14:00:42 #startmeeting Development IRC meeting 13 December 2017 14:00:42 Meeting started Wed Dec 13 14:00:42 2017 UTC. The chair is kidclamp. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:42 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:42 The meeting name has been set to 'development_irc_meeting_13_december_2017' 14:00:48 #topic Introductions 14:00:48 #info wahanui, a bot that has become sentient 14:00:58 #info Jon Knight, Loughborough University 14:01:05 #info Nick Clemens,ByWater Solutions 14:01:18 #info Claire Gravely, BSZ Germany 14:01:41 #info Jonathan Druart 14:01:55 #info Josef Moravec, Municipal Library UO, Czech Republic 14:02:08 rmaints? 14:02:08 hmmm... rmaints is kidclamp (17.11), fridolin (17.05), rangi (16.11) 14:02:39 #info Fridolin Somers, Biblibre France 14:02:44 #info Thomas Dukleth, Agogme, New York City (In California for the end of the year) 14:02:59 #Info Victor Grousset, Biblibre, France 14:03:12 qa_team is alex_a jajm marcelr khall kidclamp tcohen josef_moravec 14:03:14 qa_team? 14:03:15 qa_team is alex_a jajm marcelr khall kidclamp tcohen josef_moravec 14:03:32 #info Julian Maurice, BibLibre 14:03:39 I will try not to abuse of these 2 "aliases" :) 14:03:47 #topic Announcements 14:03:58 anyone have things? 14:04:05 Yes 14:04:06 #link https://wiki.koha-community.org/wiki/Development_IRC_meeting_13_December_2017 14:04:07 #info Alex Arnaud, Biblibre France 14:04:34 qa_team? 14:04:35 qa_team is alex_a jajm marcelr khall kidclamp tcohen josef_moravec 14:04:35 #info Katrin Fischer, BSZ, Germany 14:04:38 :) 14:04:57 Today is my happy birthday. Sadly, I will miss the next meeting for the party. 14:05:09 thd++ 14:05:12 happy birthday 14:05:14 Happy Birthday 14:05:36 #topic Update from the Release manager (18.05) 14:06:00 happy birthday thd 14:06:12 * kidclamp whispers 'that's you Joubu' 14:06:23 nothing special for me, I have started 17.12 pushing quite lot of stuffs 14:06:53 and will continue like that, as far as the QAers send me stuffs to review 14:07:24 #info Joubu is pushing all the things. QA more stuff, more stuff gets pushed. 14:07:36 as it's the beginning of the release cycle, let me know if you want to work on something big/special 14:07:38 #info Claire Hernandez, BibLibre 14:07:44 * fridolin have seen there is in master a new dependancy on ES perl lib 14:07:49 and/or need help on specific works 14:07:52 not yet in koha-perldeps 14:08:23 this will brake sandboxes, one must install manually 14:08:26 I'd like to have a koha-es with zebra-equivalent features for 18.05 14:08:42 #info Owen Leonard, Athens County Public Libraries, USA 14:08:42 or at least to know what is work to be done to reach that state 14:09:06 #info let Joubu know if you have projects so you can coordinate on what is needed 14:09:07 there is a wiki page created by Nick, you are pleased to add items to this list 14:09:12 we have ES section later on too 14:09:13 koha-es? 14:09:26 koha / Elastic 14:09:28 Wiki page URL? 14:09:28 ah.. not spanish, elastic search 14:09:33 *hides* 14:09:34 yep, juste saying koha-perldeps must be corrected 14:09:49 cause dependancy is mandatory even for zebra install 14:09:52 * oleonard hopes Bug 19474 will get some interest 14:09:52 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=19474 enhancement, P5 - low, ---, oleonard, Needs Signoff , Convert staff client CSS to SCSS 14:09:54 kidclamp: do you have the wiki page link? 14:10:07 so this page : https://wiki.koha-community.org/wiki/Elasticsearch_status I'll add things we saw or questionned last week 14:10:07 #link https://wiki.koha-community.org/wiki/Elasticsearch_status#Elasticsearch_status_.28Updated_2017-12-07.29 ES status 14:10:22 yes thanks clrh 14:10:23 yes please clrh - anyone should add or update 14:10:59 in my goal for 18.05: push all what oleonard write to move JS code (we will talk about that later) 14:11:00 moving on 14:11:07 sorry :-) 14:11:09 wait 14:11:24 I'd like to do some benchmarking (last ones are quite old) 14:11:34 Joubu++ 14:11:40 and write more Selenium tests 14:11:43 * Joubu searches a link 14:12:01 see bug 19243 for selenium tests (and related) 14:12:01 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=19243 enhancement, P5 - low, ---, alexbuckley, ASSIGNED , Selenium test for testing the administration module functionality 14:12:02 #info ES, JS updates, benchmarking, selenium tests are some golas for 18.05 14:12:33 I will also push/support tcohen's work (and others) on REST API, ofc 14:12:49 more or less the big lines for 18.05 14:12:49 API++ 14:12:52 That's all for me 14:13:02 Joubu++ 14:13:05 And the manual! :) 14:13:14 #info also REST API and manual 14:13:21 will talk about that during the next doc meeting 14:13:22 Joubu doesn't think small 14:13:46 #topic Updates from the Release Maintainers 14:13:52 Joubu++ ;) 14:13:56 Joubu+ 14:13:59 + 14:14:22 I will be pushing stuff next two days and will announce string freeze for 12/15 to stick to release on 22nd 14:14:45 #info 17.11 string freeze on 15th, pushing things now 14:14:50 fridolin? 14:14:51 fridolin is busy at the moment, I asked him to backport the bug fix 14:15:00 * fridolin RMaint 17.05 14:15:13 tests++ 14:15:14 I'm now waiting on 17.11 aka stable pushes 14:15:42 althow i have some late patches in master i can backport 14:15:55 but no major ones normaly 14:16:00 #info 17.05 waiting on 17.11 and pushing some others 14:16:20 to say something that is maybe not public: We will have secutiry fixes pushed in stable releases regularly. We are not going to announce them as "security releases", because the fixes are not critical. But we are going to follow a security workflow, in the sense of the fixes will not be published before the releases. 14:16:47 ++ 14:17:08 We (RM+RMaints) are trying to work closely and in sync. The workflow is not perfectly defined yet... 14:17:44 rangi si probably not here but he is an rmaint and we like him 14:18:07 #topic Updates from the QA team 14:18:29 Joubu: Announcing after release seems to be an important part of "security workflow". 14:18:49 I have sent some questions to the QA team members to see what we can do to improve things, when people usually QA, how much time they can spend, and what topics we have not covered 14:19:05 I got some answers already - thx! 14:19:20 what we already see is that auth patches get stuck often and easily 14:19:27 because we lack the environment to test them with 14:19:41 would it be possible that we have a cas test server to access somewhere? LDAP? Shibb? 14:20:05 it's also about documentation, if we had simple tutorials on how to set that up, maybe they would be not as hard to tackle 14:20:18 (thd: the idea was to keep "security releases" when it's really needed, not to announce releases as "very important" every months) 14:20:20 cait++ 14:20:33 cait++ 14:20:34 ldap testing is possible, there are test servers out there 14:21:01 so if someone wnat to help out there, that would be great 14:21:02 shib is harder because you need koha accessible from outside and https 14:21:15 yeah, not doable with a kohadev, which rules me outmost of the time 14:21:26 I'd like to repeat that the severity should be used when you pick a patch to SO or QA 14:21:34 we have 21 majors 14:21:39 6 waiting for QA 14:21:50 kidclamp: could you info my need for poeple helping with auth servers and tutorials? 14:22:22 there is no much sense to QA enhs when major bugs can be pushed and backported 14:22:25 #info QA needs help with auth testing and tutorials is an area where we need more people 14:22:38 please also make sure you rebase/react as fast as possible once someone has started to qa your patches 14:22:47 #info Reminder to hit majors and criticals before enhancements please 14:22:50 it's hard to start over when a lot of time is between testing and it eats our resources 14:22:59 With SAML2 authentication its worth bearing in mind there's more than Shibboleth. We use simplesamlphp as our IdP for example. 14:23:07 #info Cait sent a questionaire to QAers, please respond so we can plan how to cover all topics 14:23:22 and also... if you are interested to join QA - please let me know/sign up for vote at the next meeting, even if you just want to help with some specific part 14:23:48 #info if interested in joining QA please let cait know or add to the next meeting agenda 14:23:50 we have saml2 working for simplesaml, shibboleth, athens and adfs idps 14:24:19 Yep, works on our SAML2 IdP... but worth considering if testing patches. 14:24:20 there are always small easy bugs to QA 14:24:29 #info looking for someone to help ease testing for Shibboleth, CAS and LDAP 14:24:34 ashimema, write ups for any of those would be amazing 14:24:39 the shib wiki is pretty well done 14:24:47 (and to rewrite C4::Auth) 14:24:48 (not that I still don't muck it up) 14:24:56 rewrite C4::Auth ++ 14:25:09 (but not by me) 14:25:30 #info Please rebase/answer questins from QA quickly, so things can be finished quickly once started 14:25:53 that's it from me - but others maybe? 14:26:36 #info don't forget to assign yourself as QAer when looking at a bug 14:26:41 * jajm has nothing to add 14:26:44 nothing more here 14:26:59 #topic General development discussion (trends, ideas, ...) 14:27:12 I am thinking of bypassing QA step for some patches, but I really would like to avoid that. There are 150+ bugs in SO 14:27:33 kidclamp: thx, i forgot that myself! 14:27:47 could we organise a "QA day" before end of the year? 14:28:06 is a god idea, sprints have worked well 14:28:09 so far QAers seem to have a preference for Fridays 14:28:11 s/god/good 14:28:22 bt next week might be too close to christmas? 14:29:18 jajm: kidclamp - what do you think? 14:29:23 next week would be ok for me ;) 14:29:23 can't hurt to try? I will be here 14:29:57 i won't be there from 18th to 29th, but i can do a full QA day this friday 14:29:59 let's see after the meeting. For the last one we use a doodle-like page to know the availabilities 14:30:04 sorry josef_moravec! 14:30:25 Joubu: this year is a bit hard :) 14:30:34 i will send a suggestion to qa team tonight 14:30:45 * oleonard will be around next week to help make sure the SO list doesn't get too short 14:30:48 #action cait will try to arrange QA sprint before end of year 14:31:54 Almost anything should be tried including many other bad practises before actually bypassing QA. 14:32:31 Bribing QA team with cake? 14:32:33 i agree with Joubu we need to get the numbers down right now 14:32:42 but we will try to do it properly first ;) 14:32:56 150 is pretty hard to manage and lots probably will have to be rebased 14:33:13 so just put them all 'doesn't apply' 14:33:16 :-) 14:33:49 is the bot which made it for us still alive? 14:34:24 it has been disabled 14:34:50 greenjimll, try with beer instead ;) 14:34:57 people complained about it :( 14:35:00 we could ask rangi, but not sure it is useful 14:35:14 I prefer someone ask me a rebase than a bot 14:35:31 a bot will not SO it, so I rebased it for nothing 14:35:57 valid point 14:36:25 sure, but a tester can't test if doesn't apply 14:36:32 agree, that is good reason 14:36:52 it was to Joubu 14:36:53 maybe have it only check bugs that are awaiting QA? 14:36:57 a tester is supposed to test when rebased, right? :) 14:37:17 ideally 14:37:46 What about a script that could tell the QA team which bugs pending QA don't apply cleanly, rather than the bug assignee? 14:38:47 I think maybe QA and bot is good for next meeting? We havea few other topics for today we should cover 14:39:02 I'd suggest to move on, we have a long topic list today 14:39:27 we can continue the discussion on the ML if people has ideas 14:39:43 i have a quick question? 14:39:55 yep? 14:39:55 yep. 14:40:04 yep! 14:40:07 I was thinking similar to the questions I asked the QAers 14:40:20 shoudl we have a wiki page where we can collect 'experts' on topics and apis? 14:40:27 for people to sign up to if they want 14:40:40 like the "module maintainers"? :-/ 14:40:45 less formal 14:40:55 just a... i am interested in those topics 14:40:56 yes we can try 14:41:27 #action cait will add wiki page to list topics of interest/expertise to assist in SO and QA 14:41:28 it's basically writing down what we more or less know already. So yes it's good to have that list on the wiki 14:41:44 also maybe good for newcomers 14:41:54 is tcohen here/needed for rest api disc? 14:42:12 pm me if you have a good name suggestion for the page 14:43:19 #topic REST api 14:43:38 from the agenda: Proposal to vote for: allowing devs working on the REST api implementation to change attribute names so they are more meaningfull for api consumers. For example, rename 'borrowernumber' for 'patron_id'. Pros: we can provide a readable API spec, and people unfamiliar to Koha internals will have better time using the api. Cons: some people might be using the current (3) endpoints and would need to change some 14:43:38 attribute names. 14:43:41 With the REST API attribute name changes, will there be support for existing attribute names to preserve existing scripts, or is it viewed as a clean break? 14:44:21 "Cons: some people might be using the current (3) endpoints and would need to change some attribute names" 14:44:36 So clean break then? 14:44:45 yes 14:45:10 in that case we should change the version 14:45:17 Is the suggestion implying that the names currently in use in Koha are poorly chosen and should eventually be changed themselves? 14:45:20 was not it the purpose of the v1/v2... 14:45:21 ? 14:45:41 That's usually the use of REST API versioning. 14:45:51 Yes, version change would be the ideal solution probably 14:45:53 Joubu, yep, if you break compatibility, you should change the version 14:46:15 Often older versions are kept (for a while at least) for backward compatibility 14:47:07 And to smooth transition for external developers 14:47:16 There is some history behind some variable names in Koha for wider applicability than the name which might be most common. 14:47:17 I would like to have the list on the wiki, patron_id or user_id, etc. 14:47:48 Joubu++ 14:47:55 thd: we are talking about borrowers/members/patrons/users 14:47:58 for instance... 14:48:44 I understand the topic of discussion and the history behind Koha's current naming choice. 14:48:47 #info general discussion is version change would be good for this proposal 14:49:01 #info and a listing of naming on the wiki 14:49:19 should it be voted, or reworked with version change as proposal for vot enext meeting? 14:50:05 I think we all agree we need to pick good attribute names :) 14:50:10 An appeal to some ISO or other standard for library attribute usage would be good as a basis. 14:50:21 we need to discuss it with tcohen when we will be around 14:50:27 s/we/he 14:50:32 If REST API is aiming for inclusion in 18.05 shouldn't we vote now to give folk time to work on it? 14:50:39 we need to discuss it with tcohen when he will be around* 14:50:55 i think the current case is the patron api 14:50:58 best to comment on the bu 14:50:59 g 14:51:02 next meeting is two weeks so should still have time 14:51:06 greenjimll: we should vote with the name of the attributes 14:51:37 Joubu++ 14:51:47 we already voted to match the GUI 14:51:56 intead of the database 14:52:10 that's why patron_id instead of borrowernumber for the api 14:53:04 i think if you want to vote on each, maybe have a separate meeting soon? 14:53:17 we need to move forward with the Rest API 14:53:28 I assume its more than just "rename 'borrowernumber' for 'patron_id'" though as that's just one example. We don't have a list of attribute names to vote on do we? 14:53:55 no we don't 14:54:04 #info vote postponed to next meeting so can discuss versioning with tcohen and get a list of attributes 14:54:18 I think enough discussion to wait, let's move on 14:54:19 Why is the GUI actually different to the rest of the code? 14:54:55 #topic SQL strict mode is not the default in MariaDB 14:54:57 lol 14:55:13 yes, for me. Just a sec 14:55:41 it's that one: bug 17860 14:55:41 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=17860 normal, P1 - high, ---, gmcharlt, NEW , [SQL strict mode] MARC records will stage, but Manage for import shows 'no data in table' 14:55:46 see last comment 14:56:09 so basically we know for a while that Koha was not MySQL 5.7 compatible (with the default sql-mode option) 14:56:26 our suggestion was to switch to MariaDB 14:56:45 but they decided to follow MySQL and now also use the trans_strict_table sql mode 14:57:20 #info Koha has compatability issues withMySQL 5.7 so we suggested mariadb, but they now follow MySQL and same problem occurring 14:57:23 if people are interested (I mean, really, SO+QA) I can take a look to estimate the work and provide patches 14:57:43 last time I did that (6months, 1y ago), nobody cared 14:58:20 it should be done, we should work towards supporting the versions in stretch probably 14:58:28 we also need to update the wiki (install debian) 14:58:37 Would there be an issue for backwards compatibility to using earlier versions of MySQL or MariaDB? 14:59:00 looks like we should start to care about this 14:59:29 it's causing strange problems - reported on the mailing list and bugzilla 14:59:39 even with a workaround we should care 14:59:47 https://wiki.koha-community.org/wiki/Koha_on_ubuntu_-_packages#Ubuntu_16.04_and_MySQL_5.7 14:59:56 that section needs to be updated 15:00:00 * LibraryClaire has to run 15:00:05 #action? 15:00:45 not sure what to write - sorry 15:01:05 #info ping Joubu if you have time to dedicated to mariadb/mysql strict mode 15:01:20 * fridolin noob for SQL server 15:01:35 #action kidclamp Update the Koha_on_ubuntu_-_packages#Ubuntu_16.04_and_MySQL_5.7 section 15:01:38 :D 15:01:50 man, I weas just typing for someone else, too, I don't know the answer 15:01:56 I'll tell you when RDA is ready 15:02:15 lol 15:02:28 #action kidclamp will send an email to the list about the db issues 15:02:47 I'll tell you when HTML 5 is ready :) 15:02:49 Joubu (neither increment or decrement) 15:03:03 #topic Elasticsearch status 15:03:25 This was discussed earlier, there is more movement on ES lately, have been a few more testers out there 15:03:36 please add anything you wish to the status page 15:04:01 please focus for now on fixing the broken things and compatability so that 18.05 can havea stable ES 15:04:12 (the kanban may be more useful to keep an updated overview) 15:04:17 once that is done there are many things that can be added 15:05:03 some BibLibre guys had a look to the existant Koha implementations, a lot of questions asked to ourselves, I must edit the wiki community page and continuing testing 15:05:27 please do clrh and feel free to ping me for any questions 15:05:34 biblibre++ 15:05:48 it was last week and only a "one day focus" but I must continue 15:06:01 yep kidclamp thanks 15:06:22 #topic Javascript in footer 15:06:30 oleonard: ? 15:06:56 Thanks to everyone who has helped move these patches quickly through 15:07:23 There are very few templates yet to do 15:07:59 Getting this done is an important step not just for performance reasons but to lay the groundwork to some JS preprocessing in the future 15:08:20 It is the kind of patches we should not wait long before pushing them, they do not survive to a lot of rebases 15:09:11 josef_moravec++ for following QA there 15:09:20 Does anyone have any questions about this move, or about how to adapt to this new practice? 15:09:21 This is the current status: https://bugs.koha-community.org/bugzilla3/showdependencygraph.cgi?id=17858 15:09:40 #link https://bugs.koha-community.org/bugzilla3/showdependencygraph.cgi?id=17858 JS move status 15:09:54 I plan to finish the QA of this as soon as possible 15:10:09 QA team need to be aware of that change, to avoid introducing JS code back in the header 15:10:41 We should have a rule for this probably 15:10:53 I have no idea how to write QA tests for that, please take a look if you want to add some 15:10:54 afk - phone call 15:12:10 you volunteer? 15:12:31 #action josef_moravec add a new coding guideline about the JS move to the footer 15:12:38 thanks josef_moravec ;) 15:12:48 Joubu is on a roll today 15:13:15 moving on? 15:13:16 2 more topics, quickly? 15:13:45 aside form the coding guidelines? go for it 15:14:16 nope, I was talking about the 2 coding guidelines 15:14:44 #topic Review of coding guidelines 15:14:51 Joubu: not at all ;) I'll do it 15:14:51 #info Don't use DBIX calls in updatedatabase 15:15:16 #info see bug 19789 and bug 17292 15:15:16 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=19789 major, P5 - low, ---, koha-bugs, NEW , Remove some indirect DBIx calls from updatedatabase 15:15:17 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=17292 major, P5 - low, ---, kyle.m.hall, Pushed to Stable , Use of DBIx in updatedatabase.pl broke upgrade (from bug 12375) 15:16:26 did anyone take a look at that? 15:17:09 looking now, seems reasonable 15:17:32 I though we already agreed on that, but did not find discussion or coding guidelines 15:17:36 Is the problem slightly different to what is supposed. 15:18:09 we need to write a coding guideline and vote it during the next meeting 15:18:24 Is the manner in which DBIx is being used in error rather than using DBIx instead of DBI? 15:18:41 #action Joubu will write a coding guideline to not use DBIX in updatedatabase for voting at next meeting 15:18:50 :D 15:19:00 #action Joubu write a coding guideline to forbid use of DBIX calls in updatedatabase (and C4/Koha ns?) 15:19:06 ha! 15:19:14 No one problem is that during updatedatabase schema and db may not match 15:19:51 cc_: yes, that's the point 15:19:58 #info Reminder - git commit messages 15:20:15 so fetch/insert/etc must be done writing SQL queries 15:20:25 cc_: Yes, that is an underlying problem which should be fixed by better usage of DBIx should it not? 15:20:31 #info ('Bug 12345: ', '(follow-up)', '(QA follow-up)', see git log) + what the patch does vs what the bug is. 15:20:31 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=12345 enhancement, P5 - low, ---, oleonard, NEW , Can't Clear Reading History for Anonymous patron 15:21:05 A quick reminder about the commit messages (I need to write a coding guidelines too), please take a look at the last commit messages in master about the formating 15:21:48 I update them before push, but I may fail them soon if they are written after the anouncement 15:22:01 #action Joubu will write a coding guidline on commmit message formatting for voting next meeting 15:22:52 an common mistake is also "what the patch does" vs "what the bug is" 15:22:52 okay, Joubu. 15:23:11 is "(follow-up)" really needed ? 15:23:15 the commit message title must describe what the patch does, not a copy/paste of the bug report title 15:23:32 jajm: it's needed if you want to add a follow-up 15:23:58 I sent an email to the list few months ago about that, I explained when you should use it (in my opinion) 15:24:09 Are the commit messages (for bug 19410) supposed to be examples of good or bad practice? 15:24:09 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=19410 enhancement, P5 - low, ---, tomascohen, Pushed to Master , Add a helper function for generating object searches for the API 15:24:12 Joubu, i'll reread this email 15:24:27 I need to write a guideline to vote and put that to the wiki 15:24:31 Joubu++ 15:25:07 greenjimll: the commit messages on the bug report are different than the ones in master 15:25:22 I edit them before push 15:25:34 I'm looking at the ones in `git log`, so are these bad/good? 15:25:36 but did not attach the updated patches to the bug report 15:25:50 git log origin/master should be ok 15:26:01 --oneline 15:26:06 Ah, that's OK then as I couldn't see what you were complaining about. :-) 15:26:52 next meeting? 15:26:53 next meeting is https://wiki.koha-community.org/wiki/Next_IRC_meetings 15:27:22 #topic Set time of next meeting 15:27:29 I cannot 15:27:36 #topic Set time of next meeting 15:27:48 #chair Joubu 15:27:48 Current chairs: Joubu kidclamp 15:28:06 #topic Set time of next meeting 15:28:26 27 December 2017, 20 UTC? 15:28:31 +1 15:28:34 27 December? 15:28:52 Works for me. Something to do instead of eat sweets and watching bad Xmas TV. :-) 15:28:57 the 'Add pagination to vendors endpoint' and the one renaming attributes in patrons are good examples on how to use the helpers 15:29:17 * tcohen says while on a taxi 15:29:27 #info Next meeting: 27 December 2017, 20 UTC 15:29:32 * thd will be in an airplane returning to New York. 15:29:47 #endmeeting