IRC log for #koha, 2008-06-30

All times shown according to UTC.

Time S Nick Message
15:18 aindilis Odd_Bloke: thanks, however those are not the errors
15:18 aindilis gmcharlt: no I have not
15:19 aindilis gmcharlt: well yes I've 'create database koha;
15:19 aindilis '
15:19 gmcharlt aindilis: ok, that's what I meant
15:20 gmcharlt aindilis: you're probably safe to run sudo make install now - the syspref errors are probably a bug in test suite
15:20 aindilis okay thanks, that's what I was thinking
15:20 aindilis I just didn't want to blow what has otherwise been a flawless install
15:20 acmoore gmcharlt, I've never seen those test errors that Odd_Bloke is pointing out. (not that I'm surprised they're happening). Have you?
15:21 acmoore I wouldn't mind having that part of the test suite actually working for new installs.
15:21 acmoore aindilis, can you send me the errors that you're seeing when you run 'make test', please? I'd like to get them sorted out for other installers.
15:22 acmoore aindilis, you can use a nopaste site, or send them to
15:23 aindilis k
15:23 acmoore aindilis, or, it's probably better to send them to the mailing list, I guess.
15:23 aindilis k
15:23 acmoore thanks!
15:24 aindilis np, thank you.  Koha rocks!
15:24 Odd_Bloke acmoore: Could they possibly be related to the errors I'm now getting from Apache?
15:24 gmcharlt Odd_Bloke: what errors?
15:24 acmoore Odd_Bloke, perhaps. what kinds of errors are you getting from apache?
15:25 Odd_Bloke 04:46:00 < Odd_Bloke> I'm now getting "[Sun Jun 29 03:40:58 2008] [error] [client] Can't locate object method "preference" via package "C4::Context" at  /usr/share/koha/lib/C4/ line 30." in the Koha Apache logs whenever I try to access it.  What could be causing this?
15:29 acmoore Odd_Bloke, that's pretty odd. I can't think of what would be causing that.
15:30 gmcharlt Odd_Bloke: are you using etch or lenny?
15:34 gmcharlt Odd_Bloke: this link from mailing list archives may be useful:[…].-td17409773.html
15:39 Odd_Bloke gmcharlt: Etch.
15:41 Odd_Bloke Is there any way I could give you more debug information?
15:42 gmcharlt Odd_Bloke: output of perl -V would help; also, do you have perl 5.8 or perl 5.10 installed, or both?
15:44 Odd_Bloke gmcharlt: and only 5.8.
15:45 gmcharlt in the Apache configuration is mod_env enabled, and what is the settings of the SetEnv directives in the Apache conf for Koha
15:48 Odd_Bloke gmcharlt: It is enabled and << SetEnv KOHA_CONF "/etc/koha/koha-conf.xml" >> and << SetEnv PERL5LIB "/usr/share/koha/lib" >> are in the Apache conf.
15:53 gmcharlt Odd_Bloke: had you been able to get through the web installer on the staff interface?  or is this Apache error what you got when you tried it?
15:55 Odd_Bloke gmcharlt: This is the error I get when I tried it.
15:56 Odd_Bloke Though looking at the OPAC error log I'm getting the same one whenever I attempt there as well.
15:57 Odd_Bloke Furthermore, there is the occasional "File does not exist: /usr/share/koha/intranet/htdocs/prx1.php" in the _staff_ interface error logs, at the same time as some of the other errors.  There doesn't seem to be a pattern to these occurring, however.
15:57 Odd_Bloke gmcharlt: Thanks for helping me with this, BTW. :)
15:58 gmcharlt the prx1.php stuff is probably somebody probing your website for an unrelated vulnerability
15:59 Odd_Bloke Actually, yeah, the IPs aren't mine.
16:02 gmcharlt Odd_Bloke: if you don't mind trying a local change to your /usr/share/koha/lib/C4/, try changing line 30 to
16:02 gmcharlt my $debug_level = 2;
16:03 gmcharlt this temp change will avoid the error, and let the CGI script pass along the original error it was trying to report - that may give us more information
16:05 Odd_Bloke gmcharlt: It's worked:
16:06 gmcharlt hmm - getting warmer
16:06 gmcharlt could you paste in the first 40 lines of your /usr/share/koha/lib/C4/
16:07 Odd_Bloke gmcharlt:
16:08 Odd_Bloke Wait, I just noticed fail on my part.
16:09 gmcharlt yeah, missing semicolon
16:09 Odd_Bloke Right, now I've reinserted it, I'm just getting "Undefined subroutine &C4::Context::get_versions called at /usr/share/koha/lib/C4/ line 40." in the error logs.
16:09 Odd_Bloke And the browser is actually receiving something from Koha (though only as far as <body>).
16:10 gmcharlt ok, for the get_versions line, change to:
16:11 gmcharlt my %versions = (kohaVersion => join("\n", @INC));
16:12 Odd_Bloke gmcharlt: Aha, I'm now getting an actual error page. \o/
16:17 Odd_Bloke gmcharlt: It's working. \o/
16:18 gmcharlt Odd_Bloke: yay!  you're seeing the web installer, or does the error message now give you a clue with work with?
16:20 Odd_Bloke gmcharlt: I'm seeing the web installer, as the error message told me I was missing a dependency.
16:21 gmcharlt ahh - which one?
16:21 Odd_Bloke, which is in Net::Z3950::ZOOM in CPAN.
16:22 gmcharlt do you know if perl Makefile.PL had complained about the missing ZOOM dep?
16:24 gmcharlt Odd_Bloke: I'm going to open a bug report for this - do you mind if I place the IRC discussion in it?
16:25 Odd_Bloke gmcharlt: I don't know if Makefile.PL complained, though I did install some dependencies that it did complain about (so I expect I would've installed it as well if it were mentioned).
16:25 Odd_Bloke gmcharlt: That's fine.
16:25 gmcharlt thanks
16:30 gmcharlt entered as bug 2285
16:30 Odd_Bloke gmcharlt: Cool.
16:33 Odd_Bloke Can I find docs referencing 3.0 rather than 2.2 anywhere?
16:37 gmcharlt Odd_Bloke: a draft version of the users manual is available at[…]/koha-manual/Home
16:37 gmcharlt some admin and dev info is scattered across the dev wiki:
16:39 Odd_Bloke OK, so I'm not really looking for anything at all intensive from Koha.  Is there a way I can retrieve MARC records (and authorities) from somewhere else in bulk?
16:40 gmcharlt to get specific records, you can use Z39.50 - e.g., the Library of Congress's Z39.50 server is on the sample list that comes with Koha
16:41 Odd_Bloke gmcharlt: OK, thanks.
16:42 gmcharlt if you just want batches of MARC records to play with, you can get various datasets from
17:28 Odd_Bloke gmcharlt: So I should be seeing something from the LoC in the list of Search targets when I go to Home > Cataloguing > Add MARC Record and click on z39.50 Search?
17:30 gmcharlt it shoudl give you a search box - enter somethign in the title field, pick the LoC target, and do a search
17:31 Odd_Bloke OK, I don't have anything under the "Search targets" header (except the Library of Congress entry I've added and evidently got wrong myself).
17:34 gmcharlt Odd_Bloke: here's how you can get the defulat entries
17:34 gmcharlt delete the LoC Z39.50 server entry you created
17:35 gmcharlt then use mysql to load the following script from distribution
17:35 gmcharlt installer/data/mysql/en/opt​ional/sample_z3950_servers.sql
17:38 Odd_Bloke gmcharlt: Thanks, that's working now. :)
17:46 aindilis well all my problems were due to CGI::Session
18:49 gmcharlt aindilis: what was the CGI::Session problem, specifically?
19:15 aindilis not sure exactly, mainly that there was no CGI::Session::ID::*
19:16 aindilis which was causing a fatal error
19:17 aindilis so doing 'force install CGI::Session::Serialize::yaml' I think it was caused a version of CGI::Session (4.20) with the ID modules.
20:28 mc aindilis, i have no proper way to work around it but there is a way
20:28 mc :
20:29 mc don't use CPAN!
20:29 mc (hello all)
20:29 mc just go to
20:29 mc serach CGI::Session::Serializer::yaml
20:29 mc search CGI::Session::Serializer::yaml
20:29 mc download it
20:29 mc dh-make-perl --build
20:44 hdl_laptop mc ?
20:49 mc yep
20:50 masonj thanks for the cgi::session info aindilis and mc
20:53 mc masonj, np ... this problem just (drive|drove|drived)? me crazy last week..
20:53 masonj yeah, me too a while back..
20:54 mc drived ? drove ? drive ?
20:54 masonj ive had the problem on different installs too,  deb and ubuntu
20:54 masonj drove
20:54 mc thx
20:55 masonj and each time i fixed it differently...
20:55 mc i too
20:55 masonj and i tried *sooo* many things that i wasnt sure what i did - that fixed it...
20:55 mc but it was something i felt as weird work around
20:55 masonj it just started working.....
20:55 mc with a message "please don't touch: it miracluously works"
20:56 mc now i found the method that works really and i know why
20:56 mc i just felt as you on this point
20:57 mc i also tried to work around into koha
20:57 mc with my patch
20:57 mc but it wasn't applaied
20:57 mc applued
20:57 mc damn!
20:57 mc applied
07:41 chris evening
07:41 paul hi chris
07:41 masonj hi guys
07:41 chris hey paul and mason
07:42 paul hello masonj
07:43 paul BibLibre is continuing it's growth...
07:43 chris oh? hire another person?
07:43 paul Jacques Piton, a senior project manager will join us, as asset partner next month.
07:43 chris ohh cool
07:43 masonj woooo
07:43 chris congratulations
07:43 paul Laurence Lefaucher, half librarian, half developper will join us in october
07:44 paul we are waiting for an answer from Celine Girardeau (librarian)
07:44 paul and should decide between 2 developpers this week.
07:44 chris thats great news
07:44 paul yep.
07:44 masonj good timing for k3 too
08:07 chris paul: did you see my opac running under mod_perl?
08:07 paul yep. link transmitted to mc, that is investigating as well
08:08 chris cool
08:09 chris i can send him my httpd.conf if that would help?
08:09 paul good idea. marc.chantreux at
08:09 mc :)
08:09 mc sure it will help
08:09 mc thanks chris
08:09 chris cool, emailing it now
08:09 mc
08:09 mc
08:10 mc thanks a lot
08:13 chris on its way
09:42 frederic hello
09:52 chris hello frederic
09:54 paul hello frederic
10:28 chris hey nicole
10:32 slef Hi all.  Anyone else having problems with serials management?
10:40 frederic hello chris and paul. Have you a solution to your/our koha performance issue? Shouldn't Koha be always deploy with mod_perl?
10:41 chris frederic: it needs to be tested more, but so far it seems to work ok with mod_perl
10:41 chris certainly things like mod_expires should be used to tell the browser to cache the images, css and js
10:41 slef Koha should not require mod_perl IMO.  It will severely reduce use by smaller libraries.
10:42 chris yep, theres no reason why it can work as cgi or mod_perl though
10:42 chris s/can/cant/
10:43 frederic chris: I activated a Koha setup with mod_perl and didn't observe any improvment. Should I?
10:43 chris currently output_with_http_headers which almost all scripts use
10:43 chris tells the browser not to cache
10:44 chris so telling the browser to cache the static files helps
10:44 chris frederic: yes, a quite marked improvement
10:44 slef chris: does output_with_http_headers serve static files?  If so, have we noted that we need to fix output_with_http_headers?
10:45 frederic But browser cache all static files by default. You don't have to tell it anything...
10:45 chris not if you are telling it not too
10:46 slef Does ever find biblios for serials which have not yet got any items?
10:48 slef query has "AND itype=?" on the end... where's itype come from?
10:49 chris thats the itemlevel itemtype
10:50 slef Is there a biblio-level one?
10:50 chris frederic: my went from 2 secs to .6 seconds with mod_perl
10:50 chris biblioitems yep thats the original one
10:50 frederic chris: Do you have a koha setup with mod_perl enabled so I can compare with mine?
10:50 slef chris: sorry, I don't understand.
10:50 frederic Mine statys at around 3 s
10:51 chris slef: itemtype was on the biblioitems table originally
10:51 chris there is now one on the items table too
10:51 chris and a syspref controlling which one you use
10:51 frederic slef: take a look a record.abs. You wil find where itype and itemptype come from
10:51 chris i suspect the serials arent honouring the syspref
10:52 chris frederic:[…]nough-speed-work/
10:52 chris but now i must go to bed
10:52 chris night all
10:52 frederic chris: thks and good night
10:52 nengard firefox question ... does anyone know how i can change the default tool that my browser automaticallys subscribes to rss with? I have it set to automatically use bloglines ... but i need to turn that off to do some testing for the manual
10:53 nengard FOUND IT!!! Damn thing is hidden
10:53 slef nengard: Everything is hidden in ff3 behind "you are a stupid user" labels :(
10:54 nengard slef: hehe
10:54 slef federic: will searching either itype or itemptype find biblios with no items?
10:58 frederic slef: In MARC21, itemtype is mapped to 942$a dn itemptype is mapped to 952$y. So itemtype is biblio level item type and itype is item level itemp type.
10:58 frederic Correction itype is mapped to 952$y
10:59 slef frederic: thanks!  So what do I want to search?
10:59 frederic So to find biblio records without item, you have to search on NOT itype equals something.
11:00 slef trouble is, the context is wanting to find biblio records of type SER which may or may not have items yet
11:01 slef but at the moment, it's searching on itype = SER which doesn't find zero-item biblios
11:02 frederic Take a look on that:
11:03 slef 942$c seems to be the thing to search on
11:04 slef interesting git commit
11:05 frederic slef: Depending of item-level_itypes syspref, itype or itemtype is used in standard OPAC search
11:05 slef 942$c is itemtype
11:07 slef "If ON, enables Item-level Itemtype / Issuing Rules"
11:08 slef well, ignores that either way... I strongly suspect it should search on itemtype always
11:12 slef kados: can you review kohabug 1536 when you wake up, please?
11:13 nengard slef: kados is at ALA conference ... he may not be on much at all - just FYI
11:13 slef nengard: oh, I thought that was over the weekend.
11:14 nengard slef - it was the weekend - and today i think
11:14 slef nengard: I wish people would set their /away status ;-)
11:18 nengard sys pref "SubscriptionHistory" - can someone tell me the diff between the options?
11:18 slef nengard: oops... I just closed the koha source... 1mo
11:20 slef nengard: looks like it provides a default value of selectview in serials/
11:20 slef nengard: I don't find it doing anything else.
11:21 nengard which one of the options provides the default value? or are you saying that there probably shouldn't be two options?
11:22 nengard puppy on my lap doesn't seem to understand that i'm back to work :)
11:23 slef nengard: are the values small and full?
11:24 nengard slef: Default Value: simplified
11:24 nengard Values:
11:24 nengard    * simplified =  
11:24 nengard    * full =  
11:24 slef ok :-/
11:25 slef nengard: that disagrees with the OPAC code a bit, but doesn't seem to matter
11:25 slef simplified or small gives a one-sentence summary of the subscription
11:25 slef full gives tables by year of the subscriptions, one per line
11:26 slef nengard: the small-v-simplified discrepency may be worth a minor or trivial bug report, if you are so inclined.
11:26 nengard slef - i'll have to check the most recent install and see if it's still labeled as such ...
11:27 slef nengard: I'm checking on a 3.0rc1... I can pull and check whether it's changed
11:29 slef yes, koha-tmpl/opac-tmpl/prog/en/modu​les/opac-full-serial-issues.tmpl and the syspref tables disagree
11:41 nengard thanks slef!!
11:45 slef no problems... can you patch the syspref description?

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