15:06:37 #startmeeting Development IRC meeting 15 April 2015 - part 1 15:06:37 Meeting started Wed Apr 15 15:06:37 2015 UTC. The chair is cait. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:06:37 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:06:37 The meeting name has been set to 'development_irc_meeting_15_april_2015___part_1' 15:06:52 #topic Introductions 15:06:52 #info wahanui, a bot that has become sentient 15:06:58 please introduce yourself with #info 15:07:17 #info Jonathan Druart, BibLibre 15:07:20 #info Colin Campbell PTFS-Europe 15:07:22 #info Zeno Tajoli, CINECA Italy 15:07:28 #info Kyle Hall, ByWater Solutions 15:07:29 #info Thomas Dukleth, Agogme, New York City 15:07:48 #info Brendan Gallagher, ByWater 15:08:34 #info Katrin Fischer, BSZ, Germany 15:09:15 #info Fridolin Somers, Biblibre, France 15:09:20 #info Jacek Ablewicz, Cracow Univ. of Technology 15:10:44 hm pinged tcohen - no response 15:10:56 hi 15:11:01 oh goood! 15:11:03 sorry, really busy day 15:11:08 can you info please? :) 15:11:11 give me 2 minutes 15:11:24 #info Tomas Cohen Arazi, Universidad Nacional de Córdoba 15:11:27 i got 2... we will give you 2 too :) 15:11:29 #topic RM 3.20 comments 15:13:18 hi everyone 15:13:42 i've been troubled and i'm sorry for not being around too much, i'm almost back 15:14:03 and will try to use most of my office time for RM tasks until release 15:14:26 there's been a lag on pushing enhancements and new features 15:14:41 i tried to focus on bugfixes once i got back 15:14:51 so they hit the stable branches ASAP 15:15:01 good plan of attack 15:15:09 so, I'm in the process of testing/pushing enhancements and new features 15:15:44 tcohen++ 15:15:51 3.20 is looking good so far, stable, issues have been fixed really quick once we noticed 15:16:05 so 15:16:06 koha++ 15:16:29 we need to set a cut day for enahncements 15:16:57 and I'd say stuff that isn't passed-qa next week should be left for 3.22 15:17:14 friday next or monday next week? 15:17:23 i have some concerns about some stuff on my queue (next topic) 15:17:28 but we can talk about them 15:17:35 next week: wednesday :-D 15:17:59 so we can release a beta during the weekend or so 15:18:02 soon now :) 15:18:23 you had fun filling my queue, you had enough time .-D 15:18:31 just to be precise.. the wednesday still counts? your timezone? or tuesday midnight your timezoneß ;) 15:18:34 * cait kows the devs 15:19:07 i come to the office at 9am ART 15:19:21 around 9:10 thursday :-D 15:19:38 i'll take a screenshot at that time 15:19:43 =D 15:20:10 #info Freeze for 3.20: What's not passed QA by April 30, 9:10 Argentinian time... won't make it in 15:20:42 Good morning all 15:20:49 so you want one more week? :-D 15:20:56 arg. 15:21:07 i just can't type 15:21:20 #info CORRECTION: Freeze for 3.20: What's not passed QA by April 22nd, 9:10 Argentinian time... won't make it in 15:21:31 sorry! 15:21:46 #info CORRECTION 2: Freeze for 3.20: What's not passed QA by April 23nd, 9:10 Argentinian time... won't make it in 15:21:57 =D 15:22:01 oh my 15:22:06 could you have said that a bit earlier? 15:22:12 #chair tcohen 15:22:12 Current chairs: cait tcohen 15:22:15 concerns? 15:22:42 Is it possible, with xslt, to display the authorised value's description but build a search link on the authorised value's code? 15:23:18 francharb: currently not i think 15:23:33 #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 15:23:44 thanks cait ! Sorry, I just realised there was a meeting... 15:23:52 francharb: it might be possible 15:24:00 move on 15:24:17 francharb: (nope, we use reverse maps desc to code as a hack) 15:24:28 #topic General technical discussion 15:24:53 anyone willing to raise something on this subject? 15:25:40 Not sure if this is the topic for it, but I would *really* like to see bug 13799 signed off. I've been planning to QA it. I'd love to see it in the next release 15:25:41 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 15:26:18 i think we should move to the next topic, and discuss that there 15:26:25 ok 15:26:26 #topic Bugs 15:26:43 first one 15:26:44 somebody said first one was looking weird - it says 9 centerville, but there is also an item at liberty 15:26:47 Moose dependencies (Bug 11190) 15:26:48 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=11190 new feature, P5 - low, ---, frederic, Passed QA , sitemap.pl -- Generate a Catalog sitemap 15:26:49 I can say something about metasearch 15:27:10 Ok after this topic 15:27:33 tajoli: use the "Big stuff youre working on section" 15:27:40 Ok 15:27:56 has anyone seen the tool introduced by bug 11190? 15:28:16 it is a really interesting one, and i'd like it to make it for 3.20 15:28:27 the only concern is that it introduces Moose dependencies 15:28:46 hm in the past we had decided against that 15:28:57 true 15:28:58 is it an optional feature? could the dependency be optional? 15:29:10 it is an optional script 15:29:13 i think benchmarks indicatd that we shoudl favor moo over moose - or that was the compromise 15:29:18 that needs to be cron-ed 15:29:58 cait: is Moo a dependency already? 15:30:06 i am not entirely sure, it might be 15:31:20 hm 15:31:25 http://git.koha-community.org/gitweb/?p=koha.git;a=blob;f=install_misc/debian.packages;hb=2e1d64d056a63c0eac412c2cbb7b72da5bd4a08c - is libmoose already there? 15:31:26 it isn't 15:31:44 that's not used I think 15:32:20 cait: C4::Installer::PerlDependencies 15:32:25 ok 15:33:05 anyone? opinions besides the comment that we historically rejected Moose? 15:34:32 ok, I'll try on part 2 15:34:42 Moose is quite heavy 15:34:50 to very interesting to add the deps for 1 script 15:34:56 s/to/not 15:35:22 i'll try using Moo, probably enough for the simple sitemap.pl script 15:35:31 (and its related packages) 15:35:32 +1 agreeing with Joubu 15:35:36 it seems too heavy for 1 script if we don't intend to use it in general 15:36:03 the question is: it doesn't have any impact on how Koha responds 15:36:04 i think if Moo isn't used yet.. we shoudl have some Class::Accessor in the codebase 15:36:07 not sure if that woudl work 15:36:08 speed or anythin 15:36:25 it is just if it needs to be listed as a dependency or not 15:36:50 i'll talk to fredericd about it, maybe it is an easy change 15:37:00 ok, moving on for now 15:38:05 #info people didn't like to include Moose as dependency, even for a cli script. tcohen will approach fredericd to see if we can make the changes needed for it to make it into 3.20 15:38:39 #action tcohen to approach fredericd about Moose dependency for bug 11190 15:38:40 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=11190 new feature, P5 - low, ---, frederic, Passed QA , sitemap.pl -- Generate a Catalog sitemap 15:38:43 #info if part 2 concludes the opposite, a vote will be held ASAP 15:39:12 moving on 15:39:25 what will we do with the QueryParser ? 15:39:37 it is an awesome idea 15:39:42 but an orphan one 15:40:19 and we have a situation where a feature targets non-QueryParser setups 15:40:23 bug bug 13528 15:40:24 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=13528 enhancement, P5 - low, ---, jonathan.druart, Passed QA , Ability to search on the different isbn variations 15:40:37 Joubu: can you explain why it won't work with QueryParser? 15:41:31 I remember an email about a problem with mulitple arguments... is that related? 15:42:33 cait: don't remember exactly, I sent an email on koha-devel 15:42:40 hm that might be the one i remembe rmaybe 15:42:42 because QueryParser is broken with several operators 15:43:02 I feel like it shouldn't stop this for now 15:43:06 a OR b OR c results in a AND b AND c 15:43:09 something like that 15:43:49 so when you want to check for barcode : bc = $bc1 OR bc = $bc2 OR bc = $bc3, it causes some problems... 15:43:51 and this seems like a really good enhancement that we coudl extend later to queryparser maybe 15:44:04 but the pref shoudl be clear that there is a catch 15:44:05 What are the consequences of disabling QueryParser aside from allowing the patch for ISBN searches to work? 15:44:05 bc = 'foo' AND bc = 'bar' will never work 15:44:19 thd: we are not usign it in any of our installations currently - it's optional 15:44:20 I'm not seeing any benefits from QueryParser being in the release at present 15:44:37 and queryparser is off by default 15:45:21 gmcharlt_: available? 15:45:30 jcamins: available? 15:45:46 It does further complicate some already complicated code which is a pain for code maintenance 15:46:25 would be good to know if Robin planned to use it 15:46:28 ColinC: it should "simplify" code actually. Changing ad-hoc code for YAML configuration on how to interpret/translate operators 15:46:35 (for ES) 15:46:48 tcohen: but the problem with operators is a pretty big blocker ot use it :( 15:46:56 the problem is we have several problems that prevent it from fully replacing the ad-hoc code that "just works" 15:46:59 also, the swtich was not completed i think 15:47:00 I imagine that having some possible query parser would be a very good thing but perhaps not the present code. 15:47:04 it doesn't work in all places it could be used 15:47:05 Joubu: exactly 15:47:42 Joubu: woudl it be hard to extend your feature when the operators bug got fixed? 15:47:46 and do we have a bug report for it? 15:48:23 i think i'd be ok with pushing your patch if the problem is documented in the pref description and we also have a bug on bugzilla... so we know where the problems are 15:48:38 Who would be able to fix it? 15:49:09 i think ripping the code out now woudl not be good 15:49:17 we need to talk more about this after 3.20 15:49:22 the queryparser code 15:49:29 and fixig it... also a bit short on time now 15:51:43 um.someone else? 15:52:08 What was the question? :) 15:52:16 Should we remove it? 15:52:21 #action tcohen will go ahead with the inclusion of bug 13528, along with a followup stating that it won¡t work with UseQueryParser=on 15:52:22 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=13528 enhancement, P5 - low, ---, jonathan.druart, Passed QA , Ability to search on the different isbn variations 15:52:33 ifit should block bug 13528 that it doesn't support queryparser 15:52:34 Should we accept ft even if does not work with QP? 15:53:21 Joubu: you only found a bug on QueryParser's PQF driver 15:53:32 I think there are a number of sysprefs that dont work if QP enabled already 15:53:33 I won't answer, I am the author of the patch :) 15:53:34 i think we just need to fix it 15:53:51 'just' :) 15:54:02 I tried, I failed. Good luck 15:54:03 i mean, i don't think it blocks your feature 15:54:57 I'm in favour of pushing the feature with a docuentation followup 15:55:16 and will definitely talk to gmcharlt_ about Evergreen's use of QueryParser 15:55:23 because they've been using it for years now 15:55:40 and it doesn't fail like this, so it is specific to Koha's driver 15:56:10 tcohen: not sure you have to highligh it in the doc, the limitation is in the syspref description 15:56:12 Joubu: you wrote a regression test for the QA problem I recall? 15:56:32 Joubu: i mean the syspref docs/description 15:56:41 ? 15:56:56 i was talking about the syspref description 15:56:59 wording issue 15:57:06 move on? :) 15:57:17 i think so 15:57:29 will see later :) 15:57:52 * tcohen knows Joubu is hopeless 15:58:21 next one is bug 13912 15:58:22 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 15:58:51 tcohen: You should note my comments to the bug added today. 15:59:06 * tcohen goes to read 15:59:33 #info Barton Chittenden, BWS, Louisville, KY 16:00:36 As user of unimarc: 16:00:47 #info Mark Tompsett 16:00:55 better a develop Marc-neutral 16:01:24 but if not, Ok on name 'DefaultCountryField008' 16:01:31 While tcohen reads I will summarise that my main objection is that an empty value in a system preference should not set the country for an installation. 16:02:04 baecause now i use a setup on marc frameworks 16:02:27 but in unimarc 008 ins not present 16:02:39 so doesn't' do anything 16:02:40 I think that the MARC 21 bias is less serious but should still have some more neutral presentation. 16:02:45 bye 16:04:02 Explaining, if the user has not set the system preference, the default should also not be set and a warning should be triggered. 16:04:05 Yes on 'empty value set nothing' 16:04:23 thd: I agree with your comments, just a little difference 16:05:14 I would set the country_of_production (and for example original_language) framework-wise 16:05:31 so it would be easier to link to an authorised value list too 16:05:51 the use case that raised my concern is thesis here 16:05:57 that are written in spanish 16:06:14 tcohen: That would require work to modify the Koha MARC frameworks design to support fixed fields better. 16:06:14 but most of the bibliography is originally written in english 16:06:42 thd: just a new field for the frameworks 16:06:53 and adjust the marc_21_008.pl plugin 16:07:04 extra work, I know 16:07:47 tcohen: How would the new column for the frameworks specify the range to which part of a fixed field it applies? 16:08:29 don't need to, the the specific plugin hould know 16:09:10 anyway 16:09:44 is the introduced technical debt something to worry about? 16:09:51 tcohen: Would you have any and all values in the fixed field specified in the default field? 16:10:08 tcohen: If so, then no new column is required. 16:10:33 tcohen: There is already a defaultvalue column. 16:10:51 thd, that works for subfields 16:10:59 tcohen: Yes. 16:11:14 i would rather add country_of_production to `biblio_framework` 16:11:18 tcohen: to some extend it works for controlfields as well 16:11:39 cait: you mean adding the full fixed lenght string? 16:11:42 yeah 16:11:51 i'm being empathetic with librarians 16:11:56 empathic? 16:12:06 i am not sure this small feature is worth too much work right now 16:12:24 tcohen: That was also my current suggestion with no sympathy to librarians ;) 16:12:41 -- displaying empathy for. ;) 16:13:21 ^for^with^? Anyways... 16:13:22 * tcohen loves having idiomatic aid 16:14:45 tcohen: Adding a column with such specificity to the framework would essentially have an almost entirely empty column. You should think of a more generic approach which uses the frameworks. 16:14:48 i'll think of cait's idea of using the default value, maybe it is worth making a better value builder for the framework editing tool 16:15:22 the only problem might be if the plugin correctly detects it always 16:15:38 but it hink i would not blow this up to be too huge - and also not sure it need to be by framework, might be ok globally 16:15:41 * mtompset thinks he better scroll back to get fully context. 16:15:58 shoudl we move on? 16:16:01 cait: i give up 16:16:09 i hope not! 16:16:30 once i have a proper solution (not just ideas) that convinces me, i'll come back to this 16:16:36 There are many problems which the frameworks could be extended to address. Maintenance of the default frameworks is the difficulty. 16:16:49 thd: tru 16:16:50 e 16:17:59 thd: are you able to provide a followup for the bug? 16:18:12 (implementing the changes you propose?) 16:18:39 I actually think that we should be moving the generic form of things out of the existing Koha MARC frameworks and use a form which will work with BIBFRAME or whatever. 16:19:14 I am able to follow up on the bug. Even the full featured version would not be difficult but merely take some time. 16:19:21 Yuck, system preference?! 16:20:16 Sorry, I need to leave. I will write about metasearch on koha-devel 16:20:17 thd: i think it can't go in changig existing behaviour if the pref is not set 16:20:21 as qa point of view 16:20:22 thd: just renaming the syspref and making sure a good default behaviour? 16:20:27 so we fqa is right and it needs a follow up 16:20:28 System preferences may not be the ideal location but the Koha MARC frameworks model is too inflexible for some important tasks of which the present case is not really an example. 16:20:45 True, thd. 16:21:14 But the advantage of a column is the whole Koha::Object thing we've started. 16:21:57 I can work on changing the fix to do something more global where the plugin would also need changing to give a warning if not set. 16:22:14 And if you are worried about blank columns, split off a table? 16:22:44 thd: thanks 16:22:55 i'll keep an eye on it 16:23:12 mtompset: Yes, we should see if we could find a better place for expressing what is needed. 16:23:34 ok, we need to move on 16:23:53 thd: related to that on "remaining new features" i will mention an interesting bug 16:24:22 there have been several bugs filled proposing the addition of new fields to XSLT's 16:24:43 to cope with RDA and whatever metadata is missing on the XSLT's 16:25:11 i want to say that i'm concerned about people relying on the shipped XSLTs 16:25:15 for regular use 16:25:28 so this changes would be a behaviour change for them 16:25:39 probably a positive one 16:26:06 but I heard several times that people prefered the non-XSLT view because it was "not as bloated as the XSLT view" 16:26:29 -- Shouldn't both views aim to be identical? 16:26:31 hm i can see your point... but it woud be really easy to hide htings from the xslt if a library felt that way 16:26:54 mtompset: the normal view hasn't been maintained 16:26:59 in forever. 16:27:09 deprecated since 2013 I think 16:27:22 Ah. 16:27:55 * tcohen just realised... this discussion is void... if we make field/subfield visibility work... damn 16:27:58 we can easily help everyone who wants to to hide something... much harder to help adding something 16:28:07 also that yes 16:28:14 d*mn 16:28:16 bug 10589? 16:28:17 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=10589 enhancement, P5 - low, ---, mtompset, Failed QA , Override OpacHiddenItems based on Patron Category 16:28:26 Oops... bug 11592 16:28:27 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=11592 normal, P5 - low, ---, mtompset, Signed Off , opac detail scripts do not respect MARC tag visibility 16:28:48 Is that what you are thinking, tcohen? 16:29:09 mtompset: almost, my XSLT approach, but basically yes 16:29:42 #actions the RM won't argue about patches that add fields to XSLTs 16:29:51 #action the RM won't argue about patches that add fields to XSLTs 16:30:09 #actions tcohen not to argue about patches that add fields to XSLTs 16:30:10 :P 16:30:17 Third time's a charm! :) 16:30:25 remaining new features? 16:30:40 tcohen: please stop to argue about patches that add fields to xslts 16:30:47 Field subfield visibility is actually something which had been controlled at one time by arcane values set in the Koha MARC frameworks even for the OPAC. 16:31:20 thd: the arcane values are nice checkboxes now - bu tit's not taken into account in a lot of places 16:31:32 thd: it is but needs some fices 16:31:33 thd: we often export a COMPLETE marc record, including things like internal notes 16:31:34 fixes 16:31:44 moving on 16:31:50 "Remaining new features" 16:32:05 cait: Yes, I know that they are honoured more in the breach than the observance. 16:32:38 i0d like to raise bug 10480, which is basically passed-qa 16:32:39 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 16:32:58 (i.e. signed by two QA team members) 16:33:33 Joubu: would you encourage another QA member to take a look? 16:33:44 i see the value of the bug 16:34:03 that one is on me i think... i said i would, but time has been running away 16:34:18 my next big thing I need to work on for next week is the tax related patches 16:34:28 and they will take time 16:34:31 cait: I can take that one on then 16:34:38 thx khall :) 16:34:39 khall++ 16:34:41 np! 16:35:46 thanks again, khall 16:35:47 and 16:35:53 marcelr++ 16:36:23 is anyone doing the REST API implementation present? 16:36:48 jajm_: ? 16:37:05 no 16:37:13 too late by now i think 16:37:34 bug 13799 16:37:35 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 16:37:39 #actions cait will write Julian to ask him for a quick update to koha-devel about the status of the REST API implementation 16:38:10 ah what? 16:38:11 well, ah is that what goes after the last 16:38:13 I've already emailed Julian to ask if he can followup on robin's comments 16:38:47 ok, so maybe that's ok for now? 16:39:05 I don't know, depends on if more requests are more motivating ; ) 16:39:08 that's ok for now 16:39:18 tcohen: all patches need signoffs :) 16:39:41 ok 16:39:44 i think we didn't agree on the terms 16:39:49 holds/reserves patrons/borrowers 16:40:02 woudl be good to have a general guideline - going by the code or going by the gui 16:40:10 really? holds and borrowers 16:40:14 :-D 16:40:15 * mtompset notes that 'In Discussion' are beginning to be close to, if not exceed 'Needs Sign Off'. 16:40:36 I'd like to see our internal terminology match our external terminology 16:40:59 i'd like to make our internal terminology match the one we use externaly 16:41:08 * mtompset agreed. 16:41:10 we have members, borrowers, patrons... only consistent palce currently is the gui 16:41:16 khall: you type faster than me 16:41:16 i'd say new code should follow it 16:41:23 the new object system gives us a chance to start making that move without radical searching and replacing 16:41:42 ok, I need to end the meeting 16:42:06 a physical one requires me to leave, and I already missed lunch 16:42:09 should we info something? 16:42:15 tcohen: Do we have a next meeting date? 16:42:16 poor tcohen 16:42:19 tonight 16:42:21 22 utc 16:42:25 yup 16:42:38 tcohen: Do we have a next meeting date for after today? 16:42:39 thd, the week after the freeze 16:42:50 maybe wednesday too, same hours 16:42:57 to check the remaining stuff for the release 16:43:00 ok? 16:43:07 i'm leaviiiiing pleaseeeeeeeee 16:43:08 makese sense to me 16:43:09 go :) 16:43:12 thanks 16:43:15 :) 16:43:20 thanks everyone for attending! 16:43:27 #info next dev meeeting probably wednesday week after freeze - april 29 16:43:28 some should endmeeting. 16:43:33 #endmeeting