IRC log for #koha, 2022-09-02

All times shown according to UTC.

Time S Nick Message
00:07 tuxayo > the paint is still wet for ES7
00:07 tuxayo mtj: Do we need to have someone use it a bit in production or is that a small enough upgrade that the current automated tests are enough?
00:08 tuxayo By the way, we should use the actual version number of OS, the equivalence with ES. Because that only work with OS 1.x = ES 7.x
00:08 tuxayo https://github.com/opensearch-[…]ebsite/issues/767
00:12 tuxayo Good, 1.x is still maintained https://hub.docker.com/r/opens[…]t/opensearch/tags
00:16 tuxayo I can submit a patch for our koha-elasticsearch-icu for OS 2.x, how should the branch be named? opensearch-2.x ? And I can also submit an update for 1.x under the new convention.
00:17 mtj tuxayo: its a good idea, i think ^
00:17 tuxayo ok :)
00:18 mtj we can use this repo https://gitlab.com/koha-commun[…]elasticsearch-icu
00:19 tuxayo > OS7 passing just 3 weeks ago
00:19 tuxayo That's a good news, I though ES7 passed but OS1.x didn't so it wouldn't be a simple as hoped. But it works :D
00:19 mtj ...or create a new one?  koha-opensearch-icu
00:19 tuxayo It's already the one used for our OS1.x
00:20 tuxayo And the name is still ok. We might just want to change the description to say it's about ElasticSearch implementations.
00:21 mtj agree ^
00:22 mtj tuxayo: i think we probably need to use ES/OS7 a bit more, before we say its production ready
00:24 tuxayo Ok so we need to find guinea pigs :)
00:24 mtj yes 🐷
00:27 mtj do a wiki page with a table showing koha/es compatibility
00:28 mtj and update the about.pl with ES info stuff
00:35 tuxayo > do a wiki page with a table showing koha/es compatibility
00:35 tuxayo mtj: it would be just a subset of https://wiki.koha-community.or[…]d_recommendations
00:35 tuxayo > and update the about.pl with ES info stuff
00:35 tuxayo What do you mean?
00:38 mtj i mean we could add some ES info to koha's about.pl page
00:39 mtj ...like ES/OS version, and if ES is running/active or not
00:41 mtj it would be helpful for troubleshooting
00:50 tuxayo > we could add some ES info, like ES/OS version, and if ES is running/active or not
00:50 tuxayo I though there was, there was a bug that added the number of index record and raised a warning if it was different from the number in the DB
00:53 tuxayo It was bug 27667, I forgot that it was just a warning and not an info always displaed
00:53 huginn` Bug https://bugs.koha-community.or[…]_bug.cgi?id=27667 enhancement, P5 - low, ---, fridolin.somers, Pushed to oldstable , Display the number of non-indexed records
00:53 tuxayo *displayed
02:23 bag joined #koha
05:23 MarkHofstetter joined #koha
05:59 marcelr joined #koha
06:00 marcelr hi #koha
06:31 alex_ joined #koha
06:44 MarkHofstetter1 joined #koha
06:54 MarkHofstetter joined #koha
07:36 MarkHofstetter joined #koha
08:14 marcelr stirring up the additional_contents.code discussion; bug 31509
08:14 huginn` Bug https://bugs.koha-community.or[…]_bug.cgi?id=31509 enhancement, P5 - low, ---, koha-bugs, NEW , Additional contents: Code should be configurable sometimes
08:18 paulderscheid[m] morning #koha
08:24 marcelr o/
09:08 marcelr bug 31503 needs a signoff :)
09:08 huginn` Bug https://bugs.koha-community.or[…]_bug.cgi?id=31503 enhancement, P5 - low, ---, m.de.rooy, Needs Signoff , Allow several consent types on the consents tab of OPAC account page
09:39 MarkHofstetter1 joined #koha
10:12 oleonard joined #koha
10:19 oleonard Hi #koha
11:10 tcohen hola #koha
11:15 tcohen mtj++
11:15 tcohen mtj++
11:21 tcohen tuxayo: have you solved the --es* param vs. .env problem?
11:49 * oleonard bbl
11:59 magnuse why would showing a single ill request take 4.5 minutes?
12:00 tcohen please enable Plack
12:00 tcohen :-P
12:01 tcohen sounds like network latency?
12:01 tcohen the illrequests?method=illview page is painfully slow but we are talking about 'seconds'
12:01 tcohen not minutes
12:02 tcohen I'm working on bug 22440 this morning
12:02 huginn` Bug https://bugs.koha-community.or[…]_bug.cgi?id=22440 normal, P4, ---, tomascohen, In Discussion , ILL API response objects inconsistent with other endpoints
12:02 tcohen it will end up being a properly API-based pagination implementation
12:02 tcohen so the amount of requests on the DB won't impact load time :-D
12:02 tcohen (above page size of course)
12:06 katuramu joined #koha
12:13 magnuse yeah, this is without plack, but even with plack it is so slow apache stops waiting and gives a 500 error
12:13 tcohen the current illview retrieves ALL requests
12:13 tcohen even if the filters would make it find one
12:14 tcohen I mean the datatables filters
12:14 magnuse other parts of the system are prett fast, even without plack
12:15 tcohen right, if you look at the controller it does a plain ->search and returns 'everything'
12:16 tcohen so even with plack it is slow on sites with more than (say) 1000 historical ILL requests
12:16 tcohen then it lets datatables do in-browser pagination of that massive list
12:16 tcohen more news about this later today :-D
12:16 magnuse this site has 107 illrequests
12:17 magnuse tcohen++
12:17 tcohen you can try using POstman to make the request yourself
12:17 tcohen to the API
12:17 tcohen and test performance
12:17 tcohen what version are you working on?
12:17 tcohen because I added a small caching trick recently (21.11 I think)
12:18 magnuse loading /cgi-bin/koha/ill/ill-requests.pl takes 1.7 seconds, then /api/v1/illrequests?embed=metadata,​patron,capabilities,library,status_​alias,comments,requested_partners takes 6,8 seconds
12:18 magnuse 21.11.05, but with some patches, so the problem might be there
12:23 oleonard joined #koha
12:30 tcohen magnuse: 6.8 secs is very different than 4.6 min. And also sounds like the 'normal ILL list speed'
12:30 magnuse yeah, the trouble is detail view - illrequests?method=illview
12:36 magnuse removed the patches, but detail view still takes 4.5 minutes. i have customers on aqbout this same version using ill in production though, so i wonder if there is something wrong with this instance
12:37 oleonard In case anyone missed it yesterday, I have a sandbox up with the staff interface redesign and my search bar updates: https://staff-bug30952.sandboxes.biblibre.eu
12:44 ashimema much better oleonard
12:49 magnuse could one of the "search boxes" be "open" by default?
12:57 magnuse oleonard++ biblibre++
13:06 magnuse oleonard: coleagues are pointing out areas that a re difficult to read, with thin green letters on a grey background. we are kind of wcag-aware at the moment :-)
13:24 oleonard magnuse: The open-by-default issue is on my list to resolve if folks are still happy with the way this is going
13:24 oleonard And I agree that the contrast is overall not adequate
13:25 Dyrcona joined #koha
13:28 magnuse i'd say it looks really good overall
13:36 magnuse tcohen: looks like my problem was that i had almost 1 million STATUS_CHANGE action logs connected to my ~100 requests. i'll investigate why on monday...
13:53 oleonard Haha I hope you were looking at the sandbox at the right time because my version was overwritten :D
14:06 oleonard I posted a link to my remote branch on the bug: https://gitlab.com/koha-dev/ko[…]2-search-redesign
14:08 ashimema I just spotted it had changed
14:22 tcohen joined #koha
14:27 tcohen oleonard-away: is there another sandbox with your branch?
15:00 * ashimema sets up another sandbox with Owens branch
15:00 tcohen \o/
15:00 tcohen ashimema++
15:00 tcohen oleonard++
15:01 * ashimema has been testing again and has closed some of the github issues and added one new one
15:02 ashimema I'm still very much in a quandary with the toolbar change though.. for me, removing the container makes all the buttons meld into the page too much.. I liked that the toolbar looked like a toolbar before personally
15:03 ashimema as a note.. I have half of ptfs-e working through this all next friday as a group
15:04 ashimema we'll likely look at both branches.. i.e. with and without oleonards follow-up stuff
15:06 oleonard Thanks ashimema
15:07 ashimema https://staff-bug30952o.sandbo[…]/koha/mainpage.pl
15:07 ashimema so now we can compare and contrast with ease 🙂
15:08 tcohen this is awesome
15:09 oleonard I'll see what I can do to solve the problem of the default form being active
15:09 ashimema nice one
15:10 ashimema I'm enabling CircSidebar, UseDesks and UseRegisters on them both btw.. so we get to see them 'crowded'..
15:11 ashimema a lot of our customers are using all those things so I feel they're worthwhile being enabled on testing 😜
15:13 tcohen +1
15:13 oleonard I've turned on IntranetCatalogSearchPulldown and IntranetAddMastheadLibraryPulldown so you can see how they have been incorporated into the catalog search form
15:13 oleonard Like the patron search, I added the "controls" icon as a trigger to display the extra fields
15:14 oleonard This may be a no-go, I'm not sure.
15:17 ashimema I LOVE that!
15:20 ashimema I do vaguely like the top bar being a bit deeper now I compare them side by side
15:20 ashimema my brain hurts
15:20 ashimema time to clock off for the weekend
15:20 ashimema have a great one everyon
15:20 ashimema s/everyon/everyone/
17:32 TGoat joined #koha
21:48 lukeg joined #koha

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