IRC log for #koha, 2005-08-24

All times shown according to UTC.

Time S Nick Message
12:01 thd osmoze: diacritical marks are the marks for characters that the Americans left out of the standard for computing when they defined ASCII to include 'e' but not 'é'.
12:03 thd osmoze: My question is whether 'é' for example is stored as one byte or two in your records.
12:08 hdl hi thd
12:12 hdl inside the database, it is latin1_swedish for mysql
12:15 thd hello hdl
12:17 thd hdl: It would still be possible to store multibyte character sets as latin1_swedish.  They would simply not be indexed correctly.
12:23 thd hdl: The cataloguers at NPL overwrite the multibyte MARC-8 character set entities for characters with diacritical marks when they have been using properly encoded Library of Congress records for cataloguing.
12:25 thd hdl: NPL replaces the multibyte representations of 'é' as two bytes with the UTF-7 'é' as one byte.
12:26 thd hdl: Do the cataloguers where you are do anything similar to copy catalogued records from BnF for example?
12:26 osmoze how, i discover the nelsonville website
12:27 osmoze how adding pics of book in koha ?
12:27 osmoze there is a lot of idea for my library
12:28 osmoze a great website :)
12:29 thd osmoze: is the OPAC
12:30 osmoze yes, but some book have pic
12:30 osmoze pics
12:31 thd osmoze: kados wrote the scripts for using Amazon web services to obtain book cover images and text.
12:31 osmoze hum, ok
12:31 osmoze but, if i want to scan, it's possible ?
12:32 thd osmoze: I think the feature is documented in if your installation is up to date.
12:32 osmoze ok, i ll see tomorrow :) thanks a lot
12:33 thd osmoze: You need your own developers token from Amazon and you would need to adjust the scripts for amazon,fr.
12:36 osmoze so, whe are a little library, and that's be simply to scan book and cd, because, for exemple, a lot of books are not in amazon
12:36 thd osmoze: scan would be possible.  Then you would be doing something like the latest borrower images addition.
12:36 hdl thd : excuse me.
12:36 thd yes hdl
12:37 osmoze hdl, je t inderdit de regarder toutes mes fautes d'anglais ^^
12:37 osmoze interdis
12:37 osmoze (meme en francais...)
12:37 osmoze dinner for me
12:37 osmoze bye
12:37 thd hdl: with answers to all the world's problems?
13:16 thd hello hdl
13:17 hdl striving to answer your question.
13:22 hdl BNF's z3950 server uses UNIMARC. so it should be ISO8859-15 charset or ISO646 or ISO-5426
13:23 hdl None of our cuestomers had problems with import.
13:27 thd hdl: 8859 is not defined for UNIMARC that I can see.
13:27 thd hdl: ISO 5426 is really what I am asking about
13:29 thd hdl: osmoze had 100 $a/27-29 showing use of ISO 5426 for his records.
13:31 thd hdl: If my information is correct ISO 5426 gives multibyte characters for characters with accents.
13:33 thd hdl: the first non-spacing character would be for the ISO 5426 diacritical mark and the second character would be for the accented ISO 646 letter.
13:37 thd hdl: Therefore \xC2\x65 would be 'é'
13:39 thd hdl: \xC2 is for the acute accent and \x65 is for the 'e'
13:48 thd hdl: look at Character set standardisation: migration strategies to unicode for national bibliographic databases : final technical report of the CHASE Project[…]s/cobra/chase.pdf
13:54 thd hdl: given part of what you said, I would expect that your libraries would have similar problems to NPL with characters containing diacritical marks unless some system was automatically correcting for this and converting to UTF-7.
13:57 thd hdl: what system are your libraries using for copy cataloguing.  Are they copy cataloguing outside the Koha or within the Koha record editor?
13:59 thd hdl: Is it too late for you again today? :)
14:09 hdl no.
14:10 hdl I have a baby... And that's why :)
14:11 thd hdl: If you have a baby, then you are not even allowed to sleep :)
14:11 hdl thd : he he :)
14:11 hdl But My computer is in the baby's room
14:12 thd hdl: So your baby is allowed to sleep.
14:14 thd hdl: Does that mean you cannot answer now about what system are your libraries using for copy cataloguing?  The Koha editor or otherwise?
14:15 hdl No.
14:15 hdl They use Koha Cataloguing.
14:15 hdl for SURE.
14:16 hdl But, some asked a question about accents and it seems that when searching with accents, some have problems that others don't.
14:17 thd hdl: And there is no code hidden somewhere I did not find in Koha to convert character sets?
14:17 thd hdl: what problems do you have searching for accents?
14:18 hdl Some would find search accent-sensitive, others donot.
14:18 hdl hidden code is MARC conversion.
14:19 hdl It seems that MARC::Record uses certain type of character.*
14:19 thd hdl: What type of MARC conversion?
14:20 hdl With, when you import MARC records, you can choose US-MARC or UNIMARC.
14:20 hdl And this has some kind of charset code underneath.
14:22 thd hdl: But if you start with one record format such as MARC 21 will not convert to the other UNIMARC format will it?
14:25 thd hdl: I do not quite follow what you mean by MARC::Record itself using a certain type of character.  Character encoding is supposed to be defined in the record itself such as 100 $a/26-29
14:26 thd for UNIMARC.
14:28 hdl thd : you should be right in theory.
14:28 hdl But maybe practise is different.
14:29 thd hdl: Also, Is there a Koha or MySQL version that distinguishes those libraries that have an accent sensitive search from those that do not?
14:31 thd hdl: How would you account for the accent searching anomaly?
14:32 thd hdl: An OS setup issue?
14:33 hdl It is someone who did install without assistance.
14:33 hdl we don't know what machine he has.
14:33 hdl What he did before.
14:33 hdl How he installed the stuff.
14:34 thd hdl: Which searching behaviour was obtained without assistance?
14:34 hdl accent-sensitive search.
14:36 thd hdl: Do your other installations all have non-accent sensitive search?
14:38 hdl yes.
14:39 thd hdl: Well at least the English users probably prefer non-accent sensitive search by default.
14:40 hdl And french too.
14:40 hdl :)
14:42 thd hdl: The French ones at least have the correct keyboard for constructing accents and know how to use it even if some accented characters may be more awkward to type than ASCII, ISO 646
14:42 osmoze hdl, sais tu si en france on peux mettre les photos des livres lors des recherches (au niveau des droits )
14:46 hdl thd : hopefully.
14:46 thd osmoze: Are you asking about the copyright of cover images?
14:47 hdl osmoze : Bonne question. En fait, amazon te laisse faire cela si tu fait un lien vers eux.
14:47 osmoze yes thd
14:47 hdl En ce qui concerne la possibilité de faire cela par toi-même....
14:47 hdl J'avoue que je ne sais pas.
14:48 osmoze je pense pas, mais mieux vaut partir du debut quand on a une idée :)
14:49 osmoze ma biliothecaire me dit que c est pas faisable
14:49 osmoze a cause des droits
14:49 hdl techniquement : Tu peux utiliser la solution utilisée par la NPL pour avoir les photos des lecteurs. Elle me semble fort intéressante.
14:50 osmoze des lecteurs ?
14:51 hdl Mais d'un point de vue juridique : Pourquoi ne pas poser la question sur fmd.internet
14:51 osmoze c est ?
14:51 hdl En fait, il suffit de faire un répertoire image dans les includes par exemple, ou au même niveau.
14:52 hdl Et tu enregistres les images avec identifiant.jpg
14:52 hdl Ensuite, dans les template, tu ajoute la ligne :
14:52 osmoze thd, yes, brcause the manager of my library say that it's not possible in france to have book's picture
14:54 osmoze hdl, apparement c est bien plus interessant la photo des livres que des lecteurs
14:54 thd osmoze: That would be the law everywhere that I know.
14:55 hdl <img src="/images/<!-- TMPL_VAR name="borrowernumber" -->.jpg" />
14:55 hdl Thx owen.
14:55 osmoze hdl, logique effectivement :)
14:56 thd osmoze: You need an agreement with a publishers association like the one that Amazon would have made where either you can use the images liberally at your library or you can provide a compensating link to buy the material as Amazon requires.
14:57 hdl osmoze : fmd.internet est fr.misc.droit.internet : c'est un forum français.
14:58 osmoze thd, hum...oki oki, business is business
15:02 thd hdl: I will have to experiment with directly gathering some French UNIMARC records with accents from BnF myself and see what I find encoded to explore the ISO 5426 issue further.
15:02 hdl Do you want BNF Z3950 IDs ?
15:03 thd hdl: yes that may help/
15:05 hdl Server :
15:05 hdl Port :            2211
15:05 hdl Database :  Catalogue Bn-Opale plus
15:05 hdl Database name :   TOUT
15:05 hdl User :            Z3950
15:05 hdl Password :  Z3950_BNF
15:05 hdl Element set :     F
15:05 hdl Version :   2
15:05 hdl Record syntax :   UNIMARC
15:10 thd hdl: Does BnF have other databases?
15:11 hdl No.
15:11 hdl Why ?
15:11 osmoze opale ?
15:11 hdl Do you have a problem ?
15:11 osmoze there is no database only for audio ? (cd and other) ?
15:11 hdl Is there ?
15:12 osmoze (and google)
15:12 thd hdl; LC has had multiple databases that only gradually decrease in number.
15:14 osmoze hdl, opale is the same cataloue ?
15:14 osmoze +g
15:14 thd hdl; LC has had separate databases for in process record with only cataloguing in publication information and also PREMARC for records retrospectively converted to MARC from catalogue cards..
15:21 osmoze thd-away, i found another for your test :[…]/profil_z3950.htm
16:16 osmoze hdl, peut on consulter un server z3950 en marc21 si on est en unimarc ?
16:18 osmoze ou usmarc aussi
16:44 Genji morning all.
17:17 Jo Hey Chris.
17:17 Jo are you about?
17:18 Jo left message for you on hlt
17:19 chris heya jo
17:41 Genji hey chris, had a look at my picture bit?
17:42 chris havent had a chance yet sorry genji
20:07 kados chris: you about? got a quick borrower permissions question
20:08 kados is there any way to make a 'patron category' automatically have a 'permissions' level?
20:47 Genji kados: hmm... interesting idea.....
20:48 Genji kados: you mean a flag template?
20:48 kados Genji: well ... it would be neat to be able to specify a default set of permissions based on the borrower category
20:50 Genji kados: like, the permissions listed in modify flags, in borrowers section?
20:59 kados Genji: exactly
21:03 Genji kados: the flags are just 1's and 0's in a string of int(11) so should be easy to implement your request.
21:04 Genji kados: so, you want a "modify flags" button on the borrower group add/modify page, right?
21:05 kados Genji: borrower categories: cgi-bin/koha/admin/
21:05 kados Genji: from each category ... a way to set the default permissions for borrowers in that category
21:06 Genji kados: Thats basically what i said. Forgot the term though.
21:06 kados Genji: the problem is that you'd want to make sure you didn't override the _actual_ permissions set for individual users
21:06 kados Genji: so you'd want an inheritence scheme
21:06 kados Genji: not quite as simple ;-)
21:06 Genji kados: right. so.. if the user has flags set, those flags set precidence against the catagory flags?
21:07 kados Genji: exactly
21:07 kados Genji: while you're at it ...
21:08 kados Genji: did you see my post about an arbitrary tree for branch relationships?
21:08 Genji not yet. where is it?
21:08 kados Genji: it might be nice to unify all permissions in Koha
21:08 kados GenI'll see if I can dig it up
21:11 Genji unified permissions?
21:12 kados Genji: right ... so imagine a single set of permissions for every object
21:13 kados library systems, consortiums, branches, users, items, etc.
21:13 kados and inheritence rules
21:14 kados I can't find that post btw
21:14 Genji not sure i get it.
21:14 kados yea ... I need to write up an explaination and post it to koha-devel
21:14 kados but I haven't had a chance ... been so busy
21:18 kados Genji:[…]p?msg_id=12324431
21:18 kados Genji: that's a bit of what I mean ... not everything though ;-)
21:34 Genji kados: well.. i can certainly do the borrower categories permission bit.
21:34 Genji kados: but as for the whole unified scheme... i don't understand it.
21:35 Genji ?
21:41 kados I've really got to get to bed ;-)
21:42 kados I'll try to write a message to koha-devel tomorrow explaining my ideas
21:49 thd ambrose: are you still awake?
21:50 ambrose thd: yes, but i haven't looked at my problem
21:53 Genji kados: still there?
21:55 thd ambrose: I had an interesting discussion with hdl and osmoze earlier today where it seems they have no problems with ISO 5426 character sets in their UNIMARC records even though I would have expected to see a problem for all accented characters being multibyte.
21:57 ambrose thd: if i am correct that corruption only affects windows-1252-specific bytes (i.e., 0x80-0x9f), then it is probably natural for iso character sets (other than 10646) to have no problems
21:57 thd ambrose: They seem not to have multibyte characters where the records themselves say that they should.
21:57 ambrose thd: i don't know anything about iso 5426, but both iso 2022 and 8859 character sets won't touch 0x80-0x9f
21:57 ambrose oh
21:58 thd ambrose: ISO 5426 and MARC-8 are special character sets for libraries.
22:00 thd ambrose: They are difficult to work with because they are not supported by the major character set conversion functions.
22:02 thd ambrose: owen told me today that NPL removes any double byte MARC-8 accented character it finds from copy catalogued sources and substitutes the single byte UTF-7 equivalent.
22:04 thd ambrose: For some reason the French do not have to work around that for ISO 5426 even though their problem should be the same with multibyte accented characters.
22:21 thd ambrose: Why would your corruption issue be related to Windows specific characters such as stupid quotes?
23:01 thd ambrose: I have no glyph for the drachma, \xa5 in ISO 8859-7, on my fully encoded Debian system.  It resembles the characters with no glyphs on your image.
23:05 thd ambrose: I can have Greek but I have to trade in dollars, pounds, or euros :)
23:07 thd unless I do multibyte trading, although, I suspect that I would still lack the glyph.
23:27 ambrose thd: i don't know why the corruption would be related to Windows specific characters, but that is my observation
23:31 thd ambrose: do you know anything more about your problem such as whether the character data was being manually entered or copy catalogued?
23:44 ambrose thd: unfortunately no. i should be less lazy and do my testing....
01:38 osmoze good morning koha
02:23 osmoze do you know if it's possible to count th number of borrowers by site ?
02:35 osmoze ok, i ve modified the & tmpl
03:08 osmoze bonjour hdl
03:10 hdl Salut Osmoze
03:11 osmoze sais tu si on peux rajouter du php a l interieur d une page ?
05:10 hdl osmoze : Pourquoi pas. Mais ce ne sont pas des templates standards !!!!
05:10 hdl Alors attention aux futures mises à jour.
05:10 hdl Mieux vaut réserver cela à un portail d'entrée.
06:32 osmoze hdl : en fait, je voudrais rajouter un tout petit script php dans une des pages
06:33 osmoze j ai essayer d ouvrir une balise php dans le tmpl mais c est pas ca ^^
06:33 osmoze un indice ?
08:56 hdl osmoze : je n'avais pas vu ce que tu demandais.
08:59 hdl osmoze :En fait, normalement, si tu mets simplement le script php dans la page, ça devrait marcher
09:00 hdl hi thd. I am here.
09:00 hdl osmoze :Mais pour pouvoir t'aider efficacement, il faudrait que tu me décrives ton problème
09:09 osmoze hdl>  je mets le script dans le tmpl ou le pl ? en fait, j ai rajouter les photos des borrowers, mais je me penche sur deux choses : une page pour l upload des photos avec correspondance (exemple, upload photo avec n°de carte) et ensuite, pour l affichage, je voulais passer par un script qui utilise image magick, j ai trouver les commande pour utiliser le image::imagick de perl, mais comme je connais pas perle
09:10 osmoze pour le redimensionnement
09:10 osmoze de plus, sur la page d accueil de l opac, je veux rajouter un script aussi pour la securité (script php de log d ip etc etc)
09:11 hdl As-tu activer le php sur ta machine ?
09:12 osmoze oui ^^
09:12 hdl Ne faudrait-il pas indiquer que c'est un script php pour que Apache le traite comme tel ?
09:12 osmoze j utilise d autre site dessus
09:12 osmoze (phpmyadmin par exemple )
09:12 hdl Je me pose des questions.
09:13 osmoze si, mais dans le tmpl, les deux cas : avec ouverture de balise <? ?> et sans, c est pareil
09:13 osmoze je me demande si un print"<? xxx ?>" fonctionnerai ou pas
09:15 osmoze a moins que je rajoute un :
09:15 osmoze <script language="php">
09:15 osmoze xxx</script>
09:24 hdl osmoze : Normalement, il faut au moi <?php ?>
09:25 hdl Peut-être un script language="php" serait-il plus approprié.
09:29 hdl Un chose est sûre, c'est qu'il faut que cela soit intégré au template.
09:31 osmoze hdl>  tout essayer dans le template, rien n y fait...Peut etre faut il un module perl particulier
09:40 hdl osmoze : il semblerait que le script php ne soit pas interprété.
09:41 hdl Certainement le Template n'est pas réinterprété.
09:41 hdl Mais je vais essayer de trouver pourquoi.
09:41 osmoze hdl>  
09:41 osmoze module php pour perl
09:41 osmoze j ai fait un
09:42 osmoze perl -MCPAN -e 'install "PHP"'
09:42 osmoze et apparement ca marche, maintenant je regrde l aide sur cpan pour l utilisation
09:49 osmoze bonne soirée
10:39 thd hello hdl
10:43 thd hdl: I was up very late last night looking at my code to find some simple authentication syntax error while receiving 'init rejected' messages from BNF :(
10:44 thd hdl: I will have to look at my connection problem further.
10:46 thd osmoze, hdl: however, I have a guess about why you have no ISO 5426 problems with your records.
10:47 thd hdl: As you had said yesterday hdl, BNF supports 8859-15.
10:50 thd osmoze, hdl: BNF must be using 8859-15 to transmit despite the encoding in the records claiming ISO 5426.
10:51 thd however, I have not been able to connect yet to confirm that speculation for myself.

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