IRC log for #koha, 2006-03-17

All times shown according to UTC.

Time S Nick Message
11:23 kados paul: you around?
11:24 kados paul: about your question about new MARC editor
11:24 paul yep
11:24 kados paul: have you looked at the demo of Ohio University student's code?
11:24 paul nope
11:24 paul (an url ?)
11:25 kados it should be listed on labs.liblime.com
11:25 kados there were in fact two software teams in the same class
11:25 kados both competed with different software packages
11:26 kados team 42 had the best features I think, though teamipg has nice keystroke management
11:27 kados (it seems they are finding no results)
11:27 paul (what  can I search ?)
11:27 paul right :-(
11:27 paul what's the status of this code ? do they work on it a little ?
11:28 kados no, no more work now
11:28 kados the class is over so they fled the scene :-)
11:28 kados I can send you the existing two software packages
11:28 kados but really, we need a complete new design
11:29 paul so you think I had better start from scratch something new ?
11:29 kados that's my suggestion
11:29 paul you told me you had some ideas iirc
11:29 kados yes, I do ...
11:29 kados but today I'm quite busy ...
11:29 paul i'm listening (or you write a mail ?)
11:29 kados yes, I'll write an email
11:29 kados maybe later tonight
11:29 paul no problem, I just need to have a clear position at end of month
11:30 kados ok ... we will :-)
11:30 paul did you saw my mail about kohacon on the wiki ?
11:30 kados yes  ... thank you
11:30 kados can we now announce it on koha lists and on our websites?
11:31 kados or are you still waiting for confirmation ?
11:31 paul i'm waiting for something from jps
11:31 kados ok
11:31 paul but if he stays silent, at end of week we will announce the con
11:31 kados sounds like a plan
11:31 paul about koha extension for firefox (on labs.liblime)
11:31 paul do you plan to commit something on cvs ?
11:32 paul could be a good "add on" to Koha.
11:32 paul the 1st separate package
11:32 paul then MARC editor,
11:32 paul then opensearchportal
11:32 paul then acquisition wizard
11:33 paul (acq wizard is something ouest provence want me to developp, i have a mail to write on koha-devel)
11:54 kados yes, I do plan to commit it
11:54 kados in fact, I registered a special domain for it :-)
11:54 kados kohazilla.org :-)
11:54 kados nothing there yet, but soon
11:55 kados it will allow any library to create a custom search plugin for their system and then upload it to kohazilla.org
11:57 paul pierrick : a few months ago we were speaking of adding new features to Koha, of course, but also adding new tools. some things that could be standalone apps, but that works also with Koha
11:57 paul the 1st one that will be developped will be a MARC editor. i have a student with me for 2 months.
11:57 pierrick OK, you noticed that to me when we met
11:57 paul joshua also began opensearchportal and an extension for firefox.
11:58 paul look at labs.liblime.com
11:58 pierrick I did
11:58 pierrick where is the search performed ?
11:59 kados the opensearch portal search?
11:59 kados it searches several z3950 targets including a couple of Koha boxes
12:01 pierrick a couple of Koha boxes running Zebra (am I right so far)
12:01 paul not necessary : koha 2.2 has a basic z3950 server
12:02 pierrick I thought 2.2 had only a z39.50 client
12:02 kados I think some are using the basic z3950 server, some are zebra indexes
12:02 kados pierrick: I wrote a very basic z3950 server it's been in use at NPL for several years
12:02 pierrick OK, I understand
12:02 kados pierrick: in fact, it was the first perl program I ever wrote :-)
12:02 pierrick :-)
12:02 kados pierrick: so it's quite poorly written IMO :-)
12:03 kados but gets the job done :-)
12:03 pierrick kados: did you need to improve your perl coding style?
12:04 kados quite a bit :-)
12:04 kados in fact, I have no training as a programmer
12:04 kados (which may be obvious :-))
12:04 pierrick really ?
12:04 pierrick I don't know if it's obvious, I don't know the code enough
12:04 kados my background is in Classical languages (Greek and Latin) and Linguistics :-)
12:05 pierrick linguistics, like Larry...
12:05 kados hehe ... yea, I guess that's true
12:05 kados guess perl's a good choice for me then :-)
12:05 pierrick Perl is a great choice for many things!
12:06 pierrick is paul a programmer basically ?
12:07 paul yep.
12:07 paul but not a perl fan at the beginning.
12:07 paul now we are good friends.
12:07 pierrick reading your code, I suppose you know C
12:07 paul me ?
12:07 pierrick am I wrong ?
12:08 pierrick paul: yes, mister poulain :-)
12:08 paul wow, no, as everybody I learned C, but it was 15 years ago, and I never wrote C IRL.
12:08 pierrick (but I'm not sure I'm reading your code)
12:08 paul (I mean IBL : In Business Life)
12:09 pierrick my very first programming language was C
12:10 pierrick and I'm happy not to write real C programs outside of school
13:27 kados hey tim
13:27 tim hey kados
13:27 kados looks like your site registration is back up :-)
13:28 kados I tried to send you an email to let you know it was down but the message bounced :-)
13:28 tim Cool!  I was just gonna ask how it looks from your end.
13:28 kados yep, came back up this morning
13:28 kados if it didn't i was gonna call you cause I remembered you have an internal DNS
13:29 kados and might not be aware of it :-)
13:29 tim Yeah.  I didn't know the library switched credit card companies, so the information wasn't passed on to the registrar.
13:29 kados ahh ... trixy :-)
13:29 kados all good now though
13:29 tim We only have internal DNS for the intranet.  Our ISP handles the rest.
13:30 kados makes sense
13:30 tim They had it hard-coded, so we didn't have any problems from here or from home because I use the same ISP.
13:30 tim We just got a call letting us know our email isn't working.  I guess the fix hasn't made it everywhere yet.
13:31 kados I will try to email you
13:31 kados sent
13:31 kados no bounce, so that's good
13:32 paul kados, a quick question
13:32 kados paul: sure
13:32 paul I try to play with 3.0
13:32 paul but get a "    [109] Database unavailable -- v2 addinfo 'koha'"
13:32 kados yep
13:32 paul how to know which DB I have in zebra
13:32 paul (i think i have "koha", but it seems no...)
13:32 kados first check koha.conf
13:32 kados is it listed as 'koha'?
13:33 paul zebradb=localhost:2100/zebra in /etc/koha.conf
13:33 paul oups !
13:33 kados ok, so the db name is in fact zebra
13:33 paul stupid guy
13:33 kados not so stupid
13:34 kados you may still get the same error
13:34 kados here's what to do:
13:34 kados zebraix init
13:34 kados zebraidx create koha
13:34 kados zebraidx commit
13:34 kados zebrasrv localhost:2100
13:34 kados (after editing koha.conf and changing zebra to koha)
13:34 tim got the email kados
13:34 kados the critical step being' zebraidx commit'
13:35 kados tim: great, sounds like you're back up 100% ... may take a few more hours to propogate everywhere though
13:36 paul works fine, thx.
13:36 paul (now i have to populate the DB ;-) )
13:36 paul (with rebuild_zebra.pl)
13:38 kados :-)
13:53 kados paul: did you see Tumer's posts?
13:54 kados paul: it seems he's importing with iso2709
13:54 kados paul: which is _very_ fast
13:56 paul you're speaking of it's mail 03.10.06 "building zebra_db" ?
13:57 kados yes
13:58 kados it seems to me like he must have forked Koha as soon as perl-zoom was out without telling us :-)
13:58 paul wow, i missed the importance of this mail. I agree it's very interesting !
13:58 paul yes and no : he already uses a hacked koha version
13:58 paul (with holidays & news, that I have ported to cvs)
13:58 paul it's better than argentinian team : he submit what he did.
13:59 paul and he is very kind to help me when I ask
13:59 kados excellent
13:59 kados maybe could you ask him to share his zebra code?
13:59 kados so we can compare it to what we already have?
13:59 paul ok
14:32 thd kados: you should have the file now
14:32 paul i leave too.
14:32 paul (search with zebra works. I just have problems with ... encoding !!!)
14:32 thd kados: I hope I have excluded enough subfields
14:32 kados thd: got it
14:33 paul but it may be my XML build, i'll investigate tomorrow morning
14:33 paul bye american friends.
14:33 thd good bye paul
14:33 thd kados: I have a question about exclusion behaviour
14:35 kados thd: still too slow I think
14:35 kados thd: about 45 seconds to load the page
14:35 kados thd: and very slow to respond to page tabs
14:35 kados thd: ask your question
14:35 thd kados: I have assumed that if not editor and collapsed is populated the value is populated in a collapsed subfield.
14:36 kados thd: that's correct
14:37 kados thd: are you taking about -5?
14:37 thd kados: I have also assumed that if editor and collapsed is populated then the populated value appears uncollapsed.
14:37 kados if collapsed is populated the populated value will be collapsed
14:38 thd so y3es the first case could be -5
14:38 kados regardless of !editor or editor
14:38 thd kados: I think that the desirable behaviour would be as I described and you should consider changing it
14:39 kados thd: !!
14:39 kados thd: but that's how I had it the first time
14:39 kados thd: you have every option available to you
14:39 kados thd: just pick the one you want
14:39 thd so that collapsed functions differently between editor and not editor
14:39 kados thd: no ... collapsed is independent from editor !editor
14:40 kados *if (odd) { collapsed }
14:40 kados *if (even) { !collapsed }
14:40 thd kados: there is no option for this behaviour now as you are describing it
14:40 kados that's the rule
14:40 kados sure there is
14:40 kados if you want !editor and !collapsed use -6
14:41 kados if you want !editor and collapsed use -5
14:41 kados phone
14:42 thd kados: I want collapsed when the subfield is important enough to be included in the editor but not used all the time yet I want to see it open if I have filled one of those important ones with a value.
14:43 thd kados: at the same time I want to hide the legacy stuff in a forced collapse for not editor.
14:44 thd kados: so the distinction would be that not editor is a forced collapse while editor has only a conditional collapse for having no value yet.
14:45 thd kados: does that make sense?
14:45 thd kados: I think that would be rational and maximally efficient behaviour for working with the data.
14:47 thd kados: I thought of that shortly after you went to sleep but did not want to disturb you because we both needed rest.
14:50 thd kados: ping me when you are off phone
14:51 thd kados: I am back to removing subfields from editor
14:59 kados thd: it's true that distinction isn't currently possible
14:59 thd kados: did my suggestion above make sense?
14:59 kados thd: and it won't be
14:59 kados thd: it's too complex and requires too many new cases
14:59 kados wait ...
14:59 kados I'm wrong
14:59 kados it already is included
15:00 thd kados: really
15:00 kados no ... I'm wrong again :-)
15:01 thd kados: i think that is highly desirable behaviour
15:01 kados thd: we'll make sure it's in 3.0 then
15:01 kados thd: I have no more time to work on this feature
15:01 kados thd: too much else to do
15:01 thd kados: let me know when you are back from lunch
15:01 kados thd: let me know when the revised framework is ready
15:01 thd ok
16:56 thd kados: I sent an updated set of subfield exclusions
17:19 kados thd: the new one takes about 14 seconds to load
17:19 kados thd: much better
17:19 kados thd: but still quite slow :-)
17:19 kados thd: i need to get the addfield working
17:20 kados thd: maybe this weekend I'll have time
17:20 kados thd: I'll load the new file into the liblime demo
17:20 thd kados: I can still take out more from the editor
17:21 thd kados: at this point I will have to take out some of the fun fields
17:22 thd goodbye fun fields hope to see you again soon
17:25 kados thd: for the liblime demo I'm running mysql 4.018
17:25 kados thd: I had to adjust seealso to be 255 (as that's the max for that in 4.0)
17:25 kados thd: also had to remove the utf8 charset
17:27 thd kados: 255 will break the functionality derived from the large values as it had before when the content is arbitrarily truncated.
17:27 kados yea, I suspected as much
17:27 kados I'll need to upgrade mysql on that machine ... maybe tonight
17:28 thd kados: try changing the type to TEXT instead of VARCHAR(255)
17:28 kados k
19:33 thd kados: I have only a modest improvement so far.  I need to go to Brooklyn now.
19:34 kados ok, have fun :/
05:02 osmoze hello all
05:12 pierrick hi #koha
05:12 paul hello
05:14 pierrick bébé rétablit ?
05:14 paul oui, tout à fait.
05:14 paul 2 bonnes nuits ! ca va mieux...
05:15 paul + un génie qui a résolu les pb utf8, alors la, c'est le pied.
05:18 pierrick mouarf... regardes un peu : http://forum.phpwebgallery.net[…]topic.php?id=6014
05:18 pierrick 2 bonnes nuits, c'est cool ça.
05:27 pierrick bref, les problèmes d'UTF-8, c'est la mode en ce moment, et tant que toutes les applications n'y seront pas passé, les problèmes subsisteront :-/
06:49 paul pierrick ?
06:49 paul j'ai un pb sur les notices que je relis de zebra :
06:52 paul si je les enregistre dans un fichier après les avoir lues, elle semblent être en utf-8.
06:52 paul mais à l'affichage, j'ai des , à la place des lettres, pb typique d'une chaine iso8859-1 affichée en utf-8 je crois.
06:52 pierrick on retombe donc dans le même problème, mais on change de backend
06:52 paul j'essaye de remettre le binmode pour voir...
06:53 pierrick quoique... si tu enregistres en utf-8 dans le fichier, on a donc de l'utf-8 manipulé par Perl, le backend n'a plus d'importance
06:53 paul tu as la syntaxe exacte du binmode ?
06:53 paul binmode(STDOUT, ":utf8");
06:54 paul gagné :
06:54 paul si je remets le binmode :
06:54 pierrick ça marche avec le binmode ???
06:54 paul * ce qui vient de mySQL n'est plus bon
06:54 paul * ce qui vient de zebra est OK
06:54 pierrick on est mal alors !
06:54 paul http://bureau.paulpoulain.com:[…]pl?biblionumber=5
06:54 paul (test/test devrait marcher)
06:55 pierrick qu'est ce qui vient de Zebra exactement ?
06:55 osmoze question, j ai suivit l install sur windows, j ai une erreur : No such files or directory: couldn't spawn child process /koha/opac-cgi/opac-main.pl Une voie a me donner pour resoudre cela ?
06:56 paul pas de bol, hdl est absent jusque demain...
06:57 paul "riding" est plutôt "driving" en fait. Il est à Chambéry, chez un client.
06:57 paul pierrick : Search.pm > get_record
06:57 paul (faut le MARC::File::xml en v 0.82
06:57 paul )
06:58 paul mais je crois l'avoir.
06:58 paul tiens, question con : comment je peux vérifier facilement ce genre de choses ?
07:00 pierrick paul: c'est une question pour me tester ou c'est une vraie question ?
07:00 paul (non vraie question)
07:00 pierrick perl -e 'use MARC::File::XML; print MARC::File::XML->VERSION, "\n";'
07:00 paul OK, 0.82
07:01 pierrick tiens, je suis en 0.81
07:01 paul ca pourrait être utile de faire une chtite page qui récapitule tout ca dans Koha : les versions des paquetages installés.
07:01 paul Je vais voir ca tout de suite.
07:01 pierrick un peu comme versions.pl que j'ai écrit pour tester l'utf-8 ? ou la page about tout simplement
07:02 paul about.pl tout simplement !
07:02 pierrick (tu me laisses le faire ça ? c'est facile :-)
07:02 paul zyva.
07:02 pierrick (sans vouloir faire uniquement ce qui est facile)
07:03 paul je veux le commit dans 30mn ;-)
07:03 pierrick arf
07:03 paul petite suggestion :
07:03 paul misc/Install.pm contient déjà la liste de ce qui est obligatoire.
07:03 paul il faut y rajouter les paquetages suivants :
07:03 pierrick (oui, mais obsolete il me semble)
07:04 paul MARC::File::xml (0.82), MARC::Charset (0.95), Class::Accessor (0.22), LWP::Simple and XML::Simple
07:04 pierrick j'ajoute la version de Zebra aussi
07:04 paul très bonne idée.
07:04 paul affiche juste la liste avec les numéros de versions.
07:04 paul + Smart::Comments
07:05 pierrick lance le chrono
07:05 paul 11H05...
07:39 osmoze 11h39...
07:41 pierrick (ouais, ouais ça va ;-)
07:41 osmoze ;)
07:42 paul ne te presse pas, il reste 4mn.
07:42 paul ah, non, 3. bon alors presse toi un peu...
07:45 pierrick (y'a un truc que je ne sais pas faire en Perl, je consulte mes bookins...)
08:02 paul Search.pm/get_record :
08:02 paul $raw is correct utf-8
08:03 paul but $record is no more
08:03 paul the     my $record = MARC::Record->new_from_xml($raw, 'UTF-8');
08:03 paul is the origin of the problem.
08:03 paul investigating...
08:03 paul [paul@bureau head]$  perl -e 'use MARC::Charset; print MARC::Charset->VERSION, "\n";'
08:03 paul 0.95
08:03 paul [paul@bureau head]$  perl -e 'use MARC::Record; print MARC::Record->VERSION, "\n";'
08:03 paul 2.0
08:03 paul [paul@bureau head]$  perl -e 'use MARC::File::XML; print MARC::File::XML->VERSION, "\n";'
08:03 paul 0.82
08:04 pierrick Net::Z3950 does not offer VERSION...
08:04 paul shame, but not a problem : it will be useless in 3.0
08:04 paul (replaced completly by Perl-zoom)
08:18 pierrick at last, I've commited
08:19 pierrick designed to be easy to add modules
08:19 paul on head ?
08:19 pierrick of course
08:19 osmoze bye
08:19 pierrick bye bye osmoze
08:19 paul bye
08:20 paul farpait.
08:20 paul juste 2 suggestions :
08:20 paul * mettre un titre au tableau !
08:20 paul * mettre les paquetages dans l'ordre alpha
08:26 pierrick (c'est tout avant que je commit ?)
08:33 pierrick commit done
08:40 pierrick tiens c'est bizarre, je recois par mail la notification de mon second commit mais pas du premier !!!
08:49 pierrick paul: about coding presentation guidelines, I won't answer on fr.comp.lang.perl but on koha-devel. But before doing this, I'd like to read the discussions you had several months ago... Can you tell me when to search in the archives ?
08:59 pierrick this thread I suppose http://lists.nongnu.org/archiv[…]-05/msg00078.html

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