Time  Nick     Message
10:00 thd      All the control fields look alike when they just have a XXX number in front :)
09:58 thd      Leader stops at position 23 :) but there are similar complexities.
09:58 kados    hdl: here
09:56 hdl      kados ?
09:54 thd      Oops s/leader/008
09:50 thd      kados: You have language for example.  Language, leader positions 35-37, is mandatory and is the most reliable place to find the primary language of the work.  Language code field 41 is used when applicable to supplement positions 35-37.  Language note field 546 is optional and cannot be relied upon to appear even in national level records.
09:49 kados    thd: and that's what the koha-zebra list is for
09:49 kados    thd: and whether it's useful
09:49 kados    thd: limit to material type to start with ... but I think the best thing to do would be to ask the Index Data folks if it's possible
09:45 thd      kados: That is a a large task unless you want to limit it to material type to start with.
09:44 kados    thd: and ask whether Zebra could do one?
09:44 kados    yep ... could you post to koha-zebra a explaination of how a leader search would work
09:42 thd      kados: 2 separate values from the leader in combination
09:42 kados    I'm fairly certain Zebra could do that
09:42 kados    yep ... not too hard to to
09:42 thd      kados: material type can only reliably be determined from the leader and requires interpreting the encoded value of 2 positions together for very common cases.
09:41 kados    and koha.leader=xxx could define materials or sth
09:40 kados    yea so we could define a new namespace: koha.*
09:40 thd      kados: I am asking about what tools would expose the maximum information in a MARC record for when that is useful.  Especially information like what you had asked about, material type.
09:40 kados    thd: yep ... it's fully extensible
09:40 kados    have folks seen: http://www.google.com/ig
09:36 thd      kados: Does CQL allow extension for searching on custom indexes as a new named search parameter?
09:32 kados    erp ... search on the Index
09:32 kados    3. search on the indexer
09:32 kados    2. run the indexer
09:32 kados    1. define searchable fields and mappings, etc.
09:32 kados    thd: it's a three stage proces:
09:32 kados    thd: you'd have to setup the indexer to build an index on specific fields
09:31 thd      kados: Is there a feature of Zebra that allows searching an index of any subfield?
09:24 kados    if you can find Peter Binkley's Blog then I can't help you ;-)
09:23 kados    http://blog.a9.com/blog/ maybe?
09:22 thd      kados: your message mensions a blog without giving an exact name or URL.
09:21 kados    thd: yea ;-)
09:21 thd      kados: and past version 1.0 :)
09:20 kados    (standing on the shoulders of giants and all that)
09:20 kados    so in that way it's mature
09:20 kados    thd: it's new as far as implementation ... it was created as a response to the pitfalls of Z39.50 Bib-1 and the limitations of Google-like syntax
09:17 thd      kados: where is the blog?
09:17 thd      kados: :) usually they are not the same
09:16 thd      kados: now that I have read the message, CQL is new and mature?
09:11 kados    might have info relevant to your email ... I'm not sure
09:11 thd      kados: no I have not looked since late last night
09:10 kados    thd: about Searching, etc.
09:10 kados    thd: did you see my latest post to koha-devel?
09:10 kados    great
09:10 thd      kados: yes ther is much comment about that at great length in my almost finished email too you
09:09 kados    it defines material types too right?
09:09 kados    paul, thd it might actually be nice to use the leader for semantic distinctions
09:09 paul     I think yaz too, but it works poorly under windows
09:07 thd      What do MS Windows users of Koha generally use for a Z39.50 client?
09:05 paul     and it's not really a bug (from a tech pov), as it's un-coded : what is coded works !
09:05 paul     the only thing is that I know the problem exists.
09:05 thd      paul: You could say that it is symmetrical already if there is no bug :)
09:04 thd      :-)
09:04 paul     yes (at least if there is no bug ;-) )
09:03 thd      paul: Will MARC import and export will be symmetrical at last in 2.2.4?
09:03 thd      paul: If systems used the leader and other control fields in the OPAC etc., their importance would be obvious.  The contents is encoded so using the contents requires extra processing and imagination about what you can do with them.
08:59 thd      paul: The leader is at least mandatory for valid records it is just that systems usually ignore the contents and so it seems unimportant.
08:57 thd      :)
08:57 thd      paul: The leader is the most imortant field.  All the other data is lost without their leader?
08:56 paul     rach don't know what to do today : stay or leave ;-)
08:55 kados    hehe
08:55 paul     (0.01USD guarantee, of course)
08:55 kados    leader isn't a big deal either
08:55 kados    cool
08:55 paul     guaranteed.
08:55 paul     no loss except for leader.
08:55 kados    yes
08:55 paul     you mean with 2.2.x ?
08:55 kados    there might be data loss
08:55 kados    and every time they edit a record
08:55 kados    there might be data loss
08:54 kados    is if I import a client's MARC records into Koha
08:54 kados    one thing I always worry about
08:54 paul     some wont work anymore & could not be used anymore
08:54 kados    (I'm not sure exactly how it's done now)
08:54 kados    paul: will the current MARC tools in Koha be rewritten to use it?
08:54 paul     right.
08:53 kados    paul: when we get the Zoom API it will hopefully simplify the management of MARC records
08:52 thd      :)
08:52 paul     (september)
08:52 paul     2.2.4
08:51 thd      paul: When might leader importation be coming?
08:49 paul     ;-)
08:49 paul     that shows which subfields are used but "ignored" in framework & which are not "ignored" but never filled
08:49 thd      paul: OK that helps me too understand better.
08:49 paul     bulkmarcimport imports everything, and you can tune your framework after, using misc/migration_tools/check_marc_definition.pl
08:48 paul     when you bulkmarcimport, you are on the way to migrate your datas. So your framework may be not finished.
08:48 thd      paul: In either case they do not appear in biblio detail.
08:48 paul     it's due to a weakness of Koha that I consider as a feature finally :
08:47 thd      paul: If I import a record with bulkmarkimport I can import 001 and 008 even if they are not defined in the framework.  Yet if I import with z39.50 daemon I cannot have 001 or 008 without their inclusion in the Koha MARC framework.
08:44 paul     when you search "anywhere", you really search "anywhere". But when you look at biblio detail, you see only what you put in the framework !
08:44 thd      paul: your answer is logical but let me explain my confusion.
08:43 thd      paul: I am confused by your answer.
08:42 paul     yes but no : if we do like this, then you have data that you can search on, but not see them ! quite strange behaviour don't you agree ?
08:42 thd      paul: Why does some importation such as bulkmarkimport read whatever fields are in the record while others such as z39.50 and MARC import from within the application only read fields included in the Koha MARC framework?  I would have expected that they all would have called the same routine for the same function at some point in processing.
08:37 paul     not yet available
08:37 paul     management of the leader is on the way
08:36 thd      paul: What can be done to have the leader imported as field 000 in bulkmarkimport.pl?  Every other field is imported using bulkmarkimport.pl even if the field is not in the Koha MARC framework.
08:32 thd      paul: yes I would have guessed that but then I am always finding the use of negative numbers for boolean values in some applications instead of just 0 and 1.
08:29 paul     (yes item is lost for example)
08:29 paul     0 being "not", and anything else being "yes"
08:29 paul     not sure but it's probably 0/1
08:29 thd      paul: Do you kow what are the appropriate values for the integer (1) boolean columns in the items table such as items.itemlost or items.withdrawn?
08:27 paul     seems unused.
08:27 thd      paul: Were you able to discover anything about items.stack and items.binding?  I had grepped the source myself a couple of days ago without finding any obvious answer.
07:45 kados    paul: no ... power back to normal now
07:45 kados    paul: hehe
07:42 paul     à la question de 10H54 ? je ne sais pas
07:42 hdl      paul : quelle réponse à osmoze ?
07:01 paul     (no more power probkels ?
07:00 paul     kados : calculate yourself which date is in 8 months & 4 days (beginning of 2nd quarter)
06:39 kados    paul: is the logogle.com email a joke?
03:54 osmoze   paul, quand je cherche un abonnement dans l intra, et que je clique sur details, dans n° parus, il me met que ceux en attente, et pas ce que j ai deja, es ce normal ?
03:28 osmoze   yes
03:27 hdl      osmoze : have you created supplier
03:27 osmoze   en fait, il y a quelque chose que je comprend pas....Quand je lui dis par exemple que le 174 est arrive dans bulletin, dois-je ajouter un exemplaire ?
03:20 osmoze   bon, je cherche un peu ^^
03:20 osmoze   j ai regarder, mais en fait, j arrive pas a faire rentrer un abonnement, je crée une notice normalement, ensuite j ajoute un abonnement, mais il doit y avoir une couffe dans ma config quand je rentre l abo car apres, ca me le prend en compte mais a l opac, il me dit qu il y en a pas
03:18 paul     le bouton help ?
03:18 osmoze   paul, y a t il quelque part un how to pour les periodiques ? J ai un peu de mal a faire fonctionner
03:02 thd      paul: Do you kow what are the appropriate values for the integer (1) boolean columns in the items table?
03:01 paul     (/me will look in a few minuts)
03:01 paul     nope, sorry
03:01 thd      paul: Do you know the function of items.stack and items.binding?
02:58 osmoze   merci :)
02:57 paul     jj/mm/aaaa
02:57 osmoze   paul, dans le module de gestion des periodique, on nous demande une date de debut, si je met par exempe juin, il me le prend pas, sous quelle forme doit elle etre ?
02:54 osmoze   thanks paul, i think it's my logo who was not in a good format :)
02:53 paul     osmoze : you must type an address. Complete (ie, with http://...)
02:44 osmoze   purhups it's my logo ...
02:42 osmoze   with a url, there is nothing....But i search ^^
02:39 thd      osmoze: I had imagined it was a URL but have never tested the result.
02:38 thd      osmoze: Cannot you determine the solution with a simple experiment?
02:36 osmoze   do you know how to type address in parameter for logo ? it's the real path into the server or is the web adress ?
02:31 thd      chris: What is the function of items.stack and items.binding?
02:22 chris    it means her irc client has connected again, before the old connection has ping timedout
02:17 thd      rach: What does it mean when Rachel becomes rach?
02:13 thd-away Rachel: What does it mean when rach becomes Rachel?
02:10 thd-away What is the function of items.stack and items.binding?
02:09 thd-away osmoze: hello
01:54 osmoze   hello
16:26 rach     inside pages should be fluid
16:26 rach     default position is fluid - but if you need the homepage to be fixed width that is acceptable as long as you can justify it :-)
16:22 shaun    koha.org
16:19 rach     for what?
16:19 rach     umm I need more context
16:13 shaun    rach: fixed-width, fluid or max-width?
14:19 kados    OK ... I'm off to solve what I hope is the last of the power problems at APL: a possible bad UPS
14:18 kados    shaun: I wasn't
12:29 shaun    kados: are you aware that the Katipo folk have decided to use opencms?