19:02:19 <cait> #startmeeting Development IRC Meeting 22 March 2016 19:02:19 <huginn> Meeting started Tue Mar 22 19:02:19 2016 UTC. The chair is cait. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:02:19 <huginn> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:02:19 <huginn> The meeting name has been set to 'development_irc_meeting_22_march_2016' 19:02:26 <cait> #topic Introductions 19:02:26 <wahanui> #info wahanui, a bot that has become sentient 19:02:39 <cait> please introduce yourself with #info, following wahanui's example 19:02:47 <cait> #info Katrin Fischer, BSZ, Germany 19:02:51 <cait> #chair bag 19:02:51 <huginn> Current chairs: bag cait 19:02:59 <cait> today's agenda is at 19:03:05 <cait> #link https://wiki.koha-community.org/wiki/Development_IRC_meeting_22_March_2016 19:03:45 <cait> hm anyone around? 19:03:55 <cait> pianohacker: ? 19:03:59 <bag> hello 19:04:19 <bag> #info Brendan Gallagher - ByWater 19:04:28 <pianohacker> #info Jesse Weaver - ByWater 19:04:51 <pianohacker> we aren't too late for EU and too early for NZ, are we? 19:04:54 <gmcharlt> #info Galen Charlton - ESI 19:05:19 <eythian> wahanui: anyone around is <reply>no one but us, $who 19:05:20 <wahanui> OK, eythian. 19:05:32 <cait> it's 8pm in europe, so after work hours 19:05:45 <eythian> 8am in NZ 19:06:15 <cait> kind of in between i guess 19:07:16 <pianohacker> guess we can wait until 15 after or so for stragglers? 19:08:26 <cait> yeah, let's wait a bit 19:11:13 <cait> rangi is still on the bus 19:11:25 <cait> hi drojf - please introduce yourself 19:12:08 <tcohen> #info Tomas Cohen Arazi, Theke Solutions 19:13:13 <jajm> #info Julian Maurice, BibLibre 19:14:13 <cait> ok 19:14:21 <cait> I think we can continue 19:14:28 <pianohacker> barton is coming 19:14:35 <cait> #Announcements 19:14:39 <cait> #topic Announcements 19:14:43 <barton> #info Barton Chittenden, Bywater Solutions, Louisville KY, USA 19:14:51 <cait> bag? 19:14:51 <wahanui> it's his birthday! 19:14:55 <bag> HA 19:15:10 <bag> not much to report. working on PQA 19:15:14 <barton> I thought it was Trae's birthday. 19:15:38 <bag> I also plan to have a talk with gmcharlt later this week dealing with a URL::Encode package isue 19:15:52 <bag> gmcharlt: and I have already communicated about this. More to follow 19:16:08 <cait> ok 19:16:09 <bag> we both feel that it should be rewritten using some packages that are already in koha 19:16:28 <bag> now I just need to find someone to do that ;) 19:16:40 * gmcharlt sneezes 19:16:56 <cait> I got something too: I'd like to remind about the GBSD on Thursday this week 19:17:03 <cait> #link https://wiki.koha-community.org/wiki/2016-03-24_Global_bug_squashing_day 19:18:01 <cait> anything else? 19:18:01 <wahanui> anything else is just being crap 19:18:10 <bag> any questions for me? 19:18:14 <drojf> #info mirko tietgen, berlin, germany 19:18:48 <cait> quiet... :) 19:18:50 <cait> ok, moving on 19:19:01 <cait> #topic Review coding guidelines 19:19:11 <cait> there are no suggestions for new coding guidelines on the wiki 19:19:50 <cait> after discussion on the mailing list it was decided to remove PERL12 19:19:54 <pianohacker> just killing that one 19:20:00 <cait> and also remove the remaining versions from the source code 19:20:08 <cait> pianohacker: do you have the bug number on hand? 19:20:29 <pianohacker> no :/ 19:21:03 <rangi> #info chris cormack, catalyst 19:21:40 <cait> bug 16011 19:21:41 <huginn> 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=16011 enhancement, P5 - low, ---, jonathan.druart, Signed Off , Remove $VERSION from our modules 19:21:57 <cait> if noone disagrees i will log that as an info 19:22:27 <cait> #info PERL12 to be deleted, remaining $VERSION will be removed by bug 16011 19:22:42 <pianohacker> fine by me 19:22:54 <cait> i think no khall - and i think he won't have time for the revamp of the coding guidelnes for a bit longer :) 19:23:13 <pianohacker> nope :) 19:23:15 <cait> so i'd jump over that if noone wants to add something 19:23:21 <pianohacker> do we want to vote on the idea, at least? 19:23:28 <cait> we already did 19:23:38 <cait> he wanted to add a new page with the suggested changes 19:23:54 <cait> and that was agreed on 19:24:22 <cait> does someone have other things for thistopic? 19:25:12 <cait> ok, moving on then 19:25:27 <cait> #topic Bugs in discussion 19:25:34 <cait> bug 9254 19:25:35 <huginn> 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=9254 enhancement, P2, ---, vfernandes, In Discussion , New option to block patrons from changing messaging preferences in OPAC 19:25:45 <cait> A lot of people have asked for it, but it's kind of stuck 19:26:26 <cait> It would be great to get some comments and see this moving again 19:27:02 <cait> there was some disagreement about which options to offer 19:27:20 <pianohacker> bag: do we have any partners that can poke that who would want it? 19:27:44 <bag> not sure 19:27:59 <cait> we currently hide it with jquery for a lot of ours... but a pref would be nicer (and safer) 19:28:44 <rangi> hmm 19:29:08 <rangi> seems the patches are a bit broken 19:29:19 <cait> nengard commented she#d like it ;) 19:29:21 <rangi> Works as described. But the patch only hides the link, 19:29:21 <rangi> not disable the feature. If you refresh OPAC after unset 19:29:21 <rangi> EnhancedMessagingPreferencesOPAC, you can still change 19:29:22 <rangi> your messaging preferences. 19:29:32 <barton> pianohacker, bag: I had a ticket requesting that recently, I'll dig it up. 19:29:33 <rangi> that doesn't seem finished to me 19:29:50 <cait> thx 19:30:00 <cait> ... there are more stuck in discussion... but have to start somewhere :) 19:30:11 <cait> anyone else? 19:30:32 <cait> moving on? 19:31:02 <cait> #topic General development discussion 19:31:12 <cait> ok, there are 4 topics listed 19:31:40 <cait> we already discussed a bit about the merge of the borrowers tables at one of the last meetings 19:31:47 <cait> i am not sure if we have new info on this? 19:31:55 <pianohacker> I don't think so 19:31:58 <pianohacker> and no Joubu 19:32:11 <cait> ok 19:32:23 <cait> Preventing XSS attacks 19:32:34 <cait> our first attempt failed - anyone new info/ideas to share? 19:32:42 <rangi> no 19:32:48 <rangi> just that it is the right direction 19:32:57 <rangi> it just has to be done in a faster way 19:33:07 <rangi> ie, everything should be escaped by default 19:33:32 <rangi> The approach was dead right, it was just not performant 19:33:46 <pianohacker> why was the previous one so slow? 19:33:58 <cait> i tihnk being too many variables on some pages 19:34:23 <bag> perhaps as we get more of the caching done - then we could retry that? see if it’s still such a big hit? 19:34:25 <cait> adding up to a bigger performance issue 19:34:29 <pianohacker> right, but a lot of those are completely unavoidable (without major rearchitecting) 19:35:09 <pianohacker> bag: and the caching we've been adding wouldn't really affect that , unfortunately. 19:35:16 <bag> boo 19:35:39 <pianohacker> rangi: would marking more things as pre-escaped help any, do you think? 19:35:48 <rangi> probably not 19:36:17 <rangi> caching template fragments would though 19:36:32 <rangi> 2 secs 19:38:03 <rangi> https://perldoc.koha-community.org/Koha/Template/Plugin/Cache.html 19:39:35 <rangi> probably would need to reorganise some pages, but I think there are big chunks that don't need to be rerendered every page 19:40:01 <rangi> does that make sense? 19:40:18 <cait> i think one of he problmatic pages was authority import? 19:40:29 <cait> just trying to remember, it's probably on the bug 19:40:52 <jajm> will it prevent template variables to be escaped ? 19:41:34 <rangi> sorry I dont understand the question jajm 19:42:34 <jajm> template variables that are used only in cached template fragments 19:42:42 <rangi> no 19:42:48 <jajm> i thought all variables were automatically escaped 19:42:51 <rangi> no 19:42:59 <rangi> thats what this bug tried to do 19:43:03 <rangi> but it was too slow 19:43:23 <jajm> ok 19:43:25 <rangi> at the moment we manually escape them, and people forget to do it, and we get xss vulnerabilities 19:44:46 <rangi> what caching bits of the template means, is they get evaluated once, the same way as everything does now, and then put in a cache, and just returned from that, until it is stale. if you cache the borrowernumber as part of the key you could easily cache the top part of pretty much all the opac etc 19:45:16 <rangi> anyway, something to experiment with maybe 19:45:20 <pianohacker> yeah 19:45:31 <pianohacker> do we know why the variable escaping itself was slow? 19:45:36 <jajm> rangi, i am not sure to understand, you suggest to use template::stash::autoescaping + caching, or just caching ? 19:45:47 <pianohacker> because that's a really common practice in modern html template libraries 19:45:50 <rangi> both 19:46:04 <rangi> pianohacker: exactly, so i feel like it's something we can fix 19:46:07 <pianohacker> and, well, it's a friggin regex, perl can handle those :P 19:46:14 <rangi> it's written up on the bug i think 19:46:53 <rangi> jajm: both, i think the caching will help anyway, and I think we need to autoescape too 19:47:05 <cait> ithink more information to be found at the bug 19:47:24 <rangi> pianohacker: it may be we just looped too much, not sure 19:47:26 <cait> are we ok to move on for now? something that i should log? 19:47:45 <pianohacker> cait: just one thing 19:47:52 <pianohacker> rangi: do you have that bug number? 19:48:18 <rangi> bug 13618 19:48:19 <huginn> 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=13618 enhancement, P5 - low, ---, jonathan.druart, BLOCKED , Prevent XSS in the Staff Client and the OPAC 19:48:27 <pianohacker> thanks 19:48:27 <cait> pianohacker: it's on the agneda ;) 19:48:39 <pianohacker> I totally missed that :) 19:48:52 <cait> i'd like to talk about benchmarking first if that's ok? 19:48:59 <cait> or next 19:49:05 <cait> as we have seen a lot of activity there 19:49:18 <barton> #link https://perldoc.koha-community.org/Koha/Template/Plugin/Cache.htm 19:49:27 <pianohacker> cait: yeah, definitely 19:49:33 <cait> i think the bug linked on the agenda is not right 19:49:42 <cait> probably me getting the sequence wrong again 19:50:05 <cait> bug 15342 19:50:06 <huginn> 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=15342 enhancement, P5 - low, ---, jonathan.druart, NEW , Performance 3.22 - Omnibus 19:50:22 <cait> pianohacker: could you maybe say something about the work so far? 19:50:57 <cait> link on the agenda is fixed now 19:51:16 <pianohacker> cait: reviewing all the linked bugs real quick 19:51:24 <cait> bug 11998 is the pushed one 19:51:25 <huginn> 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=11998 major, P5 - low, ---, jonathan.druart, Pushed to Master , Syspref caching issues 19:51:30 <cait> ... which broke my opac 19:51:40 <cait> and yes, that's a complaint ;) 19:51:46 <pianohacker> which is... less than good, but we'll fix it 19:51:55 <cait> hope so :) 19:52:17 <pianohacker> bug 15350 is a big pain point, though it's less of a big deal for Plack 19:52:18 <huginn> 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=15350 enhancement, P5 - low, ---, jonathan.druart, In Discussion , DBIx::Class Startup speed 19:52:52 <bag> yes 19:53:10 <pianohacker> any suggestions there are welcome; I had a basic POC that only really helps with Plack (and even then only with the first load on a new worker) 19:54:46 <pianohacker> finally, I've been working with Joubu on a very promising set of patches 19:54:49 <pianohacker> bug 16044? 19:54:50 <huginn> 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=16044 enhancement, P1 - high, ---, jonathan.druart, Needs Signoff , Define a L1 cache for all objects set in cache 19:54:53 <pianohacker> yes, okay 19:54:53 <rangi> excellent 19:55:16 <pianohacker> this is intended to bypass the unexpectedly high cost of deserializing stuff from memcached 19:55:31 <pianohacker> Joubu added what amounts to a per-process in-memory cached 19:56:30 <pianohacker> I added some code to a) clear this new L1 cache less often, by storing a modification time in memcached and b) avoid some of the object-cloning protections Joubu/Jacek have put in place 19:57:04 <rangi> cool 19:57:10 <pianohacker> any technical comments on that bug would be very very welcome; I know that the approach we've made is _spectacular_ for performance, especially search, but we need to make sure it's safe 19:57:19 <bag> could be good to bring up on GBSD 19:57:23 <pianohacker> cait: that's all I've got 19:57:59 <cait> that's not bad - i am very glad to see work on the performance issues 19:58:09 <cait> :) 19:58:12 <bag> 61 days until release - that would be good to get in quickly so we have time to test 19:58:18 <pianohacker> agreed 19:58:19 <cait> true 19:58:31 <cait> pianohacker: could you add anything you need testers for to the gbsd page? 19:58:37 <pianohacker> cait: yup yup 19:58:40 <cait> thx 19:58:48 <rangi> pianohacker: yeah im happy to spend time testing that on gbsd 19:58:49 <bag> ES too - everyone is really excited for it - but need sign off ;) 19:58:57 <cait> ES is on it already :) 19:59:12 <bag> yuppers - just mentioning again ;) 19:59:23 <cait> ok, last topic would be item-level_itype 19:59:29 <cait> do we want to discuss or postpone? 19:59:30 <pianohacker> rangi: cool. If you have any ideas on other ways to do the memcached-modification-time thing (second to last patch) I'd love to hear them 19:59:42 <rangi> righto 19:59:45 <bag> I think we need Joubu here to talk about that cait 19:59:45 <rangi> ill take a look 19:59:49 <pianohacker> cait: we don't have much of a quorum for that touchy of a subject 19:59:53 <pianohacker> rangi: thanks# 19:59:55 <pianohacker> ! 19:59:57 <cait> so postpone 20:00:05 <cait> we can move it to top for the next meeting 20:00:09 <pianohacker> yup 20:00:17 <cait> #topic Updates from the QA team 20:00:33 <cait> nothing much from me I am afraid - I have been quite busy but hope to do more this week and probably next 20:00:44 <pianohacker> it may also be worth making a hot-topics list on the agenda that bag could copy-paste into his reminder email and perhaps encourage more people to come 20:00:44 <bag> cool 20:01:07 <cait> ah, maybe it would be good to talk about freeze dates sometime soon 20:01:29 <bag> good point - I’ll work something up this week cait 20:01:45 <cait> #action bag to set freeze dates 20:01:52 <cait> looking at the list of horrible bugs 20:02:00 <cait> there is quite a lot of red at the moment 20:02:08 <cait> also unhappy jenkins 20:02:37 <pianohacker> cait: are you looking at the "FIXME NOW" list? 20:02:41 <cait> 21 bugs critical - blocker 20:02:50 <cait> pianohacker: looking at the one linked on the agenda :) 20:03:00 <cait> #link https://bugs.koha-community.org/bugzilla3/buglist.cgi?cmdtype=dorem&remaction=run&namedcmd=FIXME%20NOW&sharer_id=1 20:03:08 <pianohacker> yup, same one, cool 20:03:17 <cait> yep, didn't see the name first 20:03:33 <cait> 21 is not so nice... should try to bring that way down before release 20:03:44 <cait> some maybe can be recategorized, but needs some work to check 20:03:45 <rangi> and get jenkins green 20:03:59 <pianohacker> tcohen: any comments on the current status of jenkins? 20:04:06 <bag> I can get on 15585 today 20:04:07 <cait> quite a few are LDAP or SIP 20:04:12 <cait> anyone able to test this, please do 20:04:28 <tcohen> i tried providing patches for several failures 20:04:43 <tcohen> but we really need to encourage the devs to fix them 20:04:57 <tcohen> probably with a warning about patches reverted 20:05:06 <cait> #idea add hot topics list to the agenda for use in the reminder email 20:05:08 <wizzyrea> one of those is a 3.8 20:05:28 <cait> fix or revert threat? 20:05:31 <bag> hmm tcohen can you talk with me on GBSD about those patches? I can do the leg work on getting some comments on those bugs 20:05:42 <tcohen> of course! 20:05:51 <bag> awesome! thanks 20:06:09 <pianohacker> wizzyrea: bug 9626? 20:06:10 <huginn> 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=9626 critical, P5 - low, ---, koha-bugs, NEW , can't pay FU fines without messing things up 20:06:20 <wizzyrea> bug 14003 20:06:21 <huginn> 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=14003 critical, P1 - high, ---, koha-bugs, In Discussion , SIP2 AF Field return 1 20:06:24 <cait> i think that's on mster too - waiting for another test from me or someone else 20:06:32 <cait> i need to verify some weird behaviour i saw is real or not 20:06:36 <cait> for the FU one 20:07:30 <cait> nothing more from me for now :) 20:08:47 <cait> #action bag and tcohen to talk about fixing jenkins 20:08:52 <cait> moving on? 20:08:57 <tcohen> yeah 20:09:02 <cait> #topic Set date for next meeting 20:09:08 <cait> when shall we meet again? 20:10:01 <bag> April 5th 20:10:02 <cait> or asked differently, which rhythm 20:10:13 <cait> next general meeting is on 6th 20:10:25 <cait> so maybe too close 20:10:28 <bag> 12th then? 20:10:42 <cait> +1 20:11:29 <bag> what time? 20:11:29 <wahanui> time is like a big ball of wibbly wobbly... time-y wimey... stuff. or generously donated to giving support in #koha but if you need more help faster than we can give it try http://koha-community.org/support/paid/ 20:12:21 <cait> if we use the one from before this one 20:12:24 <cait> it would be 15 UTC 20:12:40 <bag> ok 12 April 2016 at 15 UTC 20:12:57 <drojf> there is the catmandu hackathon. but i think its only half the day 20:13:14 <cait> oh, where is that? 20:13:17 <drojf> berlin 20:13:19 <drojf> stabi 20:13:39 <drojf> johann is organising. 11.+12.04 20:13:43 <cait> ah ok 20:13:47 <cait> 13th? 20:14:02 <cait> 13th, 15 UTC? 20:14:25 <cait> please saysomething :) 20:14:29 <drojf> woud work for me 20:14:31 <drojf> would 20:14:34 <cait> bag? 20:14:34 <wahanui> I LIKE BASEBALL! 20:14:44 <drojf> everyone else? :) 20:14:49 <bag> I LIKE BASEBALL 20:14:52 <bag> works for me 20:14:54 <bag> 13th 20:15:25 <cait> #agreed Next meeting will be April 13th, 15 UTC 20:15:26 <cait> #endmeeting