> 19:27:58 [ryan_ramage]: chewbranca: I've been asked to run the show today,
>> can I slot you in for some fauxton topic time?
>> 19:28:13 [chewbranca]: sure, I can give a quick update on that
>> 19:28:23 [ryan_ramage]: wendall911: so we good to move on?
>> 19:28:28 [wendall911]: ok, I think that's all for now on this. I'll
>> followup on ML and wiki
>> 19:28:28 [chewbranca]: I'm also very interested in alternative query
>> engines
>> 19:28:50 [chewbranca]: I would like to also propose or talk about
>> separating couchapps from view engines
>> 19:28:51 [ryan_ramage]: [off[ dch noted
>> 19:28:58 [ryan_ramage]: bah
>> 19:29:21 [benoitc]: chewbranca: can we put last topic on the l ?
>> 19:29:23 [benoitc]: i have to go
>> 19:29:29 [ryan_ramage]: ok, thanks benoitc
>> 19:29:35 [chewbranca]: ok yeah, lets do that
>> 19:29:43 [chewbranca]: I've got some code to link out for that too
>> 19:29:50 [benoitc]: cool thx
>> 19:29:50 [benoitc]: also nice
>> 19:30:28 [ryan_ramage]: ok, moving on I thinkā€¦
>> 19:31:00 [chewbranca]: sounds good
>> # 5. erlang r13b04 #
>> 19:31:06 [ryan_ramage]: dch: go
>> 19:32:22 [ryan_ramage]: or anyoneā€¦I think the idea wasā€¦ has anybody
>> tested it lately?
>> 19:32:29 [dch]: sorry, here.
>> 19:33:13 [dch]: yup this turned up when I was fiddling with migrating to
>> mochiweb 2.4.2, I had *lots* of failures, this on a std 12.04 ubuntu off
>> ec2.
>> 19:33:30 [dch]: so the question is, basically "does anybody use R13B04 at
>> all?"
>> 19:34:20 [wendall911]: dch: this is mainly centered around *very* old
>> linux distros. Centos 5 and ~3 yr old debian/ubuntu
>> 19:34:50 [dch]: wendall911: right, that's what I'm thinking. lucid for
>> example or pre squeeze debian.
>> 19:35:28 [wendall911]: I think the question is more a matter of how long
>> we can add support for very old distributions. Those are still on 1.0 or
>> earlier, so really not an issue, imo
>> 19:35:43 [dch]: well I think we can roll the discussion into COUCHDB-1696
>> (the mochiweb migration ticket)
>> 19:35:43 [wendall911]: and they are never going to see even 1.2.x
>> 19:36:20 [dch]: ok that covers it then, thanks wendall911
>> 19:36:43 [wendall911]: so it's really not a problem as I see it dropping
>> support for it, as it's barely relevant. Anyone using Centos 5 for example,
>> and needing 1.3.0 will need to probably upgrade a dozen or more packages
>> themselves anyhow
>> 19:36:58 [wendall911]: dch: np
>> 19:37:35 [ryan_ramage]: ok, can someone sum up so I can put an info or
>> action down?
>> 19:38:23 [wendall911]: One small thing to add here
>> 19:38:28 [dch]: #action dch to update COUCHDB-1696 and prod the list on it
>> 19:38:35 [wendall911]: I brought this up on the list yesterday
>> 19:39:13 [wendall911]: This could represent a breaking change for
>> packagers. So, say we strip support from 1.3.x, it shouldn't get a patch
>> bump, as it's a major dependency change.
>> 19:39:20 [wendall911]: at least not from my perspective
>> 19:40:20 [wendall911]: I don't think the api is the only important piece
>> relevant to the semantic version. Might be wrong here, but these changes
>> have implications for someone.
>> 19:40:43 [wendall911]: and if it's a major version bump in a lib, or
>> dropping support, can be problematic downstream
>> 19:41:22 [dch]: I think the most likely move is that we set spidermonkey
>> at 1.8.5 (so we can drop a lot of crufty code & autotools checks) and ditto
>> for R14+ the same.
>> 19:41:50 [dch]: for most current distros this won't break anything.
>> 19:42:08 [dch]: and for older ones I assume packagers will not be bumping
>> them anyway, unless via backports/epel.
>> 19:42:13 [wendall911]: right, but dropping spidermonkey 1.7 is
>> problematic, as it's still deemed relevant in newer distributions
>> 19:42:28 [wendall911]: no, Centos 6 is the current release and it ships
>> with 1.7
>> 19:42:31 [dch]: SRSLY.
>> 19:42:35 [wendall911]: RHEL, etc
>> 19:42:38 [dch]: oh well, noted!
>> 19:42:44 [wendall911]: yeah, spidermonkey doesn't have a history of
>> encouraging updating ;)
>> 19:43:40 [dch]: next topic? I think we are done
>> 19:43:47 [ryan_ramage]: ok
>> # 6. fauxton #
>> 19:44:24 [ryan_ramage]: I should have some updates, but I sadly say I do
>> not. chewbranca ?
>> 19:45:16 [wendall911]: I have a question about coding style. Is this
>> documented somewhere?
>> 19:45:54 [wendall911]: I can rant for a while about nodejs style of
>> javascript coding, but just wanting to know if we have a style guide.
>> 19:46:09 [chewbranca]: hey
>> 19:46:24 [chewbranca]: yeah, partially asleep right now
>> 19:46:54 [chewbranca]: so I've got a branch pushed out that's rebuilding
>> the routing system right now, specifically to address problems with full
>> page reloads and shared state between subtabs and what not
>> 19:47:32 [chewbranca]: if you haven't used fauxton yet, that probably
>> doesn't make sense, but if you're familiar with it, you will have seen the
>> random flickers and full page reloads for things that should only change
>> part of the page
>> 19:50:09 [ryan_ramage]: only thing I want to add is to make sure to add
>> *fauxton* somewhere in the branch name so others can play when ci pushes
>> out a build
>> 19:50:46 [ryan_ramage]: chewbranca: do we have a style guide somewhere?
>> wendall911 was asking
>> 19:51:27 [ryan_ramage]: ok, trying to make this quickā€¦moving on
>> # 7. GsoC #
>> 19:52:11 [ryan_ramage]: dch: ?
>> 19:52:39 [dch]: So we got into GSoC both as a project and as the
>> foundation.
>> 19:53:29 [dch]: I'd love if over the next 2 weeks we can do some
>> propaganda (twitter, emails, local user groups etc) to invite students to
>> join, and bring some more people into couch-love.
>> 19:53:38 [dch]: so call for help on 2 folds;
>> 19:54:13 [dch]: 1. everybody to review the GSoC tickets and filter out or
>> prioritise -> dch to send email
>> 19:55:43 [dch]: 2. offer to mentor.
>> 19:55:51 [dch]: that's basically it, a heads up.
>> 19:55:58 [ryan_ramage]: #action everybody to review the GSoC tickets and
>> filter out or prioritise -> dch to send email
>> 19:56:00 [dch]: #action dch to mail dev@ and user@
>> 19:56:21 [chewbranca]: for instance, I added pagination to _all_dbs, but
>> because of this problem, every single page you go through would do a full
>> page reload, which is a mistake for _all_dbs, because it does not support
>> any proper limit or range functions, and you get every single value at once
>> 19:56:21 [chewbranca]: so we don't want to reload all the things, but just
>> change a few
>> 19:56:21 [chewbranca]: anyways, started working on that this week, garren
>> has been doing some great auth work that's coming along, and we've been
>> doing a lot of misc cleanup to get things working well in fauxton
>> 19:56:21 [chewbranca]: we would love to get more people using fauxton and
>> doing some QA, if you find oddities, please file a JIRA ticket and we'll
>> get it cleaned up
>> 19:56:21 [chewbranca]: also, ryan_ramage, it doesn't look like the app on
>> has been updated in a while, any ideas on that? I
>> saw a failed test email, but haven't had a chance to look at it yet
>> 19:56:21 [chewbranca]: and wendall911, back to your actual question, there
>> isn't an official style guide, but I'm partial to most things in:
>> 19:56:21 [chewbranca]: that's it from me
>> 19:56:28 [chewbranca]: wow... netsplit
>> 19:56:36 [chewbranca]: did any of my messages get through?
>> 19:56:37 [wendall911]: dch: where is the best place to do this right now?
>> G+, cause hacker news and reddit aren't really producing much relevant or
>> useful discussion lately.
>> 19:56:38 [dch]: lols so thats what happens
>> 19:56:58 [ryan_ramage]: wowza. time machine
>> 19:56:58 [dch]: wendall911: I guess where ever you feel that works
>> 19:57:37 [ryan_ramage]: #action ryan_ramage to check fauxton builds/deploys
>> 19:57:37 [wendall911]: chewbranca: awesome. I think I agree with most of
>> that :)
>> 19:57:58 [chewbranca]: did a flood of messages from me eventually show up?
>> 19:58:21 [ryan_ramage]: chewbranca: yeah. I think they came
>> 19:58:43 [chewbranca]: wendall911: yeah, its kind of a random guide, but
>> it think its a pretty good list
>> 19:58:46 [chewbranca]: ryan_ramage: cool, thanks
>> 19:58:59 [ryan_ramage]: #info offers to mentor are available for GSoC
>> 19:59:05 [chewbranca]: wendall911: yeah good idea
>> 19:59:21 [wendall911]: dch: ok, I'd love to get some attention, just not
>> sure where all the cool kids hang out nowadays
>> 19:59:23 [ryan_ramage]: ok dch, closing GSoC topic
>> 19:59:29 [dch]: ACK.
>> 19:59:43 [dch]: They hang out in #couchdb-dev :D
>> 19:59:45 [wendall911]: hahaha
>> # 8. open discussion #
>> 19:59:51 [ryan_ramage]: anything else?
>> 19:59:59 [wendall911]: I have one comment
>> 20:00:06 [ryan_ramage]: 1 min closing bell
>> 20:00:13 [ryan_ramage]: wendall911: shoot
>> 20:00:13 [wendall911]: I'd like to see the failing test on master get fixed
>> 20:00:37 [wendall911]: so Travis can quit bugging me :)
>> 20:00:50 [wendall911]: no discussion needed
>> 20:01:35 [ryan_ramage]: wendall911: when you say failing test, is it only
>> one that is failing?
>> 20:02:03 [wendall911]: make distcheck
>> 20:02:11 [wendall911]:
>> 20:02:20 [wendall911]: oh, not a test, just distcheck is failing
>> 20:02:33 [wendall911]: For some reason I thought it was a test
>> 20:02:33 [chewbranca]: ahhh yeah looks like its failing on
>> 20:02:41 [wendall911]: I never run distcheck, only make && make check
>> 20:02:42 [dch]: wendall911: nice piece of work on this btw
>> 20:03:04 [wendall911]: dch: no problem, just a start :)
>> 20:03:12 [chewbranca]: odd, looks like simon added that file to
>> 20:03:48 [wendall911]: I have one other makefile issue that got pushed out
>> in 1.3.0 I didn't catch
>> 20:04:03 [wendall911]: I'll file a ticket and get it fixed later today.
>> Was a minor issue with the manifest
>> 20:04:26 [ryan_ramage]: ok gonna close this off
> --
> NS

View raw message