Time Nick Message 06:56 reiveune hello 07:09 magnuse \o/ 07:23 marcelr o/ 07:30 ashimema Another week of back to back meetings.. today looks a tiny bit clearer, hopefully will fit in some QA.. woop, woop 07:39 marcelr there will be not much to qa 07:39 marcelr bugs+strings 08:21 cait good morning #koha 08:21 cait guten Morgen MarkHofstetter :) 09:18 cait ok, quiet Friday... 09:18 PedroAmorim[m] o/ 09:24 Joubu ashimema: about the lang permission error, yes I thought about that but didn't fill it yet 09:24 Joubu we need to clone using the correct user, not root 09:25 ashimema I see 09:25 ashimema thanks 09:25 ashimema I got round it by escalating to root 09:25 ashimema but it did feel 'wrong' 😛 09:26 ashimema funny.. I almost never need to install languages but happened to be working on a bug in that area last night 09:26 Joubu https://gitlab.com/koha-community/koha-testing-docker/-/issues/405 09:40 MatthewBlenkinsop[m] <caroline> "are we in string freeze? I wasn..." <- 22.11 string freeze starts today, I'm sending out the email this morning :) 09:55 Joubu marcelr: Is the wiki page clear enough? 09:56 Joubu you are not supposed to not have the po directory, but I thought it was better to warn than die, see 588e834e7de8409a838906ce87f7a296893e1e23 09:56 Joubu I have added a warning to the about page as well, that point to the wiki page 10:01 * ashimema tried to wrap brain around new flow.. must admit I've not kept up here 10:01 ashimema just been impressed by it 10:01 ashimema the change looks sane to me.. but yeah.. I need to fully grok the flow 10:06 Joubu I've tried to communicate about it the better I could. But it something is still not clear please let me know. 10:08 cait Does things change for RMs? 10:08 cait just because MatthewBlenkinsop[m] just mentioned the string freeze 22.11 10:08 marcelr Joubu: i understand that the po dir is gone, but not all code seems to know that ; see your logs 10:10 Joubu marcelr: see my comment and the wiki page 10:10 Joubu https://wiki.koha-community.org/wiki/Translation_server_migration_weblate#Architecture and especyally the sync step 10:10 marcelr we probably need a translator :) 10:11 Joubu about the removal of .po, prod will have a new koha-l10n package that will contain the po 10:11 Joubu for devs, ktd will install them from koha-l10n git repo, automatcially 10:11 Joubu if you are not using ktd you need to clone koha-l10n into misc/translator/po 10:12 Joubu if you have not done it, you will have the warn in the logs + the alert on the about page 10:12 cait I mean more becuase theis will be hte first one with weblate - so no more Beranrdo form for pull request etc 10:12 Joubu cait: see my email to the release team 10:12 cait hm might have missed that ok 10:12 Joubu yes, less steps 10:12 Joubu nothing to do for the translatino files 10:12 Joubu nothing to do for the tarball 10:13 Joubu so the workflow change, but it removes steps 10:13 cait translation is huge this time aroudn :( I hope I will make it in time 10:13 cait but not weblate's fault, it's the changes to breadcrumbs etc and the ton of new modules we have 10:16 Joubu one thing I have not explicitely said, devs will work with the up-to-date strings (from weblate). You will translate a string today and see it tomorrow in your dev box 10:16 Joubu (not there yet, but almost) 10:20 marcelr ok this is just a case of non-ktd discrimination then, lol 10:21 marcelr i will add a mkdir po line in my dev Docker, haha 10:23 marcelr Joubu: there are some flatpickr issues in smart-rules, please see bug 35341 10:23 huginn` 04Bug https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=35341 minor, P5 - low, ---, matt.blenkinsop, Failed QA , Circulation rule dates are being overwritten 10:24 Joubu marcelr: how could I have done it better for you? 10:24 Joubu I should have sent an email to koha-devel for better communication 10:24 marcelr you are doing good 10:24 Joubu but running out of time yesterday 10:24 marcelr np 11:09 aude_c[m] Silly question... In bugzilla search results, is it possible to have the "Needs QA" and "Needs documentating" statuses display differently - so it's quick to see which one is which? Maybe there's just something I am doing wrong? 11:10 aude_c[m] I'm looking at https://bugs.koha-community.org/bugzilla3/buglist.cgi?quicksearch=cookieconsent The last line is Needs documenting and the two above Need signoff but I see them all as "Need" 11:11 ashimema I know what you mean 11:11 ashimema it's an annoyance with the column width 11:12 aude_c[m] ah, ok 11:12 ashimema we have control over what those status's say.. but changing them now would involved a chunk of related work to update everything else that uses them 11:12 aude_c[m] So not worth it :D 11:14 ashimema we tried to move away from maintaining a fork of bz.. though I do think we have some customisations already. 11:14 ashimema maybe submit a request in here: https://bugs.koha-community.org/bugzilla3/buglist.cgi?component=Bugzilla&list_id=484819&product=Project%20Infrastructure&resolution=--- 11:14 ashimema I think Joubu might have the access required to make such a change? 11:14 ashimema if not, rangi is the man 11:15 aude_c[m] thanks 12:00 cait we could also update the change and turn off the email sending for it 12:00 cait like we do after releases 12:01 cait actually.. we could maybe schedule it for when we do the status updates 12:01 cait but is there a good description that woudl fit in well? 12:10 cait tcohen[m] ping! 12:17 Annelisterman[m] Hi! I'm trying to test bug 31694 and install other languages to sandboxes but they are not installed. I get this message from every provider's sandboxes: readdir() attempted on invalid dirhandle $fh at /kohadevbox/koha/misc/translator/LangInstaller.pm line 71. 12:17 Annelisterman[m] closedir() attempted on invalid dirhandle $fh at /kohadevbox/koha/misc/translator/LangInstaller.pm line 72. 12:17 Annelisterman[m] Unsupported language: de-DE 12:17 huginn` 04Bug https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=31694 normal, P5 - low, ---, shi-yao.wang, Needs Signoff , MARC overlay rules presets don't change anything if presets are translated 12:31 ES[m] Hi 12:34 ES[m] Is there a way to show a particular field to the normal view in OPAC (cgi-bin/koha/opac-detail.pl?biblionumber=1) and to Staff interface Catalog Search (cgi-bin/koha/catalogue/detail.pl?biblionumber=1&found1=1) 12:46 cait Annelisterman[m]: I think you found the piece Joubu forgot :) 12:46 cait we just had a change in how translations are handled and maybe the sandboxes need some work 12:46 cait Joubu: around? 12:47 Annelisterman[m] cait, I wondered if that was the reason. :) Well, I'll have to try later. 12:48 cait this would be a good one to get in finally 12:49 cait ES[m]: can you explain a bit more about what you are trying to do? 12:56 ashimema they very likely do 13:00 ES[m] In order to locate a book we have created a field 995q (shelve number). We would like this fiedl to be shown in normal view or at least at the Exemplar (see pictures bellow) 13:01 * ES[m] uploaded an image: (54KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/evawfhIqVwglqigVkFgIGPQO/normal_view.PNG > 13:01 * ES[m] uploaded an image: (48KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/bVobUiVtXvxjkqpDjIeQIuzj/table_exemplar.PNG > 13:01 cait you can see it in the MARC detail view 13:01 cait but for the normal views you'd have to create custom XSLT files 13:01 ES[m] s/fiedl/field/ 13:01 cait the views use XSLT to transform MARCXML to HTML 13:02 cait ah ok 13:02 cait you are a UNIMARC user? 13:02 cait is your 995q mapped to a database column in items? 13:02 cait I thought we were talking biblio level... but I guess we are in the items 13:03 tcohen[m] hi all 13:03 cait and what is shelf number? can you describe it? 13:03 cait hi tcohen 13:03 tcohen[m] it seems I missed a lot 13:04 ES[m] cait: Yes. 13:04 tcohen[m] Joubu: where are we removing the po files 13:05 tcohen[m] oh, it is done :-D 13:08 cait ES[m]: which column and how are you using the sehfl number? 13:08 cait what is it :) 13:08 cait i hope ican make a suggestion 13:21 ES[m] <cait> "and what is shelf number? can..." <- Is the number of the shelf where a book is (stored/placed). I couldnt find a field so I created one (995q) 13:24 ES[m] How can I check this (see the configuration of the field in attached picture) 13:24 * ES[m] uploaded an image: (23KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/wWxWtYhXSVngjKkdfYyoNxYs/field_settings.PNG > 13:26 cait ok 13:26 cait we don't ahve a good column for that I think 13:27 ES[m] When adding an exemplar the librarian inserts it. 13:27 * ES[m] uploaded an image: (26KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/DqNGKjydnkkNGCRVhsugrwtZ/shelf%20number.PNG > 13:29 ES[m] cait: Wha tis your experience of locating books in shelf? 13:30 cait sorr yin a meeting 13:30 cait I'll try to reply in a bit 13:35 ES[m] This need of the "shelf number" came when creating the labels for the books (we print 4 fields) one of the is the "shelf number" so that our librarians can easily locate a requested book and also easily return the book to the shelf. 13:36 ES[m] I have also create an Authorized Value (stack) 13:51 ES[m] It show in OPAC in Marc view, but in Staff interface neither in Normal and MARC 13:52 * ES[m] uploaded an image: (33KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/jUeAcddUTTtpBdBQRUUXBkIn/opac_marc_view.PNG > 13:58 cait I think numbering the shelves is not done often 13:58 cait shelves often can change here 13:58 cait it's usually alla bout the callnumber 13:58 cait for locating, and the branch and maybe the location 14:05 ES[m] <cait> "it's usually alla bout the..." <- I think I red something about it. But let me do some research on "call number". Thanks for the tip. 14:11 ashimema cait around? 14:11 ashimema I have a fun searching thing with unlauts I wanted to run past you if you are 14:13 ashimema or anyone else who understands character encodings a bit more than I do 14:13 cait on phone 14:14 ashimema I'll post my weird search here just in case 14:15 ashimema We're using the main search in authorities "Search main heading ($a only)" 14:16 ashimema Searching for Karamustafa, Gülsün yields no results... searching for Karamustafa, Gülsün gets the right result (those look exactly the same to me.. but somehow they're different under the hood).. and searching for Karamustafa, Gulsun also gets the right result 14:16 ashimema any idea's what might be the difference behind the scenes on those first two and how I might get zebra to go unlaut blind on the former one as well? 14:21 caroline I don't have the answer to your question, but I think the difference between ü and ü is that one is ü and the other is umlaut + u (two different "characters") 14:21 * ashimema just found this fun https://www.coderstool.com/unicode-text-converter 14:22 caroline I often get the problem with the more ubiquitous é in french. somehow it is sometimes encoded as é and sometimes as acute + e 14:22 ashimema highlights the difference much more clearly for me, thanks 14:22 cait ES[m]: the callnumber is usually what you woud put on a label on the book 14:22 cait so we al ahve that in one field, some might use the copynumbr, but it's not as separate as yours 14:23 cait ashimema: there is something special about the Turkish ü 14:23 cait it has to do with combined and uncombined yes 14:23 ashimema interesting 14:24 cait I believe the character is not quit the same as the German umlaut 14:24 cait when I type ü that is a key, not 2 combined 14:24 ashimema I see.. 14:25 ashimema so I need to work out the zebra foo to mask whichever of the two it's not already masking 14:45 ashimema right.. stupid question time again 14:45 ashimema how do I actually type the umlaut + u combination char 14:45 ashimema I can copy the lower case one from the string I have here 14:46 caroline that is what I would do hehe 14:46 ashimema but I want to also catch upper case versions in the config change 14:46 ashimema also.. is this something we think might be helpful in community zebra config? 14:46 ashimema or is my customer just really weird 😜 14:47 caroline Is this something they did in Koha? When I see this problem, it's from records exported from other systems. I "convert" it to the correct encoding and import that file into Koha 14:49 ashimema but is it the 'correct' encoding.. or is Koha just not understanding it correctly 😉 14:49 ashimema they're importing authority files from various places all the time.. I don't think they should have to know to convert from one to the other.. especially when display wise it's impossible to tell there's any difference 14:49 caroline hm, wikipedia has an interesting explanation 14:49 caroline In some alphabets such as those of a number of Romance languages or Guarani it denotes an instance of regular U to be construed in isolation from adjacent characters with which it would usually form a larger unit; other alphabets like the Azerbaijani, Estonian, German, Hungarian and Turkish ones treat it as a letter in its own right. 14:50 caroline https://en.wikipedia.org/wiki/%C3%9C 14:50 ashimema exactly 14:51 caroline maybe then it's the language of the keyboard that typed it? Like cait said, she has a single key for it, but I have to do umlaut + u 14:51 ashimema so I would say that the letter isn't "incorrectly" encoded in the other system in this case.. it's simply that Koha doesn't treat it properly in our diacritic insensitive searches 14:51 ashimema pretty sure we can add it to word-phrase-utf.chr 14:52 ashimema I just can't find an upper case one to copy 😛 14:52 ashimema and.. probably there are loads of them to add mappings for.. not just u with umlaut 14:53 ashimema vim to the rescue 14:57 ES[m] <cait> "ES: the callnumber is usually..." <- We still use the old 4 text line labels (CDD, Autor First Letter+Title FirstLetter, Copy Number, Shelf Location) 14:58 cait are you using locations yet? 14:58 ES[m] I will deep dive in Call Number and on monday I will give a feedback. 14:58 cait in your library? 14:58 ES[m] Thanks for the guidance. 16:04 reiveune bye 16:17 ashimema anyone here able to remind me what the db passwords are for sandboxes 16:24 domm[m] I'm trying to improve Bug 33036, which takes some code from a cgi-script (cataloguing/merge.pl) and puts it into Koha::Biblio. I assume that we should also change to code from raw SQL to use DBIx::Class? If yes, is Koha::Biblio the correct place for the code? 16:24 huginn` 04Bug https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=33036 enhancement, P5 - low, ---, koha, Needs Signoff , REST API: Merge biblio records 16:24 domm[m] I somehow doubt that, as I don't find for example txn_do in Koha::Biblio 16:27 domm[m] or, a more generic question: If we want to make code that's currently only available via CGI available via the REST API, we need should not copy that code from some script.pl into Koha::REST::V1::SomeController, but move it to a generic Model (or a DBIx::Class Result/ResultSet Class, if Koha uses those for Models). 16:27 domm[m] But what's the correct place for eg a biblio model? 16:45 domm[m] Well, I found some inspirational code in Koha::Patron (as hinted in the issue in a comment..) 17:10 ashimema There's already Biblio endpoints 17:11 * ashimema reads up for context 17:12 ashimema Ok, I see.. this is building on that 17:26 KevinFurst[m] Back with more questions. In setting up the cronjobs I notice that the KOHA_CRON_PATH is defined by KOHA_USER_ is not. Is that user picked up via the Koha config file being defined? If not, is that supposed to replaced with the Koha user for the server? I'm getting cron errors like /bin/bash: line 1: __KOHA_USER__: command not found in my logs. I wasn't checking the logs regularly but now, since I've set the logs to be emailed, they're 17:26 KevinFurst[m] coming through frequently. I used the recommended cronjobs from the documentation 17:29 KevinFurst[m] * Back with more questions. In setting up the cronjobs I notice that the KOHA_CRON_PATH is defined but KOHA_USER_ is not. Is that user picked up via the Koha config file being defined? If not, is that supposed to replaced with the Koha user for the server? I'm getting cron errors like /bin/bash: line 1: __KOHA_USER__: command not found in my logs. I wasn't checking the logs regularly but now, since I've set the logs to be emailed, 17:29 KevinFurst[m] they're coming through frequently. I used the recommended cronjobs from the documentation 17:30 KevinFurst[m] * Back with more questions. In setting up the cronjobs I notice that the KOHA_CRON_PATH is defined but KOHA_USER_ is not. Is that user picked up via the Koha config file being defined earlier in the crontab? If not, is that supposed to replaced with the Koha user for the server? I'm getting cron errors like /bin/bash: line 1: __KOHA_USER__: command not found in my logs. I wasn't checking the logs regularly but now, since I've set the 17:30 KevinFurst[m] logs to be emailed, they're coming through frequently. I used the recommended cronjobs from the documentation 17:30 bag ashimema: I read that as bilbo :P 17:33 ashimema Lol 17:47 domm[m] a small questions regarding patches and perltidy: the qa scripts complains that a file I touched is dirty. I run perltidy on it, which results in a lot of changes. Shall I still commit the tidy version, even though the actual change is now hard to see? 18:07 oleonard-away domm[m]: You want to make sure your changes are tidy so that it doesn't *increase* the untidiness. You don't have to fix all of it. 18:08 oleonard domm[m]: If you did want to tidy the whole file you should submit that as a separate patch, distinct from whatever change you're making now 18:13 ashimema What he said 18:13 ashimema Tidy just the sections you changed 18:35 domm[m] huh, so that's a manual job? or can I call perltidy on a subset? 18:39 oleonard domm[m]: I've always done it manually, don't know if there's another way 18:40 domm[m] I doubt it, because perltidy will lack the context of the surrounding code 18:42 ashimema There's ways 18:42 ashimema I have a vim script to tidy just my selection 18:42 ashimema Sure Joubu had a good way too 18:47 ashimema > with --line-range-tidy (see Zeno's email) and the "keep my changes" flag you can pass to git checkout we could actually script that 18:47 ashimema --line-range-tidy 18:47 ashimema Look back https://matrix.to/#/!jlTwDfIFGIaPmcyHTc:matrix.org/$rTIy6nZ1XZ_mbXIvaEPRJY5mk5sg0yUV-OHzz2ydM8w?via=matrix.org&via=kit.edu&via=validad.net 18:48 ashimema We talk about seeing if we can try to magic the tidy challenge away into commit hooks or QA script magic 18:48 ashimema But we've just not had time to pursue it eyt 18:48 ashimema I'd like to see it 21:15 tcohen[m] I always select an area and let a VSCode shortcut do the trick. On that specific portion. It is documented on the wiki 21:30 ashimema Yup, that's the same as my vim setup 21:36 tcohen[m] The QA script checks are not perfect 21:36 tcohen[m] I have tidied a full file just to check, and it kept complaining LOL 21:51 ashimema Been there