[12:01] <mpool> hullo
[12:02] <mpool> are we here?
[12:02] <spiv> I am.
[12:04] <lifeless> dudes and dudettes
[12:05] <mpool> ah
[12:05] <ddaa> Hello, sorry for being late to the meeting, we're chatting with sabdfl now
[12:06] <lifeless> are we ?
[12:07] <ddaa> everybody please get acquainted with https://launchpad.canonical.com/BazaarMeetingAgenda while i'm answering a call of nature
[12:09] <spiv> ddaa: is my proposed item on the agenda?
[12:09] <spiv> ddaa: you probably want to subscribe to that page :)
[12:09] <ddaa> subscribed
[12:10] <ddaa> sabdfl interrupt
[12:13] <SteveA> spiv: hi
[12:13] <SteveA> I'm interested in how the smartserver work is going
[12:14] <lifeless> its coming along nicely
[12:14] <lifeless> all transport tests are passing, and a number of the bzrdir interface conformance tests.
[12:14] <lifeless> we're refactoring/extending some infrastructure at the moment to accomodate it better
[12:15] <spiv> I haven't actually been involved in the smartserver work recently, I've been doing other things.  I should organise another day with mpool.
[12:15] <SteveA> I don't know the plan for it, so while this is very interesting, it doesn't give me a picture of how things are going.   What's the one sentence goal, and how much is left towards that goal?
[12:16] <lifeless> uhhh
[12:16] <lifeless> what metric are you hoping for the answer to be in ?
[12:16] <mpool> the goal is to push & pull over smartserver over ssh, and have it be much faster than sftp
[12:17] <mpool> i would say we are in the middle third of the work
[12:17] <mpool> we plan to release it in august
[12:17] <mpool> lifeless: do you concur, doctor?
[12:17] <lifeless> I think that fits
[12:19] <SteveA> when you say "release", does that mean release as open source kind of release, or release to internal testing for launchpad / supermirror ?
[12:19] <SteveA> does it require a new bzr release?
[12:19] <lifeless> it will be a new bzr release
[12:19] <lifeless> its not a separate product
[12:21] <mpool> yes, at least in the main public bzr.dev tree, hopefully in an actual released tarball
[12:21] <mpool> SteveA: i've spoken to sabdfl about this and heard no suggestion we needed private testing - do you know something i don't?
[12:22] <lifeless> SteveA: integration with launchpad/supermirror is a second-stage feature. Its been designed with spivs input to allow that to take place, but its not being implemented in parallel at this point. Though that may be an idea.
[12:24] <SteveA> okay
[12:24] <SteveA> I just talked with mark, and this is all fine
[12:25] <spiv> Implementing in parallel would probably help uncover unexpected issues.
[12:25] <SteveA> I'd like to get the smartserver used on the supermirror for the "launchpad 1.0" october release
[12:25] <SteveA> so, we should do some planning of that sometime
[12:26] <mpool> spiv: yes, i'd like to keep you involved 
[12:26] <lifeless> SteveA: did you have some different expectations ?
[12:26] <SteveA> I had unformed expectations
[12:26] <mpool> SteveA: seems like it would be nice if spiv can take responsibility for the lp side of that
[12:26] <mpool> as he is so far
[12:26] <SteveA> I had no idea how big/long/hard/uncut a task the supermirror would be
[12:26] <SteveA> s/supermirror/smartserver/
[12:27] <mpool> is there anything about say planning of it that he would need to know but doesn'tt?
[12:27] <SteveA> he being mark?
[12:27] <lifeless> http://samba.org/~jelmer/bzr/bzr-git-kernel.png
[12:27] <lifeless> ^^ awesome :)
[12:28] <lifeless> should we cancel this meeting ?
[12:28] <lifeless> I have another in 30 minutes
[12:28] <mpool> he in that sentence being spiv?
[12:28] <mpool> s/\?//
[12:28] <mpool> ok, let's keep moving on
[12:28] <mpool> i'll talk to stevea afterwards if that's ok
[12:28] <SteveA> ah, you're in a meeting
[12:28] <SteveA> I hadn't realize
[12:28] <SteveA> d
[12:29] <lifeless> ddaa is meant to be here
[12:29] <SteveA> anyway, thanks for the status update.
[12:29] <mpool> *acion* spiv, lifeless, mbp will meet this week
[12:29] <SteveA> mark is happy
[12:29] <SteveA> ddaa is sprinting with jamesh
[12:29] <lifeless> but I understand mark to be talking with him
[12:29] <SteveA> is there anything in particular you need from ddaa?
[12:29] <lifeless> hes the meeting chair :)
[12:29] <SteveA> lifeless: would you take on that role for this meeting?
[12:29] <ddaa> yeah, I'm coming
[12:29] <lifeless> sure
[12:29] <ddaa> okay I'm here
[12:29] <SteveA> thanks lifeless 
[12:29] <lifeless> meeting agenda coming up in 10
[12:29] <lifeless> (seconds)
[12:30] <lifeless> Usual introduction *      roll call *      production status
[12:30] <lifeless> Current focus
[12:30] <lifeless> *      SFTP advertising *      vcs-import knits *      cscvs/bzr-native *      supermirror branch browser *      private branches *      cscvs/svn-symlinks *      dyson *      About 9% of hosted branches are empty directories -- are people finding SFTP service too hard to use? (spiv)
[12:30] <lifeless> Usual end
[12:30] <lifeless> *      other meeting actions *      critical bugs *      pending sysadmin tasks *      any other business
[12:30] <lifeless> meh
[12:30] <lifeless> silly toolchain
[12:31] <ddaa> -EUSER
[12:31] <spiv> lifeless: hey, it's more compact that way ;)
[12:31] <ddaa> So, lifeless, spiv, mpool are here.
[12:31] <lifeless> Usual introduction *      roll call *      production status
[12:31] <lifeless> Current focus
[12:31] <lifeless> *      SFTP advertising *      vcs-import knits *      cscvs/bzr-native *      supermirror branch browser *      private branches *      cscvs/svn-symlinks *      dyson *      About 9% of hosted branches are empty directories -- are people finding SFTP service too hard to use? (spiv)
[12:31] <lifeless> Usual end
[12:31] <lifeless> *      other meeting actions *      critical bugs *      pending sysadmin tasks *      any other business
[12:31] <lifeless> nope, GARH
[12:31] <ddaa> Jamesh, ddaa, SteveA are mostly here and being periodically NMUed by sabdfl
[12:31] <lifeless> Usual introduction
[12:31] <lifeless>     *      roll call
[12:31] <lifeless>     *      production status
[12:31] <lifeless> Current focus
[12:31] <lifeless>     *      SFTP advertising
[12:31] <lifeless>     *      vcs-import knits
[12:31] <lifeless>     *      cscvs/bzr-native
[12:32] <lifeless>     *      supermirror branch browser
[12:32] <lifeless>     *      private branches
[12:32] <lifeless>     *      cscvs/svn-symlinks
[12:32] <lifeless>     *      dyson
[12:32] <lifeless>     *      About 9% of hosted branches are empty directories -- are people finding SFTP service too hard to use? (spiv)
[12:32] <lifeless> Usual end
[12:32] <lifeless>     *      other meeting actions
[12:32] <lifeless>     *      critical bugs
[12:32] <lifeless>     *      pending sysadmin tasks
[12:32] <lifeless>     *      any other business
[12:32] <lifeless> thats better
[12:32] <lifeless> ddaa: SteveA has asked me to chair to allow you to be NMU'd
[12:32] <lifeless> production status : anyone have issues to report?
[12:32] <lifeless> 5
[12:32] <lifeless> 4
[12:32] <lifeless> 3
[12:32] <lifeless> 2
[12:32] <ddaa> problem with samba branches
[12:32] <lifeless> 1
[12:33] <ddaa> apparently jelmer's plugin is producing revisions that get through the branch puller
[12:33] <lifeless> with the scanner ? there is a bug report I believe.
[12:33] <ddaa> but cause the branch scanner to crash
[12:33] <lifeless> ok. who has the backtrace? can they enlarge on the bug report ?
[12:33] <lifeless> ddaa: is that you ?
[12:33] <lifeless> or jamesh or spiv ?
[12:33] <spiv> Anyone that gets the launchpad-errors-reports email.
[12:33] <ddaa> one of the responsibilities of the branch puller is to sanitize the data so that would not happen
[12:33] <lifeless> ok, spiv can you do that ?
[12:33] <spiv> I've replied to the bug already with the exception (though not the full traceback)
[12:34] <lifeless> ok, any other production errors ?
[12:34] <spiv> It died in get_revision, claiming invalid XML iirc.
[12:34] <lifeless> ddaa: this is a decision meeting not a design meeting
[12:34] <lifeless> ddaa: please discuss the detail in the bug report
[12:34] <ddaa> I'm not having a design discussion.
[12:34] <ddaa> no other issue
[12:34] <lifeless> well you're assuming the data is invalid, it may be skew between systems for instance.
[12:34] <spiv> (bug 53825, for the record)
[12:34] <ddaa> no other important issues that I'm aware of, at least
[12:35] <lifeless> spiv: can you do this examination please ?
[12:35] <spiv> lifeless: I've already followed up on the bug, do you want me to dissect it further?
[12:35] <lifeless> I will eyeball it and comment
[12:35] <lifeless> ok, sftp advertising
[12:35] <lifeless> this was in ddaa's queue IIRC
[12:35] <lifeless> ddaa: status  ?
[12:36] <ddaa> actions from last meeting:
[12:36] <ddaa>    * ddaa to tell jdub about existing blog entry
[12:36] <ddaa>    * jamesh to blog about team-shared branches and using them with checkouts 
[12:36] <lifeless> actually, jamesh was to blog on it
[12:36] <lifeless> the agenda is not detailed enough.
[12:36] <jamesh> I haven't written the article yet
[12:36] <ddaa> I did the first one, did not observe any resulting effect
[12:36] <ddaa> jamesh: still planning to do it, or want to hand it over?
[12:37] <jamesh> Still planning to do it
[12:37] <lifeless> ok
[12:37] <lifeless> vcsimports knits
[12:37] <ddaa> action from last meeting
[12:37] <ddaa> * ddaa to file bug about converting existing vcs-import branches to knits. 
[12:37] <ddaa> mh
[12:37] <ddaa> do not remember clearly
[12:38] <ddaa> looking
[12:38] <ddaa> https://launchpad.net/products/launchpad-bazaar/+spec/vcs-imports-knits-upgrade
[12:38] <lifeless> ok
[12:38] <lifeless> who is moving this forward - you ?
[12:39] <ddaa> I would like feedback on that spec
[12:39] <lifeless> ok, jamesh - could you review it while you are at the sprint?
[12:39] <jamesh> okay
[12:39] <lifeless> seems like a good use of the face time
[12:39] <lifeless> ok
[12:39] <lifeless> bzr-native
[12:39] <lifeless> ddaa: that seems to be coming along well. 
[12:40] <ddaa> feedback has bee requested from lifeless, jamesh and steve
[12:40] <jamesh> we've been discussing the importd changes a bit this morning
[12:40] <lifeless> on bzr-native ?
[12:40] <ddaa> feedback on vcs-imports-knits-upgrade has been requested
[12:40] <ddaa> we discussed bzr-native this morning
[12:40] <ddaa> got to order food
[12:41] <lifeless> ok, in progress. cool
[12:41] <lifeless> branch browser
[12:41] <lifeless> spiv: ?
[12:42] <ddaa> action from last meeting:
[12:42] <ddaa> * spiv to have a pre-impl call with SteveA
[12:42] <spiv> No progress worth reporting.  It's my top priority for this week.
[12:42] <spiv> Oh, I had that call.
[12:42] <lifeless> ddaa: please dont blat the actions list, we've all read the page
[12:42] <ddaa> nice, make it rock
[12:42] <spiv> And the mail sent to the list by Steve outlines the plan.
[12:43] <lifeless> spiv: do you have an expected time of completion ?
[12:43] <lifeless> spiv: like 'august' ?
[12:43] <lifeless> spiv: ?
[12:43] <spiv> lifeless: That seems a reasonable estimate.  Ask me again next meeting for a firmer answer.
[12:43] <lifeless> ok. for the next meeting, have an answer :)
[12:44] <lifeless> private branches:
[12:44] <ddaa> ACTION: spiv to give delivery estimate for branch browser
[12:44] <lifeless> still in spec form. I have feedback to give. Steve has updated the spec.
[12:44] <lifeless> to be clearer - a terminology change
[12:44] <lifeless> svn-symlinks.
[12:44] <lifeless> ddaa: is that rolled out ?
[12:44] <ddaa> lifeless: jamesh has rolled out
[12:44] <ddaa> hu
[12:44] <ddaa> has committed
[12:44] <ddaa> I rolled out
[12:44] <ddaa> It rocks
[12:44] <lifeless> cool
[12:44] <lifeless> dyson:
[12:45] <lifeless> jamesh has got it working from behind the proxy
[12:45] <jamesh> stub reported a new bug
[12:45] <lifeless> stub has yet to trial it again on staging. expecting that soon
[12:45] <lifeless> ahha!
[12:45] <jamesh> https://launchpad.net/products/launchpad/+bug/53698
[12:45] <lifeless> I did not notice that
[12:45] <jamesh> it failed when scanning a GRASS cvs snapshot tarball
[12:45] <jamesh> because the version number component of the tarball did not match the DB constrainyt
[12:46] <lifeless> 6.1.cvs_src_snapshot_2006_07_15  being the string
[12:46] <ddaa> "dyson blows on grass"
[12:46] <jamesh> I guess valid_version() is looking for valid debian version numbers
[12:46] <jamesh> which wouldn't like the underscores
[12:46] <lifeless> jamesh: can you please talk this over with Mark ? 
[12:46] <jamesh> okay
[12:46] <lifeless> jamesh: its in the product registry area which I know he cares about
[12:47] <lifeless> I think representing what upstream does in ProductSeriesRelease is important, my vote is to loosen the constraint for this table.
[12:47] <lifeless> hosted branches:
[12:47] <lifeless> About 9% of hosted branches are empty directories -- are people finding SFTP service too hard to use? (spiv)
[12:47] <lifeless> spiv: ?
[12:48] <lifeless> ddaa: can you prep a critical bugs list please.
[12:48] <ddaa> One of the problem with the sftp service is that once a branch has been created, if you did not create a proper branch the first time, for some reason, then you are screwed
[12:48] <lifeless> ddaa: as i dont have one ready
[12:48] <mpool> right
[12:48] <spiv> I got stub to add some stats about branches to cricket.
[12:48] <mpool> can i suggest we discuss this (empty branches, lp use, etc) on the lp list please?
[12:48] <lifeless> yes.
[12:49] <lifeless> I think 10% is uncomfortable, and discussion is needed.
[12:49] <spiv> Sure, I can move this to the list.
[12:49] <lifeless> any other meeting items ?
[12:49] <lifeless> 5
[12:49] <lifeless> 4
[12:49] <lifeless> 3
[12:49] <lifeless> 2
[12:49] <lifeless> 1
[12:49] <lifeless> ok
[12:49] <lifeless> critical bugs:
[12:49] <lifeless> ddaa: do you have a list handy ?
[12:49] <ddaa> doing
[12:49] <lifeless> we will come back
[12:50] <lifeless> pending sysadmin tasks?
[12:50] <ddaa> critical bugs:
[12:50] <ddaa> 31308: cannot set branch associated to productseries: lifeless to spec
[12:50] <ddaa> 37897: renaming breaks vcs-imports: david to do something about it
[12:50] <ddaa> basically, no progress on that
[12:50] <lifeless> ok, no change on those.
[12:50] <lifeless> pending sysadmin tasks.. anyone got stuff to nag about ?
[12:50] <ddaa> and 51130
[12:50] <lifeless> 5
[12:50] <lifeless> 4
[12:50] <lifeless> 3
[12:50] <lifeless> 2
[12:50] <lifeless> 1
[12:50] <lifeless> ok
[12:51] <lifeless> any other business? or meeting over time
[12:51] <lifeless> 5
[12:51] <lifeless> 4
[12:51] <lifeless> 3
[12:51] <lifeless> 2
[12:51] <lifeless> 1
[12:51] <lifeless> meeting over, thanks
[12:51] <lifeless> see some of you in 9 minutes for the review meeting
[12:51] <lifeless> mpool: can you do me a favour  ?
[12:52] <lifeless> mpool: collate the irc log into a minutes for the meeting?
[12:52] <lifeless> I need to prep my meeting
[12:52] <mpool> sure
[12:52] <lifeless> thanks
[12:52] <ddaa> s/from/for/
[01:41] <jelmer> sorry, hope I'm not too late
[01:43] <ddaa> sorry was meditating about bugs, branches and revisions
[01:43] <jelmer> (-:
[01:43] <ddaa> jelmer: sabdfl wants us to reconcile your work and importd
[01:44] <ddaa> he said, essentially "I sponsored jelmer work for the ability to commit to subversion, but I want it to work with our existing import system"
[01:45] <jelmer> ddaa: Reconcile how exactly?
[01:45] <ddaa> so that importd can do something useful with the metadata bzr commit puts into svn
[01:45] <jelmer> making existing imports compatible with bzr-svn or simply integrating bzr-svn into importd?
[01:45] <jelmer> Ah, right
[01:46] <jelmer> http://samba.org/~jelmer/bzr/mapping.txt contains a short write-up of the metadata used by bzr-svn
[01:46] <ddaa> so, id-wise, importd and your plugin have two different approaches: random and deterministic
[01:47] <ddaa> there are advantages to both
[01:47] <jelmer> yes, do those have to be reconciled as well?
[01:47] <ddaa> I do not think it is reasonably feasible
[01:48] <ddaa> I think what would make sabdfl happy would be the possibility to use your work to:
[01:48] <jelmer> I don't have a lot of time at the moment, have to go in 5 minutes - any chance we can arrange a meeting later this week?
[01:48] <ddaa> get a bzr branch produced by importd, makes some new changes there, commit to svn, and have importd get the information from there
[01:49] <ddaa> so I'm happy to have two branch universes, the one based on cscvs imports, and the one based on your import system
[01:49] <ddaa> but both should be able to roundtrip with svn
[01:49] <ddaa> jelmer: please think about it
[01:49] <ddaa> jelmer: when would a meeting suit you?
[01:50] <jelmer> ddaa: Sure, I will. There's a couple of issues we'll ahve to deal with
[01:50] <ddaa> let's arrange meeting first
[01:50] <ddaa> sabdfl focus is no importd data, your deterministic import stuff is nice, but it's not the essential stuff as far as he is concerned
[01:51] <ddaa> "is _on_ importd data"
[01:51] <jelmer> ddaa: Thursday and friday, preferably
[01:51] <ddaa> ASAP, we are in a sprint with SteveA, jamesh and sabdfl in voice range
[01:51] <jelmer> I'll be away on tuesday and wednesday
[01:51] <ddaa> Thursday, what time?
[01:51] <jelmer> oh, ok - please say hi to them :-)
[01:51] <ddaa> not between 1200 and 1300 UTC, that's all
[01:52] <jelmer> After 1300 UTC.. I'm arriving at the airport early morning, not sure when I'll be home exactly.
[01:52] <ddaa> (that's the time of the weekly launchpad mass^Wmeeting)
[01:52] <jelmer> (-:
[01:53] <ddaa> I'd like a firm hour so we can schedule effectively
[01:53] <ddaa> say, 15 UTC?
[01:53] <jelmer> ddaa: Feel free to pick one, it's all fine by me. 
[01:53] <jelmer> Yeah, 15 UTC sounds fine. 
[01:53] <ddaa> deal
[01:53] <jelmer> Great. 
[01:53] <jelmer> I've got to run. See you on thursday!
[01:54] <ddaa> see you
[04:07] <SteveA> mpool: ping?