IRC log for #koha, 2020-09-11

All times shown according to UTC.

Time S Nick Message
00:20 inlibro joined #koha
00:43 wizzycray joined #koha
01:20 inlibro joined #koha
02:02 koha-jenkins Project Koha_20.05_D11 build #31: STILL UNSTABLE in 45 min: https://jenkins.koha-community[…]oha_20.05_D11/31/
02:04 koha-jenkins Yippee, build fixed!
02:04 wahanui Congratulations!
02:04 koha-jenkins Project Koha_20.05_U20 build #72: FIXED in 53 min: https://jenkins.koha-community[…]oha_20.05_U20/72/
02:13 koha-jenkins Project Koha_20.05_D11 build #32: FAILURE in 2 min 59 sec: https://jenkins.koha-community[…]oha_20.05_D11/32/
02:20 inlibro joined #koha
03:20 inlibro joined #koha
03:35 kathryn joined #koha
04:01 alexbuckley joined #koha
04:06 koha-jenkins Project Koha_20.05_D11 build #33: STILL FAILING in 2 min 15 sec: https://jenkins.koha-community[…]oha_20.05_D11/33/
04:21 inlibro joined #koha
04:51 koha-jenkins Project Koha_20.05_D11 build #34: NOW UNSTABLE in 44 min: https://jenkins.koha-community[…]oha_20.05_D11/34/
05:08 chriss joined #koha
05:21 inlibro joined #koha
05:42 lmstrand joined #koha
05:44 cait joined #koha
05:48 did joined #koha
06:12 marcelr joined #koha
06:12 marcelr hi #koha
06:21 * cait waves
06:21 cait marcelr: i didn't want us to forget about it after Joubu pushed - no time this week to patch it myself (yml files)
06:21 inlibro joined #koha
06:21 marcelr hi cait++
06:22 marcelr yeah i do it today
06:24 cait thx :)
06:27 cait ... office day today - bbl
06:27 alex_a joined #koha
06:30 alex_a Bonjour
06:38 ashimema[m] ?
06:42 marcelr hi alex_a and ashimema[m]
06:43 magnuse \o/
06:48 ashimema[m] ils-di is also an option Patrick.. but I'd say SIP2 is the most well used/well supported.
06:49 ashimema[m] we should do OAuth2 really
06:49 ashimema[m] but don't yet support that
06:50 * ashimema[m] wishes they're chosen more different names for OATH and OAuth.. they're too easy to muddle up
06:58 lds joined #koha
07:00 JesseM joined #koha
07:06 fridolin joined #koha
07:13 cait joined #koha
07:16 cait back
07:21 inlibro joined #koha
07:57 huginn News from kohagit: Bug 26384: (QA follow-up) Fix two exceptions <[…]d5163f13e2d4c39ee>
07:57 huginn News from kohagit: Bug 26007: (QA follow-up) Fix English yml file <[…]0222ad0af6057a7f0>
07:57 huginn News from kohagit: Bug 26384: Fix executable flags <[…]a10ee9a4a7137c6f4>
07:57 huginn News from kohagit: Bug 26384: Add test to catch for missing or extra execution flags <[…]f381d2ca3be28e1e8>
07:57 huginn News from kohagit: Bug 26401: Remove xt/fix-old-fsf-address* <[…]aad25b1b7e1fd2a51>
07:57 huginn News from kohagit: Bug 26394: Update .mailmap <[…]d4bab83955a3c8a18>
07:57 huginn News from kohagit: Bug 26365: Remove comparison with 0000-00-00 for datecancellationprinted <[…]6c08cefc02e788acc>
08:11 * ashimema[m] roles up sleaves ready to dive into the mojolicious app patch.. it looks very self contained so I think I'm happy with it as a stepping stone.. I'd love to see it as a slow move towards no longer emulating CGI, but I think that's a long way off :(
08:21 Joubu ashimema[m]: yes, please! :)
08:21 ashimema[m] It's been on my list for like a year.. finally getting through the list!
08:21 ashimema[m] woop, woop
08:21 inlibro joined #koha
08:33 marcelr what is this:   sub ModZebrations {     return _common_config($_[1],'serverinfo'); }   ???
08:33 marcelr in C4/Context
08:34 marcelr seems to be unused ?
08:36 marcelr the ElasticSearch code in Koha/SearchEngine does not really make me look forward to moving away from Zebra yet..
08:40 koha-jenkins Project Koha_Master_D9 build #1427: SUCCESS in 41 min: https://jenkins.koha-community[…]a_Master_D9/1427/
08:43 koha-jenkins Project Koha_Master_U20 build #91: SUCCESS in 44 min: https://jenkins.koha-community[…]ha_Master_U20/91/
08:43 koha-jenkins Project Koha_Master_D9_MDB_Latest build #386: SUCCESS in 45 min: https://jenkins.koha-community[…]9_MDB_Latest/386/
08:57 koha-jenkins Project Koha_Master_U18 build #890: SUCCESS in 56 min: https://jenkins.koha-community[…]a_Master_U18/890/
08:57 ashimema[m] am I reading this right.. so te Mojo patch is bascially running  Mojo app that's emulating CGI undef in psgi mode.. LOL
08:57 ashimema[m] kinda makes sense considering what we're trying to achieve.. but such layers of of obfuscation do scare me
08:58 marcelr yeah ashimema[m]  but we already do such things too for Plack and CGI
08:58 marcelr real PSGI is still far
09:07 ashimema[m] indeed
09:07 ashimema[m] layer upon layer of emulation
09:09 ashimema[m] I'm kinda trying to understand why we'd want to run via mojo though now..
09:10 ashimema[m] so.. at the moment a fairly standard configuration is two plack apps running.. right.. one with opac (and opac api) and a second with intranet (and intranet api).. both basically running compiled cgi apps.
09:11 ashimema[m] with the mojo version we'd still be running two plack apps, but this time it's a plack app running a mojo app running a compiled cgi app
09:11 ashimema[m] does it actually help us move towards writing routes directly in mojo or anything like that..
09:11 ashimema[m] think those questions need to go to jajm.. I'm struggling with the high level vision
09:12 Joubu I will be mostly afk for the next 2 weeks. Bye #koha :)
09:12 ashimema[m] have a great break Joubu
09:13 did joined #koha
09:15 marcelr bye Joubu
09:16 marcelr ashimema[m]: at some point we need to make a choice or well never pass the crossroad
09:17 koha-jenkins Project Koha_Master_D10 build #347: SUCCESS in 37 min: https://jenkins.koha-community[…]a_Master_D10/347/
09:20 jajm ashimema[m], you can also run directly the mojo apps without plack/starman, but that's a detail really. I think that if we want to drop CGI emulation, it will be easier with a framework that provides everything needed, and we already use mojo for the api
09:22 inlibro joined #koha
09:27 koha-jenkins Project Koha_Master_D11 build #83: SUCCESS in 43 min: https://jenkins.koha-community[…]ha_Master_D11/83/
09:28 koha-jenkins Project Koha_Master_D10_Deps build #71: SUCCESS in 45 min: https://jenkins.koha-community[…]ster_D10_Deps/71/
09:41 cait oh. now i missed Joubu :(
09:43 ashimema[m] sounds sane jajm
09:43 ashimema[m] DO you have any form a plan in your head of what enxt steps after this might be?
09:44 marcelr thx jajm++
09:44 ashimema[m] i.e have you investigated any form of migrations paths from emulated cgi controllers to actual mojo controllers?
09:44 ashimema[m] I'm wondering how that might look...
09:45 cait marcelr: you are supposed still to run zebra and elastic in parallel
09:45 cait as default
09:45 cait that's why it inserts into zebraqueue
09:45 jajm ashimema[m], kind of... the big issue will be making authentication work within mojo controllers
09:45 ashimema[m] We recommend against running under cgi at the moment don't we (I still run my dev env under it usually, but that's just habbit).. I bet there are some out there still running that way
09:45 ashimema[m] our auth is horrible as it stands
09:45 ashimema[m] hmm
09:45 cait there is a feature now that allows for z39.50 etc. but it's still pretty new
09:46 marcelr ok cait i forgot about that
09:46 ashimema[m] for ES running we're not going to run Zebra at all at ptfs-e
09:46 marcelr it is a shame that we still do btw
09:46 ashimema[m] our aim is to use the z3950 broker that talks to ES
09:46 ashimema[m] there should cetanily be an option to fully disable zebra including the queue
09:48 ashimema[m]
09:50 jajm ashimema[m], once we have authentication working, we can rewrite cgi scripts into mojo controller one by one, but that would mean making use of mojo app mandatory (otherwise we would have to maintain both the cgi script and the mojo controller)
09:51 ashimema[m] mmm, indeed.
09:51 koha-jenkins Project Koha_Master_U16 build #63: SUCCESS in 54 min: https://jenkins.koha-community[…]ha_Master_U16/63/
09:52 ashimema[m] though.. you can run mojo as a plack app (as we do here I think).. so for those already running under plack it could be seamless, at least to begin with
09:52 ashimema[m] can't remember if you can run mojo under cgi
09:54 ashimema[m] hmm.. you can indeed run mojo as a cgi app
09:54 ashimema[m] not at all recommended
09:55 ashimema[m] but possible
09:55 * ashimema[m] wrote a plugable Authentication and Authorization module for Mojo once already.. I bet I could port it to Koha to replace when running under Mojo
09:57 jajm ashimema[m], do you have a link ?
09:57 ashimema[m] it's three year old code
09:57 ashimema[m] but it's here:[…]
09:58 jajm it looks like it is a private repo : i get a 404
09:59 ashimema[m] ack
09:59 ashimema[m] for the authorization side I contributed to[…]in::Authorization
10:00 ashimema[m] and the Authentication side was along the same lines as[…]n::Authentication
10:01 ashimema[m] We supported JWT, Cookie, Shibboleth, LDAP and OAuth2
10:01 ashimema[m] oh.. and http basic
10:01 koha-jenkins Yippee, build fixed!
10:01 wahanui Congratulations!
10:01 koha-jenkins Project Koha_Master_D9_My8 build #409: FIXED in 43 min: https://jenkins.koha-community[…]aster_D9_My8/409/
10:02 ashimema[m] I'll dig around a bit.. clean it up a little and submit something..
10:02 ashimema[m] but first.. I'll finish QAing your Mojo patch.. it's so self contained that it's a no brainer to me
10:22 kohaputti joined #koha
10:22 inlibro joined #koha
10:37 oleonard joined #koha
10:45 oleonard Hi #koha
10:53 tcohen morning
10:56 ashimema[m] PQA jajm.. wish I'd looked at this in detail earlier.. it's so self contained it's a no brainer to get it in... it's the next step of getting Auth working to allow us to write mojo centric controllers which is going to be more invasive/difficult.
11:02 tcohen hola jajm ashimema[m]
11:02 tcohen I'm really excited to see that dev in
11:02 ashimema[m] hi tcohen
11:03 ashimema[m] It's a great stepping stone.. but I fear the next stone will be more of a boulder to manipulate into place ;)
11:03 ashimema[m] but.. we can get there I'm sure
11:03 ashimema[m] and it gives us an excuse to take a fresh approach to Authentication/Authorization code
11:05 ashimema[m] cait around?
11:05 ashimema[m] I'm trying to understand your comment on bug 26268
11:05 huginn Bug[…]_bug.cgi?id=26268 enhancement, P5 - low, ---, martin.renvoize, Failed QA , Remove items.paidfor once and for all
11:06 ashimema[m] My query is.. how would you actually get data into the db field at all and where would it be displayed?
11:13 khall joined #koha
11:22 inlibro joined #koha
11:27 cait kind of around
11:27 cait if you had mapped it to 952$x it just stored the internal note
11:28 cait displaying in your item information as is now itemnotes_nonpublic
11:28 cait it was like that for ages
11:28 cait we didn't have items.itemnotes_nonpublic and at some point someone mapped the default to paidfor
11:29 cait i am not sure why this happened, but it was standard for quite a qhile
11:29 cait and we couldn't easily fix that, it was a manual process we did for our libraires, i am pretty sure there are others out there
11:29 cait who haven't figured that out
11:32 ashimema[m] so can we not migrate the data in paidfor to itemsnotes_nonpublic?
11:32 cait 13023
11:32 ashimema[m] if that's the case
11:32 cait basically.... people can have used it for anything
11:32 ashimema[m] bug 13032
11:32 huginn Bug[…]_bug.cgi?id=13032 enhancement, P5 - low, ---, koha-bugs, NEW , Command line switch to add page breaks to sent to the library
11:32 cait hm typo
11:32 cait bug 13023
11:32 huginn Bug[…]_bug.cgi?id=13023 enhancement, P5 - low, ---, kyle, CLOSED FIXED, Nonpublic note for items should be mapped to a database column like itemnotes
11:33 cait I think all we need to do is check if there is content and leave it be if it is
11:33 cait print an ote
11:33 cait note
11:33 cait people might have used it for anything - because you can map whatever you want to it
11:33 cait i know that at least in UNIMARC remapping the fields was quite common
11:33 cait and to make it worse.. for a while we shipped a now "wrong" mapping
11:34 ashimema[m] but do I need to prevent the removal of the paidfor field if there's a mapping found and there's data in the field then
11:34 ashimema[m] ok.. so for most people it would be removed
11:34 ashimema[m] basically.. I want to strongly discourage it existing.. else I'll find myself doing this for the fourth time and I don't want to do that same work yet again ;)
11:35 cait is there any issues to be expected with an extra unused column in old dbs?
11:35 cait we have plenty of those i think
11:35 ashimema[m] not that I can think of
11:36 cait i mean we should try to drop it and checking for mappings and actual data is the best way i think
11:36 cait but if there is doubt... better not to delete people's data
11:38 jajm ashimema[m], thanks for the QA, i'll start thinking about how we could best add authentication/authorization, let me know if you start working on it
11:40 jajm ashimema[m], also thanks for the followup
11:40 ashimema[m] no worries
11:40 ashimema[m] sorry it's taken me this long to get to it in my queue
11:40 ashimema[m] forever a loosing battle on that front
11:43 cait aren't we all? :(
11:43 cait there is so many things I'd really like to look at, but ....
11:54 ashimema[m] bug 26268 back in the QA queue
11:54 huginn Bug[…]_bug.cgi?id=26268 enhancement, P5 - low, ---, martin.renvoize, Signed Off , Remove items.paidfor once and for all
11:54 * oleonard always worries that "once and for all" tempts fate a little too much
11:54 ashimema[m] lol
11:55 ashimema[m] hmmm
11:55 ashimema[m] to dig into bug 24975 or not to dig into bug 24975... that is the question
11:55 huginn Bug[…]_bug.cgi?id=24975 enhancement, P5 - low, ---, julian.maurice, In Discussion , Refactor database translations
11:55 ashimema[m] seems it's your day today jajm :)
11:57 jajm great :) i really need some help on that one
11:57 * ashimema[m] is currently rebasing it
11:57 ashimema[m] then will have more of a go of wrapping my head around it
11:58 ashimema[m] why are translations so hard aye
12:03 Dyrcona joined #koha
12:16 cait ashimema[m]:  at lleast you can talk in your language ;)
12:22 inlibro joined #koha
12:46 khall_ joined #koha
13:01 Eileen joined #koha
13:01 Eileen My supervisor wants me to generate a report of all items I've cataloged since I started here a year ago. How do I do that?
13:02 TimothyAlexis joined #koha
13:02 TimothyAlexis We are upgrading from 19.05 to 20.05
13:02 TimothyAlexis "You are missing the <lockdir> entry in your koha-conf.xml file. Please add it, pointing to your Koha instance's lock dir."
13:03 TimothyAlexis "You are missing the <tmp_path> entry in your koha-conf.xml file. Please add it, pointing to the configured temporary directory for your Koha instance. The effective temporary directory is '/tmp'."
13:03 Eileen I don't understand.
13:03 Eileen I don't know what you're talking about.
13:03 TimothyAlexis This is from the /cgi-bin/koha/
13:03 janPasi_ joined #koha
13:04 Eileen I don't have much tech background.
13:04 TimothyAlexis It seems to run fine without.
13:04 Eileen I just want to know how to generate a specific report.
13:12 tcohen TimothyAlexis you should set those
13:12 tcohen tmp_path can be problematic
13:13 tcohen depending on how is systemd configured to handle tmp dir
13:14 TimothyAlexis tcohen Would 19.05 complain if these were present in koha-conf.xml?
13:16 tcohen no
13:18 TimothyAlexis so, can we just create 2 folders and use those?
13:22 TimothyAlexis <lockdir>/var/lock/koha/__KOHASITE__</lockdir>
13:22 inlibro joined #koha
13:23 TimothyAlexis Can we use <tmp_path>/var/tmp</tmp_path> ?
13:27 TimothyAlexis We're running a clustered setup for Koha with 2 servers. So we're wondering...
13:27 TimothyAlexis Do you have any documentation on what the tmp and lock dirs are actually used for?  We may or may not run into problems with koha-1 and koha-2 having separate tmp/lock dirs (since they're separate machines).
13:30 tcohen systemd put some constraints on the required permissions for services using tmp dirs. So they need to be owned by the running processes
13:30 tcohen so the idea is you have a per-instance lockdir and tmp dir
13:31 tcohen the lockdir was introduced by a patchset that added locking capabilities to scripts
13:31 tcohen mainly so no concurrent runs happen
13:31 tcohen we needed a lockfile and it made sense to generate it in a configurable place
13:32 tcohen on hte other hand, if a config entry doesn't exist in a Koha version, it means it is not used in the codebase
13:32 tcohen and then adding to the koha-conf.xml file shouldn't make any harm
13:32 TimothyAlexis OK
13:33 TimothyAlexis we're not using the instance mechanism.
13:34 TimothyAlexis So we could use /var/lock/koha and I suppose /tmp.
13:35 TimothyAlexis Is the only thing using lockdir?
13:41 tcohen TimothyAlexis I made a mistake, it was :-d
13:42 tcohen and /tmp is the main problem
13:42 tcohen you should point to /var/tmp/koha
13:42 tcohen and make sure the dir belongs to the user running the service
13:42 tcohen that's it
13:42 TimothyAlexis We're not using privatetmpdir
13:42 TimothyAlexis It's turned off
13:43 wizzycray joined #koha
13:43 TimothyAlexis We'll set it to /tmp/koha :)
13:44 ivandz joined #koha
13:44 TimothyAlexis Are <lockdir> and <tmp_path> top level nodes or do they need to be inside some other tag?
13:46 TimothyAlexis We're assuming, under <config>
13:47 ivandz Hi all. Maybe somebody know where is a file mapping for ElasticSearch ? (which path/folder) Thanks!
13:48 ivandz Hi all. Maybe somebody knows where is a file mapping for ElasticSearch ? (which path/folder) Thanks!
13:49 TimothyAlexis Thank you tcohen for the help! :)
13:49 tcohen TimothyAlexis top level under <config>
13:54 TimothyAlexis We couldn't have figured this out without you, so again, thanks.
13:58 kidclamp joined #koha
14:02 tcohen TimothyAlexis cool, will be around a few more hours if you need smth else
14:09 lukeG joined #koha
14:22 inlibro joined #koha
14:23 fridolin strangest bug ever OO Bug 26434
14:23 huginn Bug[…]_bug.cgi?id=26434 major, P5 - low, ---, fridolin.somers, Needs Signoff , Plugin dirs duplicates in @INC with plack
14:23 oleonard jajm++
14:25 fridolin ivandz: should be in admin/searchengine/elasticsearch/mappings.yaml
14:25 fridolin unless defined in koha-conf.xml
14:28 tcohen jajm about to submit the patch for Makefile.PL (SMTP servers)
14:28 tcohen I found some things don't work as expected with Makefile.PL
14:28 ivandz fridolin, Thanks!
14:42 jajm tcohen, thanks for the patch. what doesn't work with Makefile.PL ? you will submit a new bug report ?
14:42 tcohen I will
14:42 tcohen I had hard time finding out how to make the --smtp-debug option switch work as I suspected it should
14:43 tcohen nad the problem was I was doing the same --use-memcached does
14:43 tcohen its a 1 vs 'yes' thing
14:43 tcohen I don't think people uses option swithces
14:58 lukeG oleonard++
14:58 lukeG thank you for Bug 26431, I love it
14:58 huginn Bug[…]_bug.cgi?id=26431 enhancement, P5 - low, ---, oleonard, Signed Off , Use split button to offer choice of WYSIWYG or code editor for news
15:08 margaret joined #koha
15:12 cait lukeG: still around?
15:13 fridolin left #koha
15:16 lukeG cait yes
15:17 cait ah
15:17 cait can you please reconsider bug 25129
15:17 huginn Bug[…]_bug.cgi?id=25129 enhancement, P5 - low, ---, katrin.fischer, Pushed to master , Update German (de-DE) web installer files for 20.05
15:17 cait it was especially written for 20.05
15:17 cait similar for bug 25448
15:17 huginn Bug[…]_bug.cgi?id=25448 normal, P5 - low, ---, katrin.fischer, Pushed to master , Update German (de-DE) framework files
15:18 cait the first makes sure someone installing 20.05 has translated instalelr files
15:18 lukeG I will defenitly backport it!
15:18 cait and the other brings the german frameworks up to date with the english ones again
15:18 cait i just did them too late and then they got lost int he SO queue sadly
15:18 cait thx :)
15:19 cait i didn't mark them as a bug... but they could be - it's a bit of a definition question i guess. Things are not broken... but incomplete/untranslated
15:19 cait so I guess they can be bugs
15:19 lukeG sorry i skipped it!
15:20 cait all good, they were marked enh - i should just file them as bug in future :)
15:20 cait time for the weekend - cya later all :)
15:20 cait left #koha
15:23 inlibro joined #koha
16:05 tuxayo «i didn't mark them as a bug... but they could be - it's a bit of a definition question i guess. Things are not broken... but incomplete/untranslated»
16:05 tuxayo Could as bug IMHO
16:23 inlibro joined #koha
16:45 cait joined #koha
17:23 inlibro joined #koha
17:48 lisettelatah joined #koha
17:49 oleonard Is it just me or is Bugzilla running really really slow?
17:55 kidclamp you are not alone oleonard
17:55 oleonard Not with you by my side kidclamp!
17:56 kidclamp Clamp and the Leopard! (coming soon to Fox)
18:09 * oleonard spends the rest of his work day dream-casting Clamp and the Leopard
18:09 kidclamp Chris Hemsworth and George Clooney as us?
18:12 oleonard Wow Clooney and I are exactly the same height
18:12 oleonard Perfect.
18:12 wahanui perfect is, like, the enemy... (of what I forgot)
18:12 oleonard I picture him as taller than me.
18:23 inlibro joined #koha
19:23 inlibro joined #koha
20:23 inlibro joined #koha
20:48 janPasi joined #koha
21:18 khall joined #koha
21:21 irma_ joined #koha
21:24 inlibro joined #koha
21:26 dj joined #koha
22:24 inlibro joined #koha
23:24 inlibro joined #koha

| Channels | #koha index | Today | | Search | Google Search | Plain-Text | plain, newest first | summary