22:01:59 <tcohen> #startmeeting Development IRC meeting 15 April 2015 - part 2
22:01:59 <huginn`> Meeting started Wed Apr 15 22:01:59 2015 UTC.  The chair is tcohen. Information about MeetBot at http://wiki.debian.org/MeetBot.
22:01:59 <huginn`> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
22:01:59 <huginn`> The meeting name has been set to 'development_irc_meeting_15_april_2015___part_2'
22:02:13 <tcohen> #topic Introductions
22:02:14 <wahanui> #info wahanui, a bot that has become sentient
22:02:38 <tcohen> please introduce yourselves using #info name
22:02:45 <tcohen> just like wahanui did
22:03:00 <tcohen> #info Tomas Cohen Arazi, Universidad Nacional de Córdoba
22:03:27 <tcohen> eythian, wizzyrea, rangi ?
22:03:43 <wizzyrea> #info Liz Rea, Catalyst IT
22:03:58 <wizzyrea> no rangi (away) or eythian (yet)
22:04:06 <tcohen> we'll wait then :-D
22:04:21 <tcohen> unless we can count as enough
22:04:25 <wizzyrea> hah
22:04:34 <wizzyrea> think it might be just you and me :)
22:05:00 <tcohen> ok
22:05:15 <tcohen> #topic RM 3.20 comments
22:05:21 <tcohen> have you read the logs?
22:05:35 <tcohen> i basically apologize (again)
22:06:18 <wizzyrea> I hadn't - I will go do now
22:06:21 <tcohen> and i've been (and will keep) using my office hours for RM duties
22:06:32 <tcohen> focused on bugfixes since last week
22:06:41 <tcohen> and enhancements follow
22:06:57 <wizzyrea> yeah, I think we can all understand and be sympathetic to your situation there
22:07:20 <tcohen> thanks
22:07:37 <tcohen> #info wizzyrea is sympathetic to tcohen
22:07:43 <tcohen> LOL
22:07:44 <wizzyrea> lol
22:08:01 <tcohen> #info: Freeze for 3.20: What's not passed QA by April 23nd, 9:10 Argentinian time... won't make it in
22:08:30 <tcohen> i set april 22nd as the cut for things to have the chance to get included
22:08:36 <wizzyrea> sensible
22:09:19 <wizzyrea> fwiw I have no opinion on moose, the others here might
22:09:21 <tcohen> as i will get to the office to get a screenshot of the pqa list at 9 am (ART) we shift to the next day
22:09:50 <tcohen> #info please let tcohen know if your feature needs attention to make it move, or have his opinion on the odds of it making into 3.20
22:10:21 <tcohen> #topic Bugs
22:10:56 <tcohen> on part 1 I proposed to approach fredericd  about moving to Moo
22:11:22 <tcohen> khall asked people to give 13799 a chance to make it into 3.20
22:12:17 <tcohen> during part 1 we also talked about QueryParser
22:12:28 <tcohen> which has issues that need to be addresed
22:13:01 <tcohen> the question was wheter to put bugs that required not using QP on hold or not
22:13:22 <tcohen> and we agreed QueryParser having a bug shouldn't prevent us from having the feature
22:13:45 <wizzyrea> simply for my information
22:14:04 <tcohen> shoot
22:14:05 <wizzyrea> is it more desirable to have QueryParser on
22:14:10 <wizzyrea> generally?
22:14:44 <wizzyrea> because it could be a bit like the old "put it in bootstrap and backport it to prog" that we did for a while
22:15:27 <wizzyrea> personally, I feel like we should work on the more advanced framework
22:15:40 <eythian> #info Robin Sheat, Catalyst IT
22:15:43 <wizzyrea> and port features back
22:15:44 <tcohen> QueryParser is a really good idea that we should be using more widely
22:15:46 <wizzyrea> but that's just me
22:15:49 <tcohen> but
22:16:02 <tcohen> people are afraid of getting into it to fix stuff
22:16:18 <wizzyrea> well then that's the problem that needs addressing, imo
22:16:38 <tcohen> Jonathan basically found a bug on the PQF driver Koha uses
22:16:45 <tcohen> and we need to address that, correctly
22:16:51 <tcohen> you are right
22:17:02 <tcohen> i'll try to
22:17:06 <wizzyrea> because people won't change to the new hotness if we let them get by wearing the old sweatpants, if you know what I mean
22:17:14 <tcohen> Joubu bets we won't be able to fix it on time
22:17:27 <tcohen> wizzyrea: I agree
22:17:55 <tcohen> much people are conservative about how they use their daily tools
22:18:10 <tcohen> anyway
22:18:24 <wizzyrea> that may be true, that we won't be able to fix it on time
22:18:36 <wizzyrea> but we mandated "features in bootstrap, back to prod" before, we could do it here too
22:18:39 <wizzyrea> just something to think about
22:18:52 <tcohen> we agreed we should push the feature, and fill a bug for QueryParser
22:19:07 <tcohen> QueryParser is off by default, so it is expected that too little people are using it
22:19:16 <tcohen> and that's because there are some outstanding bugs still
22:19:18 <wizzyrea> yeah, I think this instance is probably fine
22:19:43 <tcohen> does HEA have information about that?
22:19:48 <tcohen> it could be handy this time :-D
22:19:49 <wizzyrea> if it's a syspref, probably
22:20:18 <thd> #info Thomas Dukleth, Agogme, New York City
22:20:23 <wizzyrea> http://hea.koha-community.org/systempreferences > find in page QueryParser
22:20:32 <wizzyrea> 59 with it enabled?
22:20:34 <tcohen> 59 out of 63 have it disabled
22:20:39 <tcohen> 0
22:20:49 <wizzyrea> other way round :)
22:20:50 <tcohen> and the rest: undef :-F
22:20:59 <wizzyrea> from what I said
22:21:08 <wizzyrea> that is, you're right
22:21:16 <wizzyrea> so yeah, nobody who's submitted using that.
22:21:18 <wizzyrea> :/
22:21:35 <tcohen> so we are safe probably
22:21:38 <wizzyrea> yep
22:21:39 <eythian> My understanding is that QP had some bugs that would make it more or less not work if you turned it on.
22:21:50 * eythian has been thinking about looking into it for ES work.
22:21:57 <tcohen> that's almost true
22:21:58 <wizzyrea> is it our intention to use it?
22:22:13 <tcohen> things like "right truncation doesn't work as expected"
22:22:22 <eythian> I have no intention, I just wanted to get a good feel for it to see how useful it is/would be.
22:22:25 <tcohen> not things like "searches don't work"
22:22:37 <eythian> ah, I thought it was closer to "not work"
22:22:40 <thd> tcohen: Are certain that a similar bug does not affect QueryParser in Evergreen?
22:22:59 <tcohen> i couldn't approach gmcharlt_
22:23:10 <tcohen> but jcamins told me Evergreen has been using it for years
22:23:15 <tcohen> without any problems
22:23:22 <tcohen> so the issues are Koha specific
22:23:31 <tcohen> and related to our PQF driver
22:23:39 <tcohen> which narrows things a lot
22:23:48 <tcohen> to put it clear (my opinion on QP)
22:24:05 <eythian> ok
22:24:14 <tcohen> our search code is full of ad-hoc stuff noone knows what to expect from it
22:24:24 <tcohen> imean, you read a line and you know what it does
22:24:39 <tcohen> but having the whole picture of all the border cases... difficult
22:24:40 <eythian> yes, it is.
22:24:46 <tcohen> the idea about queryparser
22:25:05 <tcohen> is that it is an abstract query parser and translation engine
22:25:10 <tcohen> configurable using a yaml file
22:25:17 <thd> tcohen: Could the problems have been present but unnoticed by Evergreen users?
22:25:35 <eythian> which I though would be good to generate ES queries, eventually.
22:25:53 <thd> ... or without complaint from Evergreen users even if noticed?
22:25:58 <tcohen> the problem is that parse trees traversal and what to do with it is not trivial as a regular expression
22:26:11 <tcohen> eythian: yes, a ES driver should do the job
22:26:36 <tcohen> thd: the problem jonathan mentioned today is trivial enough to make it clear they dont face it
22:27:40 <thd> I see problems everywhere I look but there are too many for me to have time to file bug reports about all of them ;)
22:28:09 <tcohen> thd: are you using QP?
22:28:18 <thd> tcohen: No.
22:28:30 <tcohen> heh
22:28:45 <tcohen> this is jonathan's post to koha-devel
22:28:52 <tcohen> explaining the bug http://koha.1045719.n5.nabble.com/QueryParser-does-not-manage-more-than-1-operator-td5823170.html
22:28:53 <wizzyrea> my only point is that if we have btter tech, we should be looking to use it
22:29:02 <tcohen> wizzyrea: we should
22:29:04 <wizzyrea> and improve it
22:29:16 <wizzyrea> and not languish because it's more comfortable.
22:29:31 <wizzyrea> that's all I'm getting at.
22:29:34 <tcohen> that's the right attitude
22:29:36 <thd> wizzyrea++
22:29:41 <tcohen> wizzyrea++
22:30:19 <wizzyrea> unfortunately enforcing that comes from the top :/
22:30:27 <wizzyrea> and that's an unpleasant thing.
22:30:50 <tcohen> #info wizzyrea reminded us that the project evolves because of our commitment to do good stuff, not to let ourselves languish in favor of comfort
22:31:21 <tcohen> wizzyrea: but... we don't have a top!
22:32:04 <thd> In the early meeting, ColinC reported that he thought that QueryParser would break multiple system preferences as it is.
22:32:38 <wizzyrea> I am certain that is true, it could be a goal of a future release to tidy all that up perhaps?
22:33:19 <tcohen> there aren't *that many* open bugs for QueryParser
22:33:28 <thd> The fault may be in the legacy query code which many people have had difficulty deciphering.
22:33:29 <tcohen> but we definitely need someone to work on that
22:33:32 <wizzyrea> or (sorry eythian) to make sure that the next iteration of searching in Koha doesn't have the same issues
22:33:33 <tcohen> bug 12742
22:33:34 <huginn`> 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=12742 enhancement, P5 - low, ---, gmcharlt, NEW , Omnibus: QueryParser
22:33:53 <wizzyrea> or rather, to learn from our mistakes
22:34:12 <thd> tcohen:  ColinC was reporting his suspicion for things which may not have been properly discovered and reported.
22:34:15 <eythian> searching for ES will be pretty easy to replace with something better.
22:35:03 <tcohen> eythian: that's awesome to hear
22:35:06 <tcohen> *read
22:35:47 <tcohen> ok, we need to move on, unless eythian, wizzyrea, thd or rangi (if absent... beware) volunteers to work on QP
22:35:50 <eythian> at the moment, it's mostly compatibility layers with the C4::Search way anyway.
22:36:41 <tcohen> eythian: on my work for a more reasonable way to filter data, i came with a Koha::Search::Result class (or something like that)
22:36:59 <eythian> "filter data"?
22:37:07 <tcohen> that only puts some stuff together (bitstream, metadata format, etc)
22:37:32 <tcohen> I think we should make broader use of Koha::RecordProcessor
22:37:39 <tcohen> with Koha::Filter::*
22:38:15 <tcohen> for example, if you need to apply a policy about fields/subfields visibility
22:38:22 <tcohen> then have a filter that does that
22:38:23 <eythian> ah yep
22:38:52 <tcohen> prepare data for display in results context, have a specific filter that applies the xslt, etc
22:39:20 <eythian> yeah, build a filter chain
22:39:25 <tcohen> and i wrote that tiny class that encapsulated both the bitstream, and the metadata format of the bitstream, thinking of having JSON data, etc...
22:39:34 <tcohen> eythian: you're damn right
22:40:01 <tcohen> ok, diverted a lot
22:40:32 <tcohen> moving on
22:40:49 <wizzyrea> <sorry>
22:40:52 <wizzyrea> <not sorry>
22:41:08 <tcohen> i was worried about the patches waiting for QA right now that introduce new fields on the XSLT
22:41:39 <tcohen> and we concluded that adding fields there is not a problem as long as they are not added to the frameworks and/or enabled for visibility
22:42:12 <wizzyrea> as long as the visibility rules are applying correctly, this is sensible to me.
22:42:12 <eythian> yeah
22:42:24 <wizzyrea> (as far as I know they are)
22:42:43 <tcohen> wizzyrea: they aren't, but shhh
22:42:52 <wizzyrea> hrmpn.
22:43:08 <wizzyrea> hrmph.
22:43:19 <tcohen> :-P
22:43:26 <cait_not_here> hmpf
22:43:41 <wizzyrea> ohh you got a double hmpf.
22:43:44 <tcohen> they are, until someone has a regression test for that
22:43:54 <tcohen> :-P
22:44:23 <tcohen> who's got the double?
22:44:25 <tcohen> me?
22:44:26 <wahanui> hmmm... tcohen is 122
22:45:01 <tcohen> wahanui: tcohen is actually 123
22:45:02 <wahanui> ...but tcohen is obsessed with packages' scripts|122|obsessed with automated testing :)|king of the lab...
22:45:13 <tcohen> ok
22:45:30 <wizzyrea> you obsess a lot.
22:45:31 <wizzyrea> :)
22:45:59 <tcohen> i think we agree that if view policy is working, there's no problem adding fields to the xSLTs, right?
22:46:05 <wizzyrea> yep
22:46:08 <tcohen> great
22:46:25 <tcohen> on part one we mentioned bug 10480
22:46:27 <huginn`> 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=10480 enhancement, P5 - low, ---, m.de.rooy, Signed Off , Improvements for framework plugins in cataloguing/item editor
22:46:45 <tcohen> that does an interesting job refactoring the cataloguing plugins code
22:46:52 <tcohen> adding wings to them
22:47:15 <tcohen> i mentioned because (1) it looks great and (2) it had been signed by two QA members
22:47:25 <tcohen> so the SO status was strange
22:47:38 <tcohen> but Jonathan said he wanted more eyes on it
22:47:44 <tcohen> and khall volunteered
22:47:47 <tcohen> khall++
22:48:35 <wizzyrea> khall++
22:49:06 <tcohen> #topic Big stuff we are working on
22:49:21 <tcohen> we talked a bit about the REST api implementation
22:49:36 <tcohen> because we needed to catch up
22:49:43 <eythian> I had a look at that yesterday, the dependencies are awkward and annoying
22:49:58 <tcohen> jonathan mentioned the work on bug 13799 needs signoff
22:49:59 <huginn`> 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=13799 enhancement, P5 - low, ---, julian.maurice, Needs Signoff , Add base for building RESTful API
22:50:19 <eythian> it's not near a state for signoff, imo
22:50:28 <tcohen> and khall said he approached julian about some called "Robin"'s comments
22:51:19 <tcohen> eythian: you mean they are not in jessie?
22:51:22 <wizzyrea> he's a taskmaster, that packaging manager of ours.
22:51:26 <eythian> they're not in anything
22:51:33 <tcohen> :/
22:51:51 <eythian> they're not even in any unreleased version
22:52:20 <wizzyrea> ^ this seems like perhaps not the best situation ever.
22:52:27 <tcohen> mojolicious and swagger2
22:52:39 <eythian> older mojolicious is in debian, swagger2 isn't at all.
22:53:12 <eythian> the next biggest issue is that the code has no documentation.
22:53:33 * tcohen hasn't had the time to take a look yet
22:53:36 <bag> http://wiki.koha-community.org/wiki/New_REST_API_RFC
22:53:40 <eythian> but anyway, it's documented on the bug :)
22:53:57 <eythian> (rather, my issues are documented on the bug)
22:54:04 <tcohen> but we agreed we don't need the latest mojolicious
22:54:23 <eythian> tcohen: swagger2 declares that it depends on mojolicious 6.0
22:56:10 * tcohen is stuck
22:56:13 <thd> eythian: Which code has no documentation?
22:56:37 <tcohen> i couldn't find that dependency, but i will sure take a look at the office tomorrow
22:56:52 <eythian> http://bugs.koha-community.org/bugzilla3/attachment.cgi?id=36705&action=diff <-- see for example the perl files here
22:57:34 <tcohen> eythian: ok
22:57:45 <tcohen> i'm worried, but we'll sort it out, for sure
22:57:50 <eythian> yeah
22:58:24 <eythian> I'm having a quick go at building mojolicious 6, it might be that it's easy, but if it can't backport to wheezy, then that's simply not an option.
22:59:31 <tcohen> ok, lets keep positive
22:59:42 <wizzyrea> butterflies!
22:59:42 <tcohen> khall nudged julian about those concerns
22:59:57 <thd> eythian: Is jessie not forthcoming at the end of the month?
23:00:00 <tcohen> we'll talk about that before the end of the week for sure
23:00:01 <eythian> thd: it is
23:01:42 <tcohen> jessie == systemd == unix is dead
23:01:47 <tcohen> </rant>
23:02:19 <tcohen> ok
23:02:22 <tcohen> anything else?
23:02:29 <tcohen> eythian: about ES'
23:02:29 <eythian> squeeze is being shot in the head
23:02:38 <eythian> come 3.20
23:02:39 * thd in no hurry to discover jessie bugs.
23:02:41 <tcohen> ah *that*
23:02:52 * eythian decides unilaterally
23:03:05 <tcohen> on the previous dev meeting we agreed that 3.20 wouldn't support Squeeze
23:03:14 <wizzyrea> \o/ that was easy.
23:03:27 <eythian> you did?
23:03:28 <thd> eythian: Is squeeze not being maintained until 2016 even if not for Koha?
23:03:32 * eythian can't find it in the logs
23:03:38 <eythian> thd: it is
23:03:55 <eythian> thd: but if you want to backport everything that newer koha wants, be my guest :)
23:04:06 * tcohen can't find them ,but is pretty sure
23:04:58 * thd just wants to avoid upgrading some squeeze systems to wheezy too suddenly
23:05:05 <tcohen> is it valid if the RM dreamed about a vote on something?
23:05:38 <bag> yes
23:05:52 <eythian> thd: wheezy was released nearly two years ago, you may need to refine your definition of sudden :)
23:05:55 <tcohen> heh, thanks bag
23:06:17 <wizzyrea> tcohen: maybe it was in the main meeting
23:06:35 <bag> I seem to remember something about it too
23:06:58 <wizzyrea> or after you closed the meeting. I don't see it in the logs but whatever.
23:07:01 <thd> eythian: My definition is not being forced to do by the end of the month something which I should have done two years ago :)
23:07:18 <eythian> thd: well, you'll only have to do it in order to upgrade to 3.20 anyway
23:07:39 <eythian> support in 3.18 etc. will remain
23:08:06 <thd> eythian: I mean my non-Koha backup systems, etc.
23:08:57 <tcohen> eythian: i'll send a new email about this to koha-devel, I think no one would disagree
23:08:58 <eythian> ah right. If it's any consolation I have an 8.04 box still running.
23:09:14 * tcohen too
23:09:29 <eythian> tcohen: I was going to announce it on both lists to make sure everyone knows about it.
23:09:30 <tcohen> ah, no 10.04
23:10:34 <tcohen> go ahead
23:10:39 <eythian> cool.
23:10:52 <tcohen> you could avoid the "unilateral" speech just in case
23:11:02 <tcohen> something more politically correct maybe?
23:11:05 <tcohen> :-D
23:11:55 <eythian> oh sure
23:12:04 <eythian> that's just for you people :)
23:12:19 <wizzyrea> "Hi, it's not possible or even reasonable to continue supporting squeeze, so we're not going to do it anymore. eom."
23:12:53 <tcohen> the release notes will mention it too
23:13:06 <tcohen> eythian: go ahead
23:13:12 <eythian> cool
23:13:26 <tcohen> we will blame you if something goes wrong about the announcement anywayu
23:14:05 <tcohen> ok
23:14:10 <tcohen> moving on then
23:14:16 <tcohen> as I said on part 1
23:14:29 <tcohen> cut date -> next wednesday
23:14:38 <eythian> wfm :)
23:14:41 <tcohen> beta -> next weekend (after the cut)
23:15:10 <tcohen> next dev meeeting probably wednesday week after freeze - april 29
23:15:15 <tcohen> objections?
23:15:34 <eythian> tcohen: I'd really like to get you to have an eyeball over bug 13979 and put it in, btw.
23:15:35 <huginn`> 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=13979 major, P1 - high, ---, robin, Signed Off , Packages uninstallable on Debian Jessie
23:15:37 <tcohen> that gives us time to coordinate efforts toward stabilization if needed
23:15:56 <thd> tcohen: What happens during Wed. before Wed. has ended?
23:15:57 * tcohen adds it to his pocket
23:16:24 <tcohen> thd: i don't follow you sorry
23:17:02 <thd> tcohen: If something passes QA on Wednesday would it have a chance of qualifying?
23:17:09 <tcohen> of course
23:17:34 <tcohen> thursday 9:10 am argentinian timezone is the actual cut
23:18:07 <tcohen> [off] you could try to brive me anyway
23:18:11 <tcohen> :-P
23:18:24 <thd> Wow, many hours of UTC time for those who can stay awake doing QA :)
23:18:55 <tcohen> yeah
23:19:10 <tcohen> #info next dev meeeting probably wednesday week after freeze - april 29
23:19:54 <tcohen> any other thing?
23:20:01 <tcohen> to talk about?
23:20:14 <eythian> ES is moving along. About to start on staff client
23:20:24 <eythian> otherwsie, nothing to report from me
23:20:25 <thd> tcohen: Did we skip some items?
23:20:26 <tcohen> eythian: awesome!
23:20:26 <wahanui> That'll be $1 for the awesome jar, tcohen
23:20:58 <bag> yes good news
23:20:59 <wahanui> good news is it looks like it's running properly.
23:20:59 <tcohen> thd: i think we skipped the country of production positions talk
23:21:26 <wizzyrea> production positions?
23:21:49 <tcohen> the item was
23:21:50 <tcohen> Language and country fixed size field positions: MARC flavour-specific vs. a general approach (Bug 13912)
23:21:51 <huginn`> 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=13912 enhancement, P5 - low, ---, gmcharlt, Failed QA , System preference for default place of publication (country code) for field 008, range 15-17
23:21:52 <thd> tcohen: That is fine, I had an idea about how to keep it with the frameworks without putting it in the frameworks.
23:22:10 <tcohen> *ublication
23:22:23 <tcohen> is it failed?
23:23:03 <thd> tcohen: It passed but I toggled the state in ignorance of the correct procedure when adding my comments this morning.
23:23:11 <tcohen> wizzyrea: my question was wheter that kind of sysprefs could be set framework-wise, and in a MARC-dialect agnostic way
23:23:42 <wizzyrea> aha
23:24:13 <thd> Additional problem is having any associated value list translatable.
23:24:32 <tcohen> right
23:25:34 <tcohen> ok, i'll finish the meeting now, i need to leave and you can keep discussing it if you want to
23:25:35 <thd> I have the idea of setting a preference per Koha MARC framework without storing the value in the framework itself.
23:25:47 <tcohen> how would you do it?
23:26:32 <thd> Maybe as a system preference but called for each bibliographic framework.
23:27:08 <tcohen> thd: i think we should discuss it on a bug or RFC
23:27:21 <thd> cait mentioned that YAML for system preferences is being translated.
23:27:44 <thd> tcohen: It certainly needs some more careful consideration.
23:28:07 <tcohen> i need to take care of home stuff (i've just arrived and it is 20:30 already)
23:28:14 <tcohen> so
23:28:25 <tcohen> thanks everyone for your time and ideas/opinions
23:28:34 <tcohen> we'll do this more often
23:28:37 <tcohen> again
23:28:38 <tcohen> :-D
23:28:42 <eythian> later
23:28:47 <tcohen> #endmeeting