couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Robert Samuel Newson <rnew...@apache.org>
Subject Re: Summary of IRC Meeting in #couchdb-meeting
Date Wed, 28 May 2014 21:17:13 GMT

I based the rebar work on yours, though I did remove the cross-compile/static build bits for
simplicity. I am also pretty confident that my work does not work on all the platforms that
rebar.config.script implies. It’s certainly a work-in-progress. Since it *does* build the
bigcouch branch correctly, it now serves a valid prototype for any work you might wish to
do on it.

for the NIF’s, I’ve no preference on the matter, as long as comparison is still fast.
Make the replacement if you wish to.

couch_mrview, I’ve not looked at your changes deeply but it does seem a simple merge.

I am terrible at roadmaps and timeframes, I think that’s no secret. I can summarize current
status as follows: the 1843-feature-bigcouch branch is in good shape for testing, for finding
regressions. I’m now working to get authentication working against a clustered/sharded _users
database rather than calling the node local one only. Like in bigcouch, we’ll call one and,
on a 404, call the other. We could revise this going forward. After that is achieved, I’ll
be working to import the work that Cloudant has done since the original Nebraska merge, our
full year of code fixes and improvements, in a further IP clearance round. 

B.

On 28 May 2014, at 21:40, Benoit Chesneau <bchesneau@gmail.com> wrote:

> As promised last day on irc, answering about the merge. .
> 
> There some redundant things in bigcouch and rcouch branches. Both have
> different builds tricks and different changes which will require some
> decisions and synchronisations I guess. Some features need probably to
> be freezed before to go further. Some questions to start:
> 
> - what do we do with the build? Lot of thing in the bigcouch branch have
>  been added recently (compared to the old public repo in github) that
> are also in rcouch. So we need to make some decisions. For the test
> suite and the release creation. The rcouch build actually differs from
> bigcouch  by targetting the cross-compilation and proposing hooks to
> customise the build. It also offer the possibility to build a release
> statically. There are good things to take from both. How do you think we
> can work on this all together?
> 
> - couch_ejson_compare and couch_drv has been removed in rcouch. Both
>  have been replaced by one nif couch_compare. It can probably fix the
> issue seen on the ml recently. Should I make the replacement?
> 
> - couch_mrview changes can probably be merged easily. Though it will
>  requires some coordination. I am not sure if the format changed in
> bigcouch. Or how do we handle format changes (there are more trees in
> the implementation of rcouch) ? Can the "changes" code be reused in term
> of code (hard question i know), should i do that?
> 
> - What is the expected roadmap? When the changes should be expected? do
>  we have a time window?
> 
> 
> - benoit
> 
> 
> On Wed, May 7, 2014 at 9:57 PM, ASF IRC Bot <asfbot@urd.zones.apache.org>wrote:
> 
>> Summary of IRC Meeting in #couchdb-meeting at Wed May  7 19:10:54 2014:
>> 
>> Attendees: Wohali, wendall911, nslater, deathbear, Kxepal, jan____,
>> chewbranca, rnewson
>> 
>> - Preface
>> - 1.6.0
>>  - Info: 1.6.0-rc.5 is out for testing, so for all systems GO
>>  - Info: 1.6.x CI is all green, formal vote coming this week
>> - rcouch merge
>>  - Info: Wohali to +1 the code in its current state and OK rcouch for
>> integration branch merge
>> - bigcouch merge
>>  - Info: BigCouch: massive progress this week. Soon branches will be
>> rewritten to clean up history. After that the call will go out for testing.
>> - Fauxton
>>  - Info: agreement that Fauxton replaces Futon in CouchDB 2.0 (post
>> merge) and that resolution of COUCHDB-1954 prior to that is critical
>>  - Info: Kxepal to start dev@ thread about this and ensuring release
>> preparation
>> - Bylaws / CoC / Diversity
>>  - Info: reminder to review nslater's draft ASAP:
>> https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=40511017
>>  - Info: comments should go to dev@ ML
>>  - Info: Wohali to post initial CoC and Diversity draft text this week
>> - New Business
>>  - Action: jan____ to document and publicize the current
>> state-of-the-world on what the project release artefacts are
>>  - Action: everyone to review nslater's bylaws @
>> https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=40511017
>>  - Info: jan____ to document and publicize the current state-of-the-world
>> on what the project release artefacts are
>> 
>> 
>> IRC log follows:
>> 
>> ## Preface ##
>> ## 1.6.0 ##
>> [Wed May  7 19:11:23 2014] <Wohali>: We have rc-5 out and I think it looks
>> good, yes?
>> [Wed May  7 19:11:47 2014] <jan____>: sounds good to me
>> [Wed May  7 19:11:48 2014] <jan____>: Wohali: I can run the meeting if you
>> are too busy
>> [Wed May  7 19:11:51 2014] <deathbear>: when is 1.6.0 scheduled to go out?
>> [Wed May  7 19:11:55 2014] <Wohali>: jan____: no it's fine
>> [Wed May  7 19:12:00 2014] <jan____>: ok
>> [Wed May  7 19:12:01 2014] <Wohali>: 1.6.0 testing is now for rc-5
>> [Wed May  7 19:12:24 2014] <Wohali>: #info 1.6.0-rc.5 is out for testing,
>> so for all systems GO
>> [Wed May  7 19:12:24 2014] <deathbear>: do me and garren need to do
>> another build of fauxton?
>> [Wed May  7 19:12:49 2014] <Wohali>: deathbear: at this point i think it's
>> locked down for 1.6.0, this release should have gone out month(s) ago
>> [Wed May  7 19:13:14 2014] <Wohali>: we're only resolving issues wiht the
>> release at this point
>> [Wed May  7 19:13:18 2014] <Wohali>: as in, tests that fail
>> [Wed May  7 19:13:20 2014] <jan____>: 1.6.x CI is all greed.Will cast a
>> formal vote later tonight or tomorrow.
>> [Wed May  7 19:13:21 2014] <Wohali>: or build errors
>> [Wed May  7 19:13:31 2014] <Wohali>: awesome
>> [Wed May  7 19:13:50 2014] <Wohali>: #info 1.6.x CI is all green, formal
>> vote coming this week
>> [Wed May  7 19:13:52 2014] <Wohali>: hooray!
>> [Wed May  7 19:13:57 2014] <Wohali>: anything else 1.6.x related?
>> [Wed May  7 19:14:18 2014] <Wohali>: moving on
>> ## rcouch merge ##
>> [Wed May  7 19:15:03 2014] <Wohali>: My limited testing and Alex Shorin's
>> testing shows that it's functional. Haven't seen any additional feedback
>> from the community, tweets or mailing list. Going to propose we get this
>> onto the integration branch given lack of additional info.
>> [Wed May  7 19:15:07 2014] <Wohali>: benoitc: are you here?
>> [Wed May  7 19:16:03 2014] <Wohali>: will give him another 60s but I think
>> that's all there is
>> [Wed May  7 19:16:25 2014] <Wohali>: #info Wohali to +1 the code in its
>> current state and OK rcouch for integration branch merge
>> [Wed May  7 19:16:47 2014] <Wohali>: ok moving on
>> ## bigcouch merge ##
>> [Wed May  7 19:16:53 2014] <Wohali>: rnewson take it away
>> [Wed May  7 19:17:05 2014] <rnewson>: hi
>> [Wed May  7 19:17:12 2014] <rnewson>: so I made some strong progress this
>> week on the merge.
>> [Wed May  7 19:17:24 2014] <jan____>: yaaay
>> [Wed May  7 19:17:31 2014] <rnewson>: couchdb.git's 1843-feature-bigcouch
>> now points to 'master' branches of many subprojects
>> [Wed May  7 19:17:31 2014] <Wohali>:  ACTION applauds
>> [Wed May  7 19:17:45 2014] <rnewson>: each of which represent the merge of
>> bigcouch and couchdb-from-feb-2014
>> [Wed May  7 19:18:05 2014] <rnewson>: I hope to push the last big piece
>> (couch_replicator) tomorrow once I squash a bug.
>> [Wed May  7 19:18:18 2014] <Wohali>: I think I saw chewbranca's view work
>> go by as well?
>> [Wed May  7 19:18:21 2014] <rnewson>: the other 'master' branches to
>> create are in places with no or little merge work
>> [Wed May  7 19:18:25 2014] <rnewson>: yes, views work now.
>> [Wed May  7 19:18:28 2014] <rnewson>: clustered views
>> [Wed May  7 19:18:44 2014] <Wohali>: awesome, i know a lot of people will
>> be very excited for that
>> [Wed May  7 19:18:45 2014] <rnewson>: the new couch_replicator will be an
>> improvement over what Cloudant have today.
>> [Wed May  7 19:18:53 2014] <rnewson>: yes
>> [Wed May  7 19:19:05 2014] <rnewson>: It would useful if folks spent a
>> little time building and trying things
>> [Wed May  7 19:19:06 2014] <Wohali>: What are the next steps once
>> couch_replicator goes through?
>> [Wed May  7 19:19:11 2014] <Wohali>: testing
>> [Wed May  7 19:19:17 2014] <rnewson>: the sooner I find out about
>> regressions, the faster I can clean up histor
>> [Wed May  7 19:19:34 2014] <rnewson>: I will be rewriting any branch
>> (except rcouch) at will until I say otherwise, though, so be careful
>> [Wed May  7 19:19:47 2014] <rnewson>: notably, I will rewrite 'master' of
>> couch_mrview as soon as the git hooks are disabled
>> [Wed May  7 19:19:51 2014] <rnewson>: it's important to have clean history.
>> [Wed May  7 19:20:04 2014] <rnewson>: after couch_replicator, it's the
>> other sub projects, then it's testing.
>> [Wed May  7 19:20:14 2014] <Wohali>: rnewson: is the intent after that to
>> then merge what you've got onto the proposed integration branch and deal
>> with the rcouch bits?
>> [Wed May  7 19:20:34 2014] <rnewson>: Oh, I also, after talking with
>> chewbranca, restricted the supported erlang versions to the ones we expect
>> to work
>> [Wed May  7 19:20:42 2014] <Wohali>: ah yes I'll add that as another topic
>> [Wed May  7 19:20:43 2014] <rnewson>: I am not merging rcouch, benoitc is.
>> [Wed May  7 19:20:56 2014] <chewbranca>: yeah, going to send an email to
>> dev@ today about erlang versions
>> [Wed May  7 19:21:07 2014] <Wohali>: chewbranca: hold please
>> [Wed May  7 19:21:19 2014] <Wohali>: so in the ticket that was referenced
>> it showed both rcouch and bigcouch hitting this new integration branch,
>> then that being merged to master
>> [Wed May  7 19:21:33 2014] <Wohali>: is that not the plan?
>> [Wed May  7 19:21:46 2014] <Wohali>:
>> https://issues.apache.org/jira/browse/COUCHDB-1999
>> [Wed May  7 19:21:48 2014] <Wohali>: this ascii art
>> [Wed May  7 19:21:53 2014] <rnewson>: I think it is, but the merge effort
>> to get rcouch does not fall on me.
>> [Wed May  7 19:22:11 2014] <Wohali>: well, it shows rcouch hitting
>> integration before nebraska, that's all I mean
>> [Wed May  7 19:22:18 2014] <rnewson>: given we have lots of (new) repos,
>> the integration branch can be 'master'.
>> [Wed May  7 19:22:25 2014] <Wohali>: ok.
>> [Wed May  7 19:22:33 2014] <rnewson>: yes, that did not happen.
>> [Wed May  7 19:22:47 2014] <Wohali>: I just want to avoid a race to "who
>> merges first" and sticking the other person with difficult work
>> [Wed May  7 19:22:51 2014] <rnewson>: the rcouch branches sit in each
>> repo, they "just" need merging down to master, once I'm done.
>> [Wed May  7 19:23:09 2014] <Wohali>: ok, so in your proposal "nebraska"
>> hits first, then rcouch follows
>> [Wed May  7 19:23:16 2014] <rnewson>: or benoitc can take a copy of
>> current master, merge rcouch into that, and we'll merge that to master when
>> it's not being rewritten
>> [Wed May  7 19:23:26 2014] <Wohali>: got it
>> [Wed May  7 19:23:36 2014] <Wohali>: actually, the issue doesn't specify a
>> strict ordering
>> [Wed May  7 19:23:47 2014] <Wohali>: so that's fine with me
>> [Wed May  7 19:23:58 2014] <Wohali>: can you help me summarize that into
>> some #info points?
>> [Wed May  7 19:24:04 2014] <rnewson>: I'm happy to help with rcouch
>> merging to a degree, but it's not my task.
>> [Wed May  7 19:24:09 2014] <Wohali>: yes, understood
>> [Wed May  7 19:24:13 2014] <rnewson>: that is, any help I give is my own
>> time.
>> [Wed May  7 19:25:08 2014] <Wohali>: ok so i think i am hearing
>> [Wed May  7 19:26:01 2014] <Wohali>: massive progress this week, branches
>> will be rewritten to clean up history, need help testing.
>> [Wed May  7 19:26:16 2014] <rnewson>: only one subproj master needs a
>> force-push
>> [Wed May  7 19:26:20 2014] <Wohali>: the only thing that differs in what
>> you just said vs. what was agreed to in COUCHDB-1999 is using master as the
>> integration branch
>> [Wed May  7 19:26:20 2014] <rnewson>: aka, I only messed up once.
>> [Wed May  7 19:26:25 2014] <rnewson>: (that I know of)
>> [Wed May  7 19:26:26 2014] <Wohali>: does anyone here have issues with
>> that?
>> [Wed May  7 19:27:27 2014] <Wohali>: #info BigCouch: massive progress this
>> week. Soon branches will be rewritten to clean up history. After that the
>> call will go out for testing.
>> [Wed May  7 19:28:02 2014] <Wohali>: rnewson: if you are planning to merge
>> to master directly can you update 1999?
>> [Wed May  7 19:28:20 2014] <Wohali>: as I am hearing no objections, but
>> there are many not present and dev@ should be notified
>> [Wed May  7 19:28:34 2014] <wendall911>: Quick question.
>> [Wed May  7 19:28:50 2014] <jan____>: Wohali: yeah +1, I’m not deep enough
>> in to make a statement either way
>> [Wed May  7 19:28:53 2014] <wendall911>: What will the next version be
>> once this is merged?
>> [Wed May  7 19:29:00 2014] <rnewson>: 2.0
>> [Wed May  7 19:29:19 2014] <Wohali>: yeah, semver implies 2.0
>> [Wed May  7 19:29:35 2014] <rnewson>: well, 2.0.0, but yes.
>> [Wed May  7 19:31:57 2014] <Wohali>: ok
>> [Wed May  7 19:32:05 2014] <Wohali>: any other topics?
>> [Wed May  7 19:32:14 2014] <Wohali>: on bigcouch merge
>> ## Fauxton ##
>> [Wed May  7 19:32:38 2014] <Wohali>: wendall911: you have the floor
>> [Wed May  7 19:33:07 2014] <wendall911>: Oh, I just saw that 1.6.0 was
>> missing some things from master, including a couple important Fauxton
>> updates
>> [Wed May  7 19:33:29 2014] <Wohali>: gotcha
>> [Wed May  7 19:33:41 2014] <wendall911>: To not hold up 1.6.0 release,
>> this stuff was left out. There should probably be a 1.6.1 update shortly
>> after 1.6.0 ships to address this.
>> [Wed May  7 19:33:52 2014] <Wohali>: djc isn't here but i know he has been
>> eager to avoid creep
>> [Wed May  7 19:34:00 2014] <Wohali>: we have an rc.5 cut already and it's
>> passing all the tests
>> [Wed May  7 19:34:02 2014] <Kxepal>: wendall911: the most important
>> updates are already landed
>> [Wed May  7 19:34:15 2014] <Wohali>: wendall911: which changes
>> specifically?
>> [Wed May  7 19:34:29 2014] <wendall911>: Mostly the init script change
>> you'd made Wohali
>> [Wed May  7 19:34:40 2014] <wendall911>: I'd like to see some testing on
>> that, and anything else relevant
>> [Wed May  7 19:34:43 2014] <Kxepal>: but I agree, that Fauxton@master has
>> already a little bit more than Fauxton@1.6.x
>> [Wed May  7 19:35:02 2014] <jan____>: wendall911: the release schedule
>> says 4 weeks after the previous release. unless there are security issues
>> to address, that’s when the 1.6.1 can go out
>> [Wed May  7 19:35:05 2014] <Wohali>: wendall911: that init script change
>> is in 1.6.0 and has been reviewed/tested
>> [Wed May  7 19:35:20 2014] <wendall911>: Anyhow, I guess it could wait a
>> while, just didn't know how far we wanted to let master diverge considering
>> the upcoming merge
>> [Wed May  7 19:35:40 2014] <wendall911>: jan____: ok, that makes sense
>> [Wed May  7 19:35:54 2014] <Wohali>: yeah or if the release is DOA
>> [Wed May  7 19:35:56 2014] <wendall911>: I think that is a reasonable
>> window, I just didn't know the particulars
>> [Wed May  7 19:35:59 2014] <Wohali>: but we're not seeing that
>> [Wed May  7 19:36:18 2014] <Wohali>: cheers, glad to hear fauxton
>> continues improvement
>> [Wed May  7 19:36:28 2014] <Wohali>: anything else fauxton related?
>> wendall? robertkowalski? deathbear ?
>> [Wed May  7 19:36:33 2014] <wendall911>: Ok, I don't have anything else to
>> add
>> [Wed May  7 19:37:29 2014] <Kxepal>: COUCHDB-1954
>> [Wed May  7 19:38:51 2014] <Kxepal>: do we agree to migrate from Futon to
>> Fauxton on 2.0 release or still have to make a plan and migration decisions
>> for 1.x release?
>> [Wed May  7 19:38:59 2014] <wendall911>: Kxepal: I'd like clarification on
>> that as well, maybe ask in the dev list?
>> [Wed May  7 19:39:07 2014] <Wohali>: Definitely a topic for dev@
>> [Wed May  7 19:39:15 2014] <Wohali>: my feeling is that 2.0 would not
>> include futon
>> [Wed May  7 19:39:22 2014] <Wohali>: but it's just a feeling
>> [Wed May  7 19:39:47 2014] <Wohali>: but yes having some sort of nice
>> resolution for things would be useful
>> [Wed May  7 19:39:50 2014] <Kxepal>: Wohali: actually, I also feel that
>> migration to Fauxton with 2.0 would be the easiest way
>> [Wed May  7 19:39:59 2014] <jan____>: yeah, we may want to swap in fauxton
>> ina  1.7.0 release and then have it ready for 2.0.0, or just got for 2.0.0
>> right away
>> [Wed May  7 19:40:19 2014] <Wohali>: Kxepal: Move that bug to fix version
>> 2.0.0 and if necessary we can pull it back to 1.7.0. also can you start the
>> dev@ thread on it?
>> [Wed May  7 19:40:45 2014] <wendall911>: From a marketing perspective,
>> having a flashy new UI with a major version might be something to rally
>> around
>> [Wed May  7 19:40:54 2014] <Wohali>: well indeed
>> [Wed May  7 19:41:18 2014] <Kxepal>: Wohali: sure, I'll do. the decision
>> on this issue is almost depends on Fauxton-ready status. only deathbear
>> garren  and robertkowalski knows the answer on this question
>> [Wed May  7 19:41:18 2014] <Wohali>: there might be desires to "backport"
>> it to 1.x and I don't see any huge burdens there
>> [Wed May  7 19:41:22 2014] <Wohali>: but let's take it to dev@ please
>> [Wed May  7 19:41:31 2014] <wendall911>: +!
>> [Wed May  7 19:41:34 2014] <Wohali>: this isn't the right venue to decide
>> that
>> [Wed May  7 19:41:48 2014] <jan____>: yeah, but no harm, pre-introducing
>> it in a hypothetical 1.7.0 either. 2.0.0 will draw many news eyes and
>> Fauxton will be new to them :)
>> [Wed May  7 19:41:57 2014] <jan____>: either way, Futon has no place in a
>> 2.x world
>> [Wed May  7 19:42:03 2014] <Kxepal>: +1
>> [Wed May  7 19:42:06 2014] <Wohali>: +1
>> [Wed May  7 19:42:30 2014] <Wohali>: #info agreement that Fauxton replaces
>> Futon in CouchDB 2.0 (post merge) and that resolution of COUCHDB-1954 prior
>> to that is critical
>> [Wed May  7 19:42:44 2014] <Wohali>: #info Kxepal to start dev@ thread
>> about this and ensuring release preparation
>> [Wed May  7 19:42:46 2014] <Wohali>: good?
>> [Wed May  7 19:43:03 2014] <Kxepal>: yes (: thanks!
>> [Wed May  7 19:43:07 2014] <Wohali>: great moving on
>> ## Bylaws / CoC / Diversity ##
>> [Wed May  7 19:43:46 2014] <Wohali>: #info reminder to review nslater's
>> draft ASAP:
>> https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=40511017
>> [Wed May  7 19:43:58 2014] <Wohali>: #info comments should go to dev@ ML
>> [Wed May  7 19:44:28 2014] <Wohali>: #info Wohali to post initial CoC and
>> Diversity draft text this week
>> [Wed May  7 19:44:33 2014] <Wohali>: any comments?
>> [Wed May  7 19:44:38 2014] <jan____>: this is an open todo for me. I <3
>> the initiatives, just hadn’t had time to dig in yet
>> [Wed May  7 19:44:51 2014] <Wohali>: me as well, getting set up in my new
>> job is taking a lot out of me
>> [Wed May  7 19:44:54 2014] <Wohali>: nslater: anything more to add?
>> [Wed May  7 19:45:12 2014] <nslater>: checking
>> [Wed May  7 19:45:22 2014] <nslater>: nope
>> [Wed May  7 19:45:53 2014] <wendall911>: Progress on this looks excellent!
>> [Wed May  7 19:46:00 2014] <Wohali>: great
>> ## New Business ##
>> [Wed May  7 19:46:06 2014] <Wohali>: anything else?
>> [Wed May  7 19:46:34 2014] <Wohali>: i'll give it 120s while I drink decaf
>> [Wed May  7 19:46:46 2014] <Wohali>: banner week, really. so much exciting.
>> [Wed May  7 19:46:58 2014] <Wohali>: such progress. so merged. very happy.
>> wow.
>> [Wed May  7 19:47:28 2014] <wendall911>: Even though 1.6.0 was a little
>> delayed, it looks super sharp
>> [Wed May  7 19:47:29 2014] <jan____>: +100 :)
>> [Wed May  7 19:47:46 2014] <Wohali>: oh actually i have a topic
>> [Wed May  7 19:47:51 2014] <Wohali>: if no one else has anything
>> [Wed May  7 19:47:51 2014] <chewbranca>: yeah fun times seeing everything
>> coming together
>> [Wed May  7 19:47:59 2014] <Wohali>: it'll probably end up on dev@ but
>> [Wed May  7 19:48:06 2014] <jan____>: Wohali: go!
>> [Wed May  7 19:48:12 2014] <Wohali>: daleharvey wanted us to reconsider
>> issues w.r.t. getting couch onto distributions
>> [Wed May  7 19:48:24 2014] <Wohali>: can I get a) a clarification of our
>> positoin going forward and b) suggestions?
>> [Wed May  7 19:48:38 2014] <rnewson>: what's the problem?
>> [Wed May  7 19:48:40 2014] <Wohali>: his specific comment was around how
>> you can't get couch easily on ubuntu beyond 1.0.x
>> [Wed May  7 19:49:09 2014] <rnewson>: um.
>> [Wed May  7 19:49:15 2014] <Wohali>:  < daleharvey> I would like to think
>> so, but its not great for our users,
>> [Wed May  7 19:49:16 2014] <Wohali>:                there isnt a decent
>> way to install a recent couch on ubuntu
>> [Wed May  7 19:49:16 2014] <Wohali>:                last time I tried
>> [Wed May  7 19:49:16 2014] <Wohali>:  < daleharvey> its a popular distro,
>> its yalls choice as couch devs but
>> [Wed May  7 19:49:16 2014] <Wohali>:                imo its one where
>> couch should put effort into being usable
>> [Wed May  7 19:49:32 2014] <Wohali>: i believe our position is that we're
>> moving to tarball releases, correct?
>> [Wed May  7 19:49:43 2014] <rnewson>: raring is 1.2.0, saucy is 1.4.0,
>> trusty and beyond are 1.5.0
>> [Wed May  7 19:49:47 2014] <Wohali>: chewbranca also had some comments on
>> this
>> [Wed May  7 19:49:51 2014] <Wohali>: rnewson: LTS
>> [Wed May  7 19:49:58 2014] <rnewson>: if dale thinks we should be
>> backporting to old versions of ubuntu, then he's high.
>> [Wed May  7 19:49:59 2014] <Wohali>: which of course is mired in the past
>> [Wed May  7 19:50:01 2014] <jan____>: I think the golden solution is that
>> we will want to have a apt-chanel, that we maintain that works on latest
>> ubuntus
>> [Wed May  7 19:50:08 2014] <rnewson>: nooooo
>> [Wed May  7 19:50:08 2014] <wendall911>: yeah, LTS stuff anywhere isn't
>> pretty
>> [Wed May  7 19:50:18 2014] <rnewson>: we *just* agreed to *stop* building
>> installable tarballs entirely.
>> [Wed May  7 19:50:33 2014] <rnewson>: no init.d, no pidfile, we're
>> throwing that all away.
>> [Wed May  7 19:50:46 2014] <Wohali>: (no heart)
>> [Wed May  7 19:50:47 2014] <rnewson>: is someone now volunteering to make
>> new packages?
>> [Wed May  7 19:50:50 2014] <chewbranca>: depending on whatever version of
>> Erlang is installed on various distros is problematic
>> [Wed May  7 19:50:58 2014] <rnewson>: chewbranca: there's that too, yes.
>> [Wed May  7 19:51:11 2014] <rnewson>: but the right way to install couchdb
>> will be to build an erlang release and run it.
>> [Wed May  7 19:51:13 2014] <Wohali>: i'm promting chewbranca here because
>> he has a thread to post to dev@ about how, with 2.0, we may have to
>> restrict to very recent Erlang releases.
>> [Wed May  7 19:51:25 2014] <Wohali>: this brought up questions about
>> consumability
>> [Wed May  7 19:51:32 2014] <Wohali>: i didn't want to try and resolve this
>> here
>> [Wed May  7 19:51:37 2014] <Wohali>: i just wanted a pointer to our
>> current policy
>> [Wed May  7 19:51:40 2014] <rnewson>: it's important that couchdb includes
>> the version of erlang, it can't come from the platform, we need that for
>> upgrades.
>> [Wed May  7 19:51:47 2014] <Wohali>: and to prompt getting a discussion
>> going on dev@
>> [Wed May  7 19:51:53 2014] <wendall911>: I'll continue to maintain
>> couchdb-rpm, but it isn't really a very viable thing
>> [Wed May  7 19:51:56 2014] <rnewson>: why does dale need/want 1.0.x? We
>> don't even support that.
>> [Wed May  7 19:52:00 2014] <Wohali>: he doesn't
>> [Wed May  7 19:52:06 2014] <Wohali>: he just wants couchdb-current on an
>> LTS release
>> [Wed May  7 19:52:06 2014] <chewbranca>: yeah, I've been researching
>> different erlang releases and R15*, R16B, R16B01 are all bad and should not
>> be used
>> [Wed May  7 19:52:23 2014] <rnewson>: hm.
>> [Wed May  7 19:52:24 2014] <Wohali>: again, chewbranca's email will
>> possibly ignite a new heated discussion on this
>> [Wed May  7 19:52:32 2014] <Wohali>: i just want our current policy
>> somewhere so we can point to it
>> [Wed May  7 19:52:43 2014] <rnewson>: if someone wants to volunteer to
>> make an unofficial .deb for ubuntu LTS, that's fine.
>> [Wed May  7 19:52:48 2014] <Wohali>: presumably the last decision was made
>> on dev@, in JIRA, or possibly in person in Vienna?
>> [Wed May  7 19:52:49 2014] <rnewson>: Is there such a volunteer? (clue:
>> not it)
>> [Wed May  7 19:52:52 2014] <Wohali>: anyone?
>> [Wed May  7 19:52:52 2014] <jan____>: rnewson: yup, my thinking is, *once*
>> we have all that, it would be nice to have an apt-channel that makes it
>> easy for ubuntu folks to install
>> [Wed May  7 19:53:00 2014] <jan____>: best to dev@ this, I think, like
>> Wohali said :)
>> [Wed May  7 19:53:08 2014] <Wohali>: point of order
>> [Wed May  7 19:53:22 2014] <jan____>: rnewson: yeah, that’s my thinking,
>> I’m not saying it should be project policy, it would be a good situation to
>> be in
>> [Wed May  7 19:53:26 2014] <Wohali>: just looking for documentation as to
>> the agreement recently reached that rnewson mentioned.
>> [Wed May  7 19:53:29 2014] <jan____>: not worried about finding someone
>> [Wed May  7 19:53:31 2014] <Wohali>: then we take it to dev@
>> [Wed May  7 19:54:09 2014] <jan____>: I don’t know if we have written it
>> down anywhere :/
>> [Wed May  7 19:54:14 2014] <Wohali>: that's a problem
>> [Wed May  7 19:54:27 2014] <Wohali>: jan____: can I ask you to take an
>> action on this?
>> [Wed May  7 19:54:39 2014] <jan____>: can do
>> [Wed May  7 19:54:43 2014] <Wohali>: rnewson needs not another distraction
>> from the merge :)
>> [Wed May  7 19:54:44 2014] <Wohali>: tyvm
>> [Wed May  7 19:54:46 2014] <rnewson>: we might get some resources to make
>> the package thing happen.
>> [Wed May  7 19:55:01 2014] <wendall911>: I'd like to see rpm based
>> packaging get some attention
>> [Wed May  7 19:55:02 2014] <rnewson>: for today, it's enough to raise that
>> we are considering building them, if we have resources
>> [Wed May  7 19:55:10 2014] <wendall911>: always ubuntu only discussons,
>> and that's silly
>> [Wed May  7 19:55:15 2014] <Wohali>: #info jan____ to document and
>> publicize the current state-of-the-world on what the project release
>> artefacts are
>> [Wed May  7 19:55:18 2014] <Wohali>: fair?
>> [Wed May  7 19:55:24 2014] <rnewson>: I think I can safely say that we are
>> not good at building .deb or .rpm right now
>> [Wed May  7 19:55:32 2014] <rnewson>: so whoever is doing it, is someone
>> new
>> [Wed May  7 19:55:40 2014] <wendall911>: yep
>> [Wed May  7 19:55:46 2014] <rnewson>: .deb for me includes debian
>> [Wed May  7 19:55:53 2014] <jan____>: I think we’ve captured it!
>> [Wed May  7 19:55:53 2014] <rnewson>: I would never voluntarily use ubuntu
>> myself. :)
>> [Wed May  7 19:55:57 2014] <Wohali>: sorry to start a mess, i just wanted
>> a simple point of order :)
>> [Wed May  7 19:55:58 2014] <Wohali>: ok
>> [Wed May  7 19:56:03 2014] <Wohali>: any other business before i close the
>> meeting?
>> [Wed May  7 19:56:05 2014] <rnewson>: we're done on this topic
>> [Wed May  7 19:56:29 2014] <Wohali>: ah that should have been #action sorry
>> [Wed May  7 19:56:38 2014] <Wohali>: #action jan____ to document and
>> publicize the current state-of-the-world on what the project release
>> artefacts are
>> [Wed May  7 19:56:51 2014] <Wohali>: #action everyone to review nslater's
>> bylaws @
>> https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=40511017
>> [Wed May  7 19:57:17 2014] <Wohali>: ASFBot: meeting end
>> 
>> 
>> Meeting ended at Wed May  7 19:57:17 2014
>> 


Mime
View raw message