IRC log for #koha, 2024-01-04

← Previous day | Today | Next day → | Search | Index

All times shown according to UTC.

Time Nick Message
02:12 C4R7 joined #koha
02:13 C4R7 Hello
03:11 C4R7 joined #koha
06:35 cait joined #koha
07:14 cait joined #koha
07:17 cait joined #koha
07:28 magnuse ashimema: "ancient perls" - are those the perls shipped with debian and ubuntu?
07:34 ashimema Yup, we lock ourselves to the lowest common denominator.. i.e lts debian
07:35 5.14 perl
07:36 magnuse that is pretty low...
07:36 ashimema 5.24 would give us a fair bit from memory.. and the thinks that people are getting excited about in the perl world are brand new.. objects, signatures, etc
07:36 magnuse there are versions of debian/ubuntu we say we support that has perl as old as that?
07:37 ashimema 5.38 is current perl
07:37 Yup
07:37 magnuse how hard would it be to install a newer perl?
07:39 ashimema Looks like 5.28 is in buster.. the older non-lts debian
07:39 Whilst we stick to system perl, fairly hard
07:40 If we opted to try and go local lib it would mean some infrastructure change but then not so hard
07:40 I don't know the answer to be honest
07:40 Right, swim time for me.. be back in an hour
08:04 lds joined #koha
08:13 cait joined #koha
08:44 fridolin joined #koha
08:45 fridolin salutations
08:48 t/db_dependent/api/v1/erm_counter_registries.t failing on master 23.11
08:48 and t/db_dependent/api/v1/erm_sushi_services.t
08:49 maybe a change in the WS itself
08:49 any idea ?
09:00 Joubu PedroAmorim[m]: ?
09:01 t/db_dependent/api/v1/erm_sushi_services.t is failing because it compares an expected output with https://registry.projectcounte[…]a5f-6c113f7ffa0b/
09:01 and there is "migrations" there
09:01 which may have been added recently
09:01 this test is really bad (ie. depending on an external resource)
09:02 unless it's what we want, but it's not clear what the test is actually testing
09:06 * cait waves
09:07 cait can anyone help with a password reset on the wiki? It looks like emails are not being sent out
09:07 Joubu fridolin: added a comment on bug 35218
09:07 huginn` 04Bug https://bugs.koha-community.or[…]_bug.cgi?id=35218 blocker, P5 - low, ---, martin.renvoize, RESOLVED FIXED, No tests for /erm/sushi_service
09:08 fridolin Joubu: tanks a lot
09:18 PedroAmorim[m] I was not aware of this
09:18 can take a look
10:20 dolf joined #koha
10:24 dolf Hi there. I'm trying to upgrade an old Koha (21.05) step by step, starting with 21.05 -> 21.11. It's on debian, using apt. I updates the codename to 21.11 in /etc/apt/sources.list.d/koha/list and ran `apt update` followed by `apt upgrade`. The database schema upgrades started at 21.06.00.000 and went swimmingly until 21.06.00.041, when I got this error: ERROR - {UNKNOWN}: DBI Exception: DBD::mysql::db do failed: Row size too large. The maximum row size for the
10:24 used table type, not counting BLOBs, is 8126. This includes storage overhead, check the manual. You have to change some columns to TEXT or BLOBs at /usr/share/koha/lib/C4/Installer.pm line 743 . I'm not sure what to do with this...
10:24 cait you don't need to do a step by step update
10:25 but the error you ahve needs to be resolved, give me a moment
10:25 Bug 28267  - Older databases fail to upgrade due to having a row format other than "DYNAMIC"
10:25 dolf I realize that, but I tried to upgrade directly to the latest version before, and got the same error, so I reverted the state of the VM and tried again with a smaller increment. Also, our staff members like to test things thoroughly in between upgrades. Anyway, thanks for your time and attention!
10:26 cait yes, but testing every versin in between might be a bit of wasted energy
10:26 dolf Ah, it's the row format again. I think I ran into this on another instance a while ago!
10:26 cait have a look at the bug I posted, the link is https://bugs.koha-community.or[…]_bug.cgi?id=28267
10:26 huginn` 04Bug 28267: critical, P1 - high, ---, jonathan.druart+koha, RESOLVED FIXED, Older databases fail to upgrade due to having a row format other than "DYNAMIC"
10:28 cait comments 20/21 especially
10:30 dolf Thanks. I also found https://irc.koha-community.org[…]a;date=2023-08-16 where I discussed the same problem. I'll just pick up from there! Thanks again.
10:48 paulderscheid[m] morning #koha
10:50 I think we should make some efforts to get to 5.036 at least as one of perl's flagships (even if not this cycle). And I want these type checks as soon as they're in core :D
10:53 dolf Has the wiki account creation / email sending been sorted out yet? Back in August, davidnind tried to help me get my wiki account activated, but none of the e-mails are reaching me, so I can't (re)set my password. I would like to improve this page: https://wiki.koha-community.or[…]tabase_row_format
10:54 cait dolf: we just found it not workign today/yesterday for another user
10:55 maybe you could comment here too: Bug 34637        - Wiki - email notifications aren't being sent (account registrations, password resets, etc.) - I left a comment earlier, but strangely I am receiving page update notifications to my email
11:01 Joubu paulderscheid[m]: we can already improve some of our old code. Waiting for a future version of Perl is just an excuse to procrastinate :D
11:02 paulderscheid[m] You are right
11:02 ˆˆ
11:03 Joubu there are several places where prototype of subs is bad and can be improved/cleaned already
11:11 oleonard Hi #koha
11:50 cait hi oleonard
12:18 dolf cait: I commented on both issues. On Bug 28267 (the one about the DYNAMIC rows) I posted a link to the script that I used to convert everything to DYNAMIC row format. After that, the database upgrades worked, and this Koha instance is now on 21.11 at last. However, now I'm seeing another issue that I also had in August, and never got resolved (I ended up rolling back the VM). The discussion started here https://irc.koha-community.org[…]3-08-17#i_2503120
12:18 . In summary, the search results and normal view on the biblio page are both omitting lots of data, even though the MARC data is all there and in good condition.
12:18 huginn` 04Bug https://bugs.koha-community.or[…]_bug.cgi?id=28267 critical, P1 - high, ---, jonathan.druart+koha, RESOLVED FIXED, Older databases fail to upgrade due to having a row format other than "DYNAMIC"
12:21 dolf Unfortunately I did not check in between the row format conversion and the "apt upgrade" step, so I don't know which one is the problem.
12:25 It's looking fine on the intranet, but on the OPAC in "Normal view" on the opac-detail.pl page, no record details are shown
12:26 On the OPAC search page, only the covers are shown: https://library.refstudycentre[…]sort_by=relevance
12:34 lds joined #koha
12:37 dolf brb
12:39 cait @later tell dolf sorry, I was afk. That sounds like you should check your frameworks - especially the visilbility settings. It didn't work in the past and then we fixed it. So field need to be set to be visible in the OPAC
12:39 huginn` cait: The operation succeeded.
12:40 cait and you will always want to have "default" in the XSLT prefs (non-XSLT views have been removed in newer versions anyway)
12:48 @later tell dolf: also check the logs for errors, XSLT errors can make themselves visible like this. In the frameworks you want to check for biblionumber 999 to be set to visible in the OPAC for a start.
12:48 huginn` cait: The operation succeeded.
12:58 lds joined #koha
13:52 dolf joined #koha
13:53 Dyrcona joined #koha
13:58 dolf cait: Thanks, I'll look into that. Although I don't recall ever changing the frameworks or XLST stuff since installing Koha for the first time back in 2011.
13:59 By the way, I reverted my VM to after converting the rows to DYNAMIC, and before upgrading to 21.11 (so it's still on 21.05) and now everything is working. So it's not the row format that is causing trouble. It must be one of the updates not playing nicely with my settings?
14:03 cait dolf: there can be different reasons
14:03 is it only in the opac or the staff interface as well?
14:04 meaning: does the result list and detail page in staff interface display normally?
14:05 dolf Staff displays normally, yes. The problem is only on the OPAC (both in search results and on the detail page)
14:06 cait ok, did you check the logs yet?
14:06 or first: go to your frameworks
14:07 check that all your frameworks have 999$c set to visible in the OPAC
14:07 dolf Will do!, as soon as I get the problem replicated on another VM. I need to keep it in a stable position for the following week until my next maintenance window. So I'm keeping it on 21.05 for now so that the staff can do their work.
14:08 cait ah ok, a separate system for testing might be useful :)
14:09 tail -f /var/log/koha/instance/*.log when you go to result list and look for anything error'y
14:23 dolf cait: Under /cgi-bin/koha/admin/marc_subfields_structure.pl?o​p=add_form&tagfield=999&frameworkcode=#subcfield the OPAC check box is ticked next to "Visibility" under "Advanced constraints". Is this the right place? I noticed that it's not possible to modify that check box.
14:23 cait that's odd
14:23 dolf (Still on 21.05, just poking around)
14:23 cait but yes, visibility
14:23 is it checked?
14:24 dolf Yes, Checked are OPAC, Intranet and Collapsed. Unchecked are Editor and Flagged.
14:28 On the "Tag 999 Subfield structure" page, it says "subfield ignored" in the "Constraints" column of subfield "c". What does that mean?
14:35 marie-luce joined #koha
14:52 cait i think that's ok
14:53 I think checking for an error would be the next step when you try to update again, in the logs, when you do a search
14:56 dolf Will do. In the mean time. I see that "OPACXSLTDetailsDisplay" and "OPACXSLTListsDisplay" and "OPACXSLTResultsDisplay" are empty instead of "default". Maybe that is the problem?
14:56 What did you mean when you said "non-XSLT views have been removed in newer versions anyway" ?
14:58 oleonard-away dolf: There used to be two ways to show those pages, XSLT or non-XSLT.
14:59 caroline joined #koha
15:05 dolf Without upgrading (i.e. still on 21.05) I changed all OPACXSLT*Display settings from empty to "default", and now I see the same problem as when I upgraded: Only the book cover picture is shown. The rest is missing, both on the book detail page and on the search results page.
15:06 Changing back to empty fixed the problem. I don't understand this setting. Maybe the default xslt is missing or corrupted? Is it part of the deb package installed via APT?
15:07 oleonard dolf: I'm sorry if you've answered this already, but have you customized the default XSLT somehow?
15:07 dolf I have not. I would not know how. How can I check if it's still at "factory settings"?
15:08 cait the package update would also overwrite any local changes I think
15:08 we really need the error from the logs
15:08 if you see the problem now too, maybe check the logs now?
15:08 I thin it's something in the configuration
15:09 dolf Good thinking, I should have thought of that. Doing that ASAP
15:12 There are some errors, but nothing new appears in the log when visiting opac-detail.pl , even though I see the issue (no details being displayed other than the book cover and the holdings table)
15:13 cait hm an XSLT bug should be logged soemwhere
15:13 are you checking all logs or only a specific one?
15:14 dolf I'm using tail -f /var/log/koha/instance/*.log as you suggested
15:14 with instance being "rsc"
15:15 cait ok
15:15 hm
15:15 dolf I can paste some of the errors I see, but their timing don't coincide with my page reload
15:15 cait and when you go to results, there is nothing new?
15:15 dolf I don't understand what you mean
15:15 cait tail -f let's you watch the logs when you do the thing
15:15 so you can add empty lines and then reproduce the error, see if something turns up
15:16 dolf Yes, that's what I did. Nothing new pops up when I reload opac-detail.pl
15:16 cait and same for opac-results?
15:16 dolf Yes, same.
15:16 cait i mean the result list
15:16 dolf Got it
15:17 cait yep hm
15:17 there was a short ime when we didn't log XSLT errors right, but that would be a very unlucky coincidence
15:17 dolf I see some other errors popping up, but they seem unrelated. This is a production site, and people are searching, browsing, etc.
15:17 cait but you set XSLT to default now?
15:17 can you share a link?
15:18 I just want to check something real quick
15:18 dolf Yes, three OPACXSLT*Display settings all set to "default", which makes my stuff hidden.
15:18 cait in the browser
15:18 dolf https://library.refstudycentre[…]s&weight_search=1
15:18 https://library.refstudycentre[…]iblionumber=38779
15:19 cait the bit where the output of the XSLt woudl be is completely missing from the page source
15:20 dolf Strange... Where do these default XSLT files live? Can I check them manually?
15:20 cait trying to remember the path for a package installation
15:20 they live in .... opac/bootstrap/xslt i think
15:21 the opac template directory
15:22 dolf I have /usr/share/koha/opac/htdocs/​opac-tmpl/bootstrap/en/xslt . Is that right?
15:22 Other one is /usr/share/koha/lib/Koha/XSLT
15:22 cait yes, that looks right
15:22 no the bootstrap one
15:22 the ohter is the perl module, that shoudl be alright
15:23 dolf I see 12 .xsl files here
15:24 compact.xsl MARC21slim2OPACDetail.xsl MARC21slimUtils.xsl NORMARCslim2OPACResults.xsl plainMARC.xsl UNIMARCslim2OPACResults.xsl MARC21Languages.xsl MARC21slim2OPACResults.xsl NORMARCslim2OPACDetail.xsl NORMARCslimUtils.xsl UNIMARCslim2OPACDetail.xsl UNIMARCslimUtils.xsl
15:25 oleonard dolf: MARC21slim2OPACDetail.xsl & MARC21slim2OPACResults.xsl are for the OPAC details page and the OPAC search results page
15:26 dolf: What's your full Koha version number?
15:26 dolf `apt show koha-common` gives me `21.11.26-1`
15:27 oh wait, I thought I was still on 21.05.
15:28 Ah, `apt show` does not give the installed version, but the latest available one in the apt repos
15:28 oleonard dolf: The Koha "About" page will show you
15:28 dolf It
15:29 It's 21.05.00.000
15:29 oleonard https://gitlab.com/koha-commun[…]lt?ref_type=heads
15:36 dolf I downloaded from git and did a diff. The  MARC21slim2OPACResults.xsl files are identical. The MARC21slim2OPACDetail.xsl files have a diff: https://pastebin.com/qh6Atnbx
15:37 It looks like mine is just slightly older. Should I update to the latest 21.05.* ?
15:39 oleonard dolf: It couldn't hurt to try
15:41 cait I still think itmight be something int he data cuasing an error
15:41 but we need the error to be able to tell more
15:41 maybe if you try to update again, repeat checking the logs
15:51 dolf Something in the data: You mean in my MARC records? We have tens of thousands, and I did a spot check – all have the same problem.
15:51 I updated to 21.05.21.000 now. Checking again ...
15:54 Still no new lines in any log files when loading opac-detail or opac-search. When I change the OPACXSLT*Display settings back to empty, I can see the book details again.
15:55 cait can you still paste what you have in the logs?
15:55 dolf Yes, I'll do that. But I have to go soon. Thanks for all the help!!!
15:56 cait have to go soon too, but maybe someone will also read later
15:57 dolf opac-error.log https://pastebin.com/LD3JHkmg
15:57 Any other files? There are many, but I assumed opac-error.log is the only relevant one.
15:57 Bye now :)
15:57 cait i was thinking of anything that appears when you load the opac-rsult or opac-detail page
15:58 but ther eis nothing obvious there
15:58 dolf Nothing appeared when doing tail -f *.log during a reload.
15:58 cait hm
15:59 sorry, I am out of ideas :(
15:59 have a nice evening/rest of day #koha
16:00 bag joined #koha
16:36 paulderscheid[m] Is metacpan.org broken for anyone else (Search specifically)?
16:55 ashimema All works for me
17:00 cait joined #koha
17:04 paulderscheid[m] thanks ashimema
17:04 now it works for me too
17:05 Have you played w/ Mojo::JWT already ashimema?
17:07 ashimema Yeah, a few times over the years
17:11 paulderscheid[m] Would you recommend it for JWTs for Koha via OpenAPI or rather one of the other packages?
17:11 I think there's also JSON::WebToken
17:11 And many more
17:51 tuxayo paulderscheid:  https://github.com/orgs/Perl-A[…]lo/discussions/49 => is that about optional typing?
17:51 hi all :)
17:56 alexted[m] joined #koha
17:59 alexted[m] hello, I have jsut installed Koha on Debian (following the official Wiki: https://wiki.koha-community.or[…]i/Koha_on_Debian). My question is: which is the default password assigned to the system user "library-koha" created by the "koha-create"? Thanks!
18:11 tuxayo alexted: hi :)
18:11 https://wiki.koha-community.or[…]the_web_interface
18:11 > When you see the login for the Koha installer, the username and password are in the koha-conf.xml file for the instance.
18:12 > You can view the password with:
18:12 > sudo koha-passwd libraryname
18:12 I think it's that
18:17 alexted[m] tuxayo: hi! thanks for you reply :)
18:19 the password you are referring to is that of the MySQL user (koha_library), I was referring to that of the Debian system user (koha-library)
18:39 lukeg joined #koha
18:40 lukeg hi
18:41 cait alexted[m]: what are you trying to do?
18:41 you can use sudo koha-shell instance to switch to this user
19:06 alexted[m] cait: hi cait, thanks for your answer
19:07 i'm tring to rebuild Zebra with "sudo koha-rebuild-zebra -f -v  instancename" and i don't know how to find the password for the koha system user
19:57 cait you don't need it
19:58 that's your normal root user, not the koha one
19:58 all the sudo koha-... commands are just run with your own password
19:59 and if you need to run other scripts you switch to the koha user using the command I gave earlier
20:20 paulderscheid[m] Hi tuxayo => yes
20:20 Also: global big squashing days came up during the last dev meeting, just fyi if you want to organise
20:30 alexted[m] caitok, so so the "koha-rebuild"-zebracommand must be run as root user? thanks!
20:32 cait it will take care of using the right user internally I believe
20:32 you have the instance as parameter
20:41 alexted[m] ok, but the Koha Installation Guide on Debian https://wiki.koha-community.or[…]ki/Koha_on_Debian says that "A system user is created, called library-koha. All things to do with this instance will be run as this user."
20:41 This is why I was wondering if the "koha-rebuild-zebra" command should be run as the library-koha user

← Previous day | Today | Next day → | Search | Index

koha1