Time  Nick       Message
12:06 gmcharlt   nahuel: about?
12:06 nahuel     gmcharlt, yes ?
12:07 gmcharlt   nahuel: noticed that you had emailed a few applied patches (2868, 2799) back to me
12:07 gmcharlt   just wondering why
12:08 nahuel     hmmm when ?
12:08 gmcharlt   about 3 hours ago
12:08 nahuel     ahaha :)
12:08 nahuel     git-send-mail sends the patch to the commiter too...
12:09 gmcharlt   ah, OK
12:09 nahuel     I just made a git-format-patch too long, and sent all patches
12:09 nahuel     a little mistake, sorry
12:09 gmcharlt   no problem - just want to make sure you hadn't found something wrong in those patches that you wanted me to fix
12:09 nahuel     no no :)
17:16 frederic   Does anybody get a bug which corrupts MARC21 952 items classification subfields ($6 en $o)?
17:18 frederic   Both subfields are fulfilled with those kind of string: Ã<U+0083>Â<U+0083>Ã<U+0082>Â
17:48 gmcharlt   frederic: haven't see that one
18:05 atz        frederic: sounds like a unicode handling issue
18:05 atz        that in particular looks like it may have been using 3 combining characters to represent 1 unicode character
18:06 atz        it makes sense that you would see it in MARC21 since it lacks the encoding brains of UNIMARC
18:18 atz        most of the time i see that in z3950-imported records
18:19 atz        and there is some question about whether the remote system has their representation correct or not
18:19 atz        i see it less often with Koha-original catalogued items
18:21 frederic   atz: thanks. Not sure MARC21 brain is inferior than the UNIMARC one but...
18:21 frederic   This mess is concentrated on 952 classification subfields. All other fields are properly UTF8 encoded.
18:22 frederic   Is it possible that during import something get wrong because classification contains extended characters? é for example. I'm sure it has worked...
18:22 atz        interesting... so the local info was entered at a different time than the original biblio info
18:24 frederic   I have in phone a librarian in tears who don't believe in Christmas anymore...
18:27 cait       atz: is the right form for diacritics in koha one single unicode-character? we had a problem with german umlauts and i wondered whis is the right form
18:30 ryan       cait: the trend in the unicode world is to move to standardise on combining characters rather than using the composed (single-character) form.
18:30 ryan       I'm not sure which is more likely to cause problems in koha though.
18:31 cait       with german umlauts combined character breaks the search in opac
18:31 cait       you simply cant search for them, so we changed it to ü instead of u + ..
18:33 gmcharlt   cait: you're using Zebra, right?  if so, it can be configured to ignore combining vs. noncombing character distinctions
18:33 cait       and the display is slightly wrong, the points are a litte bit out of alignment
18:33 cait       where can i do that? and how?
18:33 cait       and yes, we like the zebra :)
18:34 gmcharlt   cait: I'll work up an example for you in a day or two
18:34 cait       that would be great
18:35 cait       there is also a problem in acquisition - when an order is placed with a "fast add" and there is an umlaut in it, it breaks acquisition, but i think there is a bug report somewhere
18:37 chris      morning
18:38 gmcharlt   hi chris
18:39 chris      i have a question
18:39 chris      we have a new listing on the koha users page
18:39 chris      for a library in jerusalem
18:40 chris      its in the asia section at the moment, which isnt quite right :) do you reckon i should make a middle east section?
18:40 gmcharlt   chris: I'd say so  - I think there are a couple other Koha users in the Middle East
18:41 atz        chris: yes, there are also known users in Saudi Arabia, and Iran
18:41 chris      yes, if we put the section there, maybe they will list in it :)
18:41 chris      cool ill do that then
18:46 chris      done
18:48 gmcharlt   cool
18:49 atz        gmcharlt: i'm surprised, but it seems this hasn't been done yet in Koha: write a hash-to-TMPL_LOOP converter.
18:49 atz        true?
18:49 gmcharlt   atz: not sure what you mean by that
18:50 atz        TMPL_LOOP is a reference to an array of hashrefs
18:52 atz        take something like:  $hash = ( foo =>1, bar=>2, tre=> {foo=>100; bar=>200}};
18:53 atz        er, missing a ), but you get the idea
18:56 atz        you'd want:   $loop = [( {foo=>1, bar=>2, tre=>[({foo=>100, bar=>200})] } )];
19:07 chris      http://xkcd.com/519/   for anyone who hasnt seen it yet :)
21:10 acmoore    areinmeyer: you ought to watch http://www.shadowcat.co.uk/archive/conference-video/yapc-eu-2008/dbic-masterclass/ for some interesting stuff about how to artfully use DBIx::Class.
21:11 areinmeyer acmoore: thanks I'll check that out
21:11 areinmeyer been making good progress today on Members
21:13 acmoore    great. I haven't worked much on ImportBatch. I spent most of the morning putting together a patch that I sent over to the DBIx::Class guys. It makes Schema::Loader work a bit better.
21:13 atz        cool
21:14 areinmeyer I was playing around with the DBIx::Class:ResultClass::HashRefInflator
21:15 ryan       HashRefInflator goes nicely with TMPL_LOOPs
21:15 acmoore    yes, it does. I thought I was probably going to ahve to write something to make that easier, but it looks like we can just use that.
21:15 areinmeyer http://search.cpan.org/~ribasushi/DBIx-Class/lib/DBIx/Class/ResultClass/HashRefInflator.pm
21:17 areinmeyer I started writing something as well, then figured someone smarter prb already did it
21:55 acmoore    gmcharlt: In my process of changing C4::ImportBatch to use DBIx::Class instead of direct SQL commands, I'm generating a lot of test cases for C4::ImportBatch. They're valid even if we never switch to DBIx::Class. They just test current functionality. If I package them up and submit them, you'll likely accept them, won't you?
21:56 gmcharlt   acmoore: sure
21:56 acmoore    OK, good. That might cut down on the patchbomb that we're going to have with DBIx::Class. thanks.
21:56 acmoore    areinmeyer: that may mean that you can submit C4::Members tests as you get them done, too.
21:58 areinmeyer super.  I had thought the same thing that they should be valid regardless
22:03 gmcharlt   acmoore++ # DBIx::Class::Schema::Loader
22:06 acmoore    thanks. those guys seemed pretty happy about it, too.
22:06 acmoore    I asked them if it were possible to have that functionality, and they said "sure, patch it". I asked for svn access and their reaction was "holy cow, the last 12 people just disappeared when we said that".
22:07 acmoore    so, now I have commit access to the DBIC repository.
22:08 gmcharlt   explains why you sent a svn link :)
22:14 acmoore    OK. I sent some test scripts for C4::ImportBatch. I have some for C4::ClassSource I can send after I split them apart from actual code changes. I guess I'll do that next Friday.
22:17 ryan       acmoore: DBIx::Class::Schema::Loader  <--  what's that ?
22:29 gmcharlt   ryan: reads schema from DB and generates classes
22:48 liz        would it be welcome from the koha community if a paying customer paid to have a feature (such as the SIP functionality) both enhanced and then fully documented?
22:48 liz        i mean, does anybody do that?
22:49 gmcharlt   liz: sounds like a good idea - documentation is as worthy of support as is the code
08:26 chris      http://opac.koha.workbuffer.org/  now thats a lot of languages