IRC log for #koha, 2005-08-02

All times shown according to UTC.

Time S Nick Message
12:01 kados thd: do you have an opinion about proper indicators for 952 and 942? i guess there is no standard so leaving them blank should be ok right?
12:03 thd kados: Well then, how do you vary the circ rules set by biblioitems.itemtype for a particular item that you want to treat differently from other items for that biblio because that item is special.
12:03 kados thd: that's handled entirely by Koha using the itemtypes specified
12:04 thd kados: indicators would not really be important for 942 and 952, but if 952 was modelled on 852 it might be nice to keep them the same.
12:08 thd kados: But in MARC Koha you only are allowed one biblioitems.itemtype for all items because biblioitems.biblioitemnumber is keyed uniquely to biblio.biblionumber, which in turn is keyed uniquely to a MARC record.
12:11 thd kados: So, if one of your items needs to go to the bindry for repair or be held in a glass case because it has the author's own margin notes how do you prevent those from circulating while allowing other copies of the same biblio to circulate in MARC Koha?
12:12 kados thd: you can't
12:12 kados thd: it's a weakness of the data model
12:13 thd kados: That weakness is not present in in non-MARC Koha.
12:13 kados of course ;-)
12:13 kados thd: we'll need to design Zebra-Koha to fix that problem
12:14 thd kados: Were you not trying to address that very problem at NPL a week ago for damaged books?
12:14 kados thd: yes ... never found a solution
12:15 thd :-]
12:16 thd kados: call number search is broken at NPL.
12:16 kados thd: the same weakness also creates another problem
12:16 thd kados: What other problem?
12:16 kados thd: its impossible to track item costs per item
12:18 thd Why, there is items.replacementprice?  Is cost stored elsewhere?
12:19 kados thd: items.price
12:19 kados thd: but I'm not sure why both are there
12:19 kados thd: I think it's a katipo thing
12:20 thd my mistake, items.price is the original cost and replacementprice is what it says that may change over time.
12:22 kados thd: is that a MARC distinction?
12:23 thd items.replacementprice date is the date when the replacent price was last set so one has an idea of how valid the replacement price may be if unchanged years later.
12:25 kados thd: so price and replacementprice should both be set ... and replacementpricedate too (generated from purchase date I suppose in my automated script)
12:26 thd Yes, however, the MARC 21 equivalent of replacementpricedate is not in a MARC holdings field.  I had left it out of my posted mapping.  Replacement price should really be generated dynamically with a script that periodically queries a vendor or vendors.
12:27 thd From my revised mapping that I had not posted yet.  Map 365 $f to 952 $r for items.replacementpricedate.
12:28 thd Oops I have 2 values mapped to 952 $r
12:31 thd $u may do
12:33 kados thd: I"m pulling replacementprice from 952$9
12:38 thd kados: I think 365 is rarely used in practise and 876-878 $c is repeatable to provide for cost and replacement cost.  Ther is just no date subfield to accompany 876-878 $c unless placed in a parenthetical text note like the one specifying the type of cost.  39.99 (replcacement cost in U.S. dollars, set Jan. 10, 2001) for example.
12:40 thd kados: call number search is broken at NPL.  Could that be because MARC is not linked to bibliotems.dewey and biblioitems.subclass at NPL.  Call numbers can be found with a key word search at NPL but then match any number in the record.
12:41 kados thd: could be
12:42 thd the error message rather funny :)
12:43 thd is
12:47 thd kados: I imagine that call number search is most useful if not designed to necessarily find an exact item.  Closely related items in the classification could be found by their proximity in the classification scheme.
12:49 thd So searching for 005.43 would find that DDC class and subclass.
12:50 thd Searching for 005.4 does not also find 005.4 as a key word search for obvious reasons.
12:51 thd s/find 005.4/find 005.43/
12:51 thd s/find 005\.4/find 005\.43/
12:52 thd s/\(find 005\.4\)/\13/
13:09 thd I believe the intention of the bblioitems.dewey v. biblioitems.subclass distinction is to separate 005 from 43 in the example above.  Therefore, if call number is DDC split 082 $a or 852 $h on the '.' with the first string going to biblioitems.dewey and the second string going to biblioitems.subclass.
13:12 thd I believe the intention of the bblioitemIf call number is LCC split 050 $a or 852 $h between the letters and the numbers with the first string going to biblioitems.dewey and the second string going to biblioitems.subclass.
13:14 thd LCC example QA76.73 becomes QA for dewey and 76.73 for subclass.
13:45 kados thd: you still around?
13:45 kados thd: just ran the MARC editing script for the first time
13:46 kados thd: seems to have worked like a charm
13:59 kados thd: I'm importing into Koha now
14:07 thd kasos: yes, I am here
14:08 thd kados: does call number search work for you, with the new import?
14:19 kados thd: yep
14:47 kados thd: classification  050  a  Classification number
14:47 kados thd: dewey  942  k  dewey
14:48 kados thd: ok ... borrowers inserted ;-)
14:48 kados thd: much easier ;-)
14:48 thd oh that's why.  050 $a does not appear.
14:49 kados thd: what do you mean?
14:49 thd kados: Well, do all the records have 050 $a?
14:50 kados thd: probably not
14:50 thd kados: Maybe very few.
14:50 kados thd: could be ... what was the alternative mapping?
14:50 thd hardy any :)
14:52 kados thd: only 57 in fact ;-)
14:52 kados thd: out of 50,000 ;-)
14:52 kados thd: huh ... how did I miss that
14:53 thd Well one was 852 $k $h $i $m and another would be 082 $a $b, but 852 should be preferred for actually finding the item.
14:53 kados thd: i remember now ... I wanted them to have both LOC and Dewey available
14:53 kados thd: and 050 is LOC classification
14:54 kados thd: there are three
14:54 kados thd: classification and dewey in biblioitems
14:55 kados thd: and items.itemcallnumber
14:55 kados thd: I think I'm going to leave it as is
14:56 kados thd: because that way, the OPAC can display all three (when they exist)
14:56 kados thd: and maybe I can even automate the process of getting the LOC numbers
14:56 kados thd: do you know of any programs to generate LOC from Dewey?
14:57 thd kados: concatenate 852 $k, $h, $i, $m  along with 852 $k, 050, $a, 050 $b, 852 $m and put them bothe in the same field.
14:58 kados thd: er?
14:58 kados thd: could you explain why that's a good idea?
14:59 thd There is no perfect one to one coresponcence between DDC and LCC.  Mappings have been attempted but do not really work well enough.
14:59 thd kados: As you wrote above, had not you wanted to have them both available.
15:00 kados thd: I want them both available ... but not concatenated
15:00 thd kados: Right now there are no call numbers in the detail view.
15:00 thd kados: separate them with a space I had meant to say.
15:01 kados thd: but that's a template problem I think
15:02 thd kados: you have almost no records with a call number in biblioitems.callnumber unless you want to remap the template.
15:05 kados thd: do you mean biblioitems.classification?
15:05 kados thd: here's my mapping:
15:05 kados biblioitems.classification is LOC
15:05 thd kados: The MARC view template was obviously designed to use items.itemcallnumber while the detail and search results templates were designed to use biblioitems.classification
15:06 kados biblioitems.dewey is dewey
15:06 kados thd: right ... but that's an easy template change
15:06 kados thd: in fact, I think NPL changed it to use classification back in the day and that became the default
15:07 kados thd: look at the MARC view on a record
15:10 thd Yes, the MARC view looks nice but most people prefer the detail view or its equivalent on every system to the MARC view.
15:10 kados thd:[…]map_ddctolcc.html
15:10 kados thd: for mapping DDC to LOC
15:11 kados thd: yes I will change the default view to use items.itemcallnumber as the default callnumber
15:11 kados thd: (which makes more sense than using classification anyway)
15:12 thd kados: however, If you make that change then you loose your LCC numbers unless you append them.
15:12 kados thd: nope ... because I can just make a new column  with the LCC numbers ;-)
15:13 thd oh yes :)
15:14 thd There are several DDC to LC mappings.  OCLC probably has the best but in all of them you will notice mapping between a single number and a range of numbers.
15:16 thd A better solution is to get the LCC number from LC for each book using LCCN or ISBN in a batched z39.50 search.
15:18 thd Softcover ISBNs could be a probem in that approach if LC does not have the softcover ISBN as it sometimes does not, especially for older titles.
15:24 thd I have lots of experience running batched z39.50 searches against LC and other targets.  Using softcover ISBNs can sometimes be a significant problem without a database to match them.  There are potentially expensive commercial databases for what they have.  I wish XISBN worked better for this problem to match the manifestation or edition.
15:28 kados thd: that would be nice
15:30 kados thd: here are some problems
15:30 kados thd: items.itemcallnumber is only 30chars
15:30 kados thd: biblioitems.classification is only 25
15:31 kados thd: biblioitems.dewey is double(8,6)
15:32 kados thd: so the callnumbers are getting truncated
15:33 thd kados: there has been some recent discussion of this on koha-devel.  The suggestion is to change the format and size of the column.
15:33 thd kados: I guess you have to rerun your import :(
15:33 kados thd: well i'd have to anyway ;-)
15:34 kados thd: still waiting to hear back from the client on correct format->itemtype and location->location mappints
15:34 kados thd: they had very unnormalized records
15:34 kados some formats were locations and some locations were formats ;-)
15:35 thd kados: I can see some of that.
15:36 kados thd: we need a 'status' for items
15:38 thd kados: Having all subject subdivisions in 650 $a instead of 650 $a $x $etc. does partially fix a bothersome Koha though for linking between detailed subject headings in the detail view.
15:39 thd kados: If you write charged in and out status to the MARC records constantly you will have a risk of thrashing the system would you not?
15:40 thd kados: Well, on second thought, that is a much lower risk than joining tables for every search to check status.
15:41 kados thd: I'm not sure about the load issues yet
15:42 kados thd: we'll have to run some tests with the Zoom API to see whether it's load-effective to do transactions off the MARC data or not
15:42 kados thd: I suspect we will continue to use mysql for transactional stuff
15:42 kados thd: and maybe update the records in the background with a batch process
15:45 thd kados: you still need DDCnumbers in the call number column of course, as most detail records still have no call number.
15:45 kados thd: working on that now ;-)
15:46 kados thd: weird ... the dewey is coming out all strange like
15:46 kados thd: I wonder if I messed up the conversion somehow
15:48 kados thd: ahh ... it's the format of the field
15:48 kados thd: :-)
15:48 thd :)
15:49 thd varchar (100) is the popular suggeston for the DB on koha-devel
15:49 kados yea ... that's what I was thinking
15:49 kados for all of them I guess
15:55 kados thd: right ... well I've still got to hack on my LFS a bit today
15:55 kados thd: enough Koha for one weekend ;-)
15:57 thd kados: Wow, itemnumber seems to be a database wide autoincremented value rather than incremented for each MARC record.  Not what I had expected.
15:57 thd kados: LFS?
15:58 kados thd: Linux From Scratch ... I'm working on my own distro for use with DiscrimiNet (smaller footprint, more security, etc.)
15:59 thd kados: Debian is too dangerous :-)
15:59 kados thd: :-)
18:09 thd chris, rach, si: good morning
18:09 thd chris, rach, si: what is the usage of biblioitems.dewey and biblioitems.subclass, as compared to biblioitems.classification?
18:10 chris you can use them how you want really
18:10 chris all three are available to the templates
18:11 chris most ppl just chuck everything in dewey
18:11 thd chris: what was the original design?
18:11 chris but it could be book 809.18 f
18:12 chris prefix dewey suffix basically
18:13 thd chris: where is prefix?
18:13 chris classification,
18:13 chris they dont have to work together, and they werent designed so that they had 2
18:14 chris just if you only have numbers in dewey, its much faster to search
18:14 chris and order
18:14 chris right back to work i go
18:14 thd prefix meaning 'J' for juvenile for example?
18:15 chris whatever you want
18:15 chris i dont see it as koha's job to enforce a library's cataloguing rules
18:15 chris thats up to them
18:16 chris and now im really back to work, ill be able to chat again after work
18:16 chris (monday is busy)
18:16 thd can I find 809.18 if I search for 809.1*
18:18 thd chris: sorry, go back to work and I will accost you on Tuesday :)
02:37 paul salut hdl, tu m'appelles qd tu es disponible ST P?
02:37 hdl hi
07:54 kados hi all
07:55 paul 'morning joshua
07:55 kados paul: I've got a few questions
07:55 kados I've tried importing records with to see if the leader is imported as well
07:55 kados it seems it's not (unless I'm looking in the wrong places)
07:56 paul what did you do ?
07:56 kados well first I added tag 000 with subfield @
07:56 paul i did not test bulkmarcimport yet, but manual adding of a biblio works well for me.
07:56 paul ok, that's what is required.
07:57 kados Tab 0; not repeatable; not mandatory
07:57 paul ok, well done too.
07:57 kados then imported with
07:57 paul is your uptodate ? for example :
07:57 kados and now I look for the leader from the MARC view or from the 'edit biblio' screen
07:57 paul line 372, you should have :
07:57 paul &MARCaddsubfield($dbh,$bibid,'000',''​,$fieldcount+1,'',1,$record->leader);
07:58 kados yes ... up-to-date
07:59 paul mmm... try : select count(*) from marc_subfield_table where tag='00'
07:59 paul (should give you >0)
07:59 kados yes I tried that
07:59 kados the 000 tags are there
07:59 kados but they appear blank
07:59 kados let me do a select
08:00 paul you mean they contains space(24) ?
08:01 kados mysql> select distinct(subfieldvalue) from marc_subfield_table where tag='000' limit 0,20;
08:02 kados +--------------------------+
08:02 kados | subfieldvalue            |
08:02 kados +--------------------------+
08:02 kados |                          |
08:02 kados | 00982nam  2200265Ka 4500 |
08:02 kados | 01458cam  2200325Ia 45x0 |
08:02 kados | NULL                     |
08:02 kados +--------------------------+
08:02 paul does this sound OK ?
08:02 paul is it what you were expecting ?
08:03 kados I think they are correct
08:03 kados but only 2 records out of 50,000 have leaders ;-)
08:03 kados maybe it's a problem with the records
08:03 kados :-)
08:04 kados I will check with
08:04 paul try to enter a record manually & see if it's correctly stored & shown after creation
08:04 paul (go to edit biblio again)
08:05 paul maybe i've forgotten to commit something, but i don't think so.
08:05 paul maybe my code is still buggy ;-)
08:06 kados using I see leaders with every record
08:06 kados LDR 00575nam  2200145 a 4500
08:06 kados LDR 00408nas  2200157 u 4500
08:06 kados LDR 00435nas  2200181 u 4500
08:06 kados some examples
08:07 paul ou mean your 50 000 records have a leader, and only 2 of them are imported with the leader ?
08:07 paul wow !!! big bug !!!
08:07 kados yes ... ;-)
08:07 paul mmm, i will investigate my code, it seems to be highly buggy ;-)
08:08 kados so ... I've some more questions too
08:08 kados if you have time
08:08 paul ok, i always have time for you.
08:09 kados the default mappings for things like biblionumber, biblioitemnumber and itemnumber ... and for reserves, restricted, etc.
08:09 kados you mention them in the help file
08:09 kados but I haven't found them anywhere ... I'd like to map them to the MARC record as I think it will be helpful when we move to zebra
08:11 kados so should they be mapped? how can we get them into the MARC record?
08:11 kados (and will Koha update them when the biblio/item information changes?)
08:13 paul i'm not sure i understand your question
08:13 kados sorry ... it wasn't very clear
08:13 paul biblionumber & biblioitemnumber are mapped in 090 by default
08:14 kados ok ... I am confused about the way MARC Koha handles holdings
08:15 kados it seems to me that _one_ 952 is mapped to _one_ Koha items table field
08:15 paul you're right on this
08:15 kados and _one_ 942 is mapped to _one_ biblioitems table field
08:16 kados hmmm ... I think I just answered my own question ;-)
08:16 kados ok nevermind that one ;-)
08:17 paul yes but no.
08:17 paul in Katipo DB model, you can have 1 biblio => X biblioitems => Y items
08:17 paul in MARC model, you just have biblio & items.
08:17 kados right
08:17 paul so, you have only 1 biblio = 1 biblioitems.
08:18 kados right ... it's too bad ... with Zebra we should investigate to see if we can do a FRBRization of our data
08:18 kados if it's worth it
08:18 kados but that's another topic ;-_
08:18 kados ok ... i've some questions about Koha fields
08:19 kados in biblio: serial and copyrightdate
08:19 kados do you know of good MARC mappings for these two?
08:20 kados in items:
08:20 kados classification  050  a  Classification number
08:20 kados dewey  942  k  dewey
08:21 kados so the client can use both LOC and Dewey (they are currently using Dewey but want to switch to LOC)
08:21 kados but I'm not sure what to use 'subclass' for
08:21 kados also volumedate, volumeddescc, and marc
08:22 kados sorry ... the above are biblioitems
08:22 kados in items I have several questions:
08:23 paul the "marc" field is useless.
08:23 paul it will be used in 3.0 to store the raw marc record, in iso2709, probably
08:23 paul subclass was done to have dewey class & subclass i think
08:24 paul volumedate & volumedesc are probably for serials.
08:24 paul I use them for serials anyway ;-)
08:24 kados could you explain how to use them for serials?
08:24 kados do you use biblio 'serial'?
08:25 kados the records i am importing contain many serials and periodicals
08:25 kados but I"m not sure how to distinguish between serials/periodicals and books
08:25 kados periodicals are journals
08:25 paul how do you call a book that is published in 4 parts ?
08:25 kados hmmm ...
08:25 kados multi-volume maybe
08:26 kados not a serial
08:26 paul I use volumedate & volume desc to store infos about multi-volume books
08:29 paul kados : what do you think of my new dynamic menus in rel_2_2 ?
08:29 kados paul: i was just going to tell you that I love them ;-)
08:29 kados very nice
08:30 paul quite simple & improves a lot useability imho
08:30 kados definitely
08:30 kados is 'booksellerid' linked to acquisitions?
08:31 kados is 'multivolumepart' related to volumedate and volumedesc?
08:31 kados 'price' 'replacementprice' and 'replacementpricedate' ???
08:32 kados what is 'stack' and 'binding'
08:32 paul all this comes from katipo, you should ask them for a correct answer ;-)
08:32 paul about booksellerid :
08:32 kados are we using statuses in items: 'itemlost' 'wthdrawn' 'restricted' 'paidfor' etc.?
08:33 paul it seems it contains the booksellerid. but it's not handled in the MARC record. 100% sure.
08:33 paul about items statuses : i'm sure that itemlost is used.
08:33 kados IIRC some vendors put their id into the MARC record (in the 001 field?)
08:34 paul wthdrawn probably too, all other are unknown to me.
08:34 paul we speak of acquisition module here, not MARC record origin.
08:35 kados right ... but I've been reading about EDI and it seems that some book vendors put their ID into the MARC record for identification purposes for the ILS
08:35 paul OK :
08:35 paul if ($iteminformation->{'notforloan'} > 0) {
08:35 paul $issuingimpossible{NOT_FOR_LOAN} = 1;
08:35 paul }
08:35 paul if ($iteminformation->{'wthdrawn'} == 1) {
08:35 paul $issuingimpossible{WTHDRAWN} = 1;
08:35 paul }
08:35 paul if ($iteminformation->{'restricted'} == 1) {
08:35 paul $issuingimpossible{RESTRICTED} = 1;
08:35 paul }
08:36 kados I see ...
08:36 kados what do you think about having user-defined statuses?
08:36 kados so the statuese are not hard-coded
08:36 paul we already have ;-)
08:36 kados we do?
08:36 paul yes, of course... I explain :
08:37 paul define an authorized value called "status"
08:37 paul enter :
08:37 paul 0 available
08:37 paul 1- not for loan
08:37 paul 2- being repaired
08:37 paul 3- withdrawn
08:37 paul ...
08:37 paul go to the field that is mapped to notforloan.
08:37 paul and use authorized value 'status' for it
08:38 paul now you have a status list that means :
08:38 paul 0- book can be issued
08:38 paul >0 = book can't be issued
08:38 kados right ... what about
08:38 kados book is in transit
08:38 kados (to a branch)
08:38 paul don't NPL use a "transit" branch for this ?
08:39 kados can staff easily set the status?
08:39 kados no ... NPL doesn't use transfers in Koha
08:39 paul I thought you were doing :
08:39 paul * transfert to "transit branch"
08:40 paul * when the book arrives in dest branch, "return" it, Koha automatically set the branch to "dest branch" for the item
08:40 paul or is it katipo way of doing transfer ?
08:41 kados it's katipo's way of doing it ;-)
08:41 kados I just checked: we dont' have a 'transit' branch defined
08:41 paul how do you do transfers ?
08:41 kados a book comes in with a flag that it's on reserve for someone at another branch
08:42 kados staff acknowleges the flag and puts the book in a box for that branch
08:42 kados when the box arrives, the librarian returns all the items
08:43 kados paul: so where do I define the authorized value called 'status'?
08:43 paul and what is wrong with this way of doing things ?
08:43 kados paul: I"m just learning frameworks
08:43 kados paul: there is no way to inform the patron of the status of the item
08:43 paul Koha >> parameters >> authorised values
08:43 kados paul: (whether it has arrived or not)
08:44 kados paul: (owen could explain this better than me)
08:44 paul (in default templates, dunno with owen's one)
08:44 paul then, New category >> type what you want as category name, enter your list
08:44 kados great ... thanks!
08:45 kados paul: I think that's all my questions for now ;-)
08:45 kados paul: I did have an idea yesterday
08:45 kados paul: that I would like to get your opinion on
08:46 kados paul: when we move to zebra non-MARC libraries will have a problem
08:46 kados paul: because the search in Zebra will be to a MARC database
08:46 kados paul: so my idea is
08:46 kados paul: instead of using MARC for non-MARC libraries ... we should use XML
08:46 kados paul: since Zebra can search XML documents just fine
08:47 kados paul: what do you think?
08:47 paul good idea.
08:47 paul but don't see how complex it will be to code ;-)
08:57 Sylvain hi all !
08:57 paul hello amigo !
08:57 paul bienvenue ici
08:57 Sylvain merci :) Alors la planete koha va bien en ces mois d'été ?
08:58 paul ca va, ca bosse bien.
08:58 paul les idées pour la v3.0 sont bien posées maintenant, hdl et moi travaillons sur des nouveautés intéressantes pour la 2.2.3
08:58 paul (2.2.4, pardon)
08:58 paul dont la gestion des leaders, réclamée par tout le monde...
08:59 Sylvain des leaders ?
08:59 kados paul: how do you handle internal development for a client (like customized templates) do you have your own internal CVS? if so, how do you sync it with sourceforge?
09:00 paul i don't have customized templates.
09:00 kados ahh :-)
09:00 paul I convince all my customers of the problems due to local templates
09:00 paul ;)
09:00 kados good plan ;-)
09:00 paul (i often speak of owen in fact ;-) )
09:00 kados hehe ... yes it is a real problem
09:00 kados very hard to sync everything
09:00 paul with css opac stylesheet, we can do really nice things on opac.
09:01 paul (like what I did for bibliotheque de sociologie)
09:01 paul that's enough for most libraries.
09:19 thd kados: If non-MARC is in XML how will it enjoy the benefit of copy cataloguing?
09:22 kados thd: map MARC->XML fields
09:23 thd kados: Then why not just keep the MARC record and MAP that to the Koha columns?
09:28 thd kados: I am just supposing that it would be much easier to define a simple set of MARC fields and subfields to map than it would be to map to an additonal format, such as XML.
09:36 thd kados: Koha should be able to export and otherwise make use of a variety of XML formats at some point; but I Imagine it would be easier to support feature development if MARC was underlying, along with Koha tables, than it would be to add new features using MARC, XML, and Koha tables.
09:36 paul kados : i just tested bulkmarcimport on 100 biblios, and it works fine (including leader)
09:37 paul are you sure you are not doing anything to your datas before calling NEWnewbiblio ?
09:37 paul mmm... I think I understand : do you use standard bulkmarcimport ?
09:37 paul if yes, then you should add :
09:38 paul $newRecord->leader($record->leader());
09:38 paul just after
09:38 paul    my $newRecord = MARC::Record->new();
09:38 thd paul: What is nonstandard bulkmarkimport?
09:38 paul thd : all libraries I migrate want to do some cleaning on datas before importing.
09:39 paul so I have almost as many versions of bulkmarcimport as I have migrated libraries
09:39 paul (for example, the one i'm migrating atm has ISBD punctuations signs inside the MARC record. And we need to remove them)
09:40 thd paul: These nonstandard versions are not in CVS are they?
09:40 paul no, of course.
09:40 paul because they are library dependant
09:40 paul (& some libraries don't want to show to the world their cataloguing rules as they are awful :-D )
09:42 thd paul: So, you had fixed the code on a nonstandard version that you have withoout adding it to the standaerd bulkmarcimport for distribution?
09:42 paul right.
09:42 paul if kados confirms, i commit the fix immediatly ;-)
09:43 thd paul: When you had been needing to convert MARC 21 to UNIMARC for a library, had you investigated USEMARCON?
09:44 paul no.
09:44 paul I have some home made tools
09:51 thd paul: USEMARCON would have most likely accomplished the task, but the free version only has rules defined for UKMARC included.  The free version is from the British Library.  I have failed to find a well publicised source for other rule sets to download directly.  BnF seems to have been using USEMARCON to convert between INTERMARC and UNIMARC.
09:51 paul looking at usemarcon website...
09:52 paul I remember now why i didn't use this tool :
09:52 paul I have no windows ;-)
09:52 paul mmm... new versions seems to work on linux.
09:53 thd paul: The GUI version also runs on UNIX.
09:54 thd paul: The British Library version runs from the command line.
09:55 thd paul:[…]ic/usemarcon.html
09:56 kados paul: I'm testing leader import no
09:56 paul I now, i'm compiling it atm
09:56 kados paul: now ...
09:57 thd paul: What would it take for MARC::Record to support INTERMARC or some other MARCs?
09:57 paul my opinion is that MARC::Record is poorly named.
09:57 paul it should be called iso2709::Record
09:57 paul because it supports any MARC flavour without problem
09:58 paul the only problem is the $record->title() & record->author() that sends MARC21 title/author field
09:58 paul that's why Koha never uses it.
09:58 kados paul: the leader is working
09:59 kados paul: in I see:
09:59 kados LDR 00502nam  2200145 a 4500
09:59 paul hurray !
09:59 kados and in catalog I see:
09:59 thd paul: What does Koah use instead?
09:59 kados 00502nam  2200145 a 4500
09:59 kados so it's perfect ;-)
09:59 paul thd : the marc2koha links
09:59 kados thanks paul!
09:59 paul ie : the field mapped to biblio.title
09:59 paul (try the plugin now)
09:59 paul to see if it's working well too.
09:59 paul it should
10:00 thd kados: hooray, thank you paul for 000 :)
10:00 paul (shouldn't we have a debugging session in less than 1mn ?)
10:00 kados thd: if you're trying to map between different flavours of MARC Zebra does this automatically
10:00 kados paul: yes ...
10:00 kados now in fact ;-)
10:02 kados I count 60 bugs:
10:02 paul (I squashed 5 today)
10:03 kados so bug 893:  Can't get past biblio info page in simple acquisitions
10:04 paul 1st question : who's there ?
10:04 hdl I
10:04 kados paul: good point ;-)
10:05 paul i'm afraid there is only hdl, me, you, and the logbot (& thd, but he's not a coder, so won't endorse any bug ;-) )
10:05 kados :-)
10:05 kados should we post-pone?
10:05 owen Oh, bug squashing today?  I forgot
10:05 kados owen finally has internet ;-)
10:06 kados ok ... the note on bug 893 refers to bug 1002 ... my mistake
10:07 paul the bug 893 is for chris, as it's MARC=OFF related.
10:07 kados yep
10:07 thd paul: hopefully, one day I will squash the bugs that I introduce :)
10:07 paul lol
10:08 paul bug 472 is the next on the list. I'll take care of it ASAP
10:08 kados great
10:08 paul (was trying few minuts ago, but not simple)
10:08 paul 993 / 994 / 991 are MARC=OFF related too.
10:08 paul 670 requires owen's opinion...
10:08 kados yep ... so assigned to chris then?
10:09 paul yep
10:10 kados paul: to test 670 do we only need a few lines from and
10:10 kados paul: if so I will add them after the meeting, owen can test and report back
10:10 owen to fix bug 670 we need to do some work on
10:11 owen Right now the template can't show whether an item has reached its destination branch
10:11 paul (that's why katipo uses a "transit" branch...)
10:11 paul (going back 2 hours ago ;-)
10:11 kados owen: could you expand just a bit on the transfer process that we're using? maybe we need to change it.
10:12 owen It's pretty simple, really: we just need to grab two pieces of information in the script, and compare them
10:12 owen holdingbranch (where the book is now) and pickup branch (where the reserve says the book should go)
10:12 owen If they're the same, the 'waiting' status should be shown: the book has arrived
10:12 kados right ... that does sound simple
10:13 owen If they're different, the book must still be in transit, and thus the patron shouldn't be told the book is "waiting for them"
10:13 kados paul: and as far as I know it wouldn't break anything else in Koha (should work with existing transit code)
10:13 paul right.
10:13 paul who checks ?
10:13 kados owen: I'll look at it after the meeting
10:13 paul ok
10:13 owen Maybe it's simpler to put this information in the reserves table somehow, but to fix the bug now the above method works
10:13 kados yep
10:14 owen (I just don't know if it's the most efficient method)
10:14 kados 975: serials : renewal of a subscription
10:14 kados (is this next on everyone's list?)
10:14 paul nope
10:14 paul it's 864 for me
10:15 kados (IIRC the order is not consistant because the Severity level is normal)
10:15 hdl ok for me.
10:15 kados so if it's ok with everyone we'll go in my order
10:16 paul no prob for me
10:16 hdl ok.
10:16 paul I'm the reporter AND the owner of this one. to be fixed, for sure.
10:16 kados ok
10:16 paul nothing to add ;-)
10:16 kados 790: opac search on 'the' gives internal server error
10:17 kados is this a problem with stopwords?
10:17 owen Although doesn't show waiting reserves at the moment, it should.  So that page needs to be updated as well.
10:18 paul 790 : right, it's a stopwords problem.
10:18 kados paul handling it?
10:18 paul didn't owen add something to warn the library when there is no stopword ?
10:18 owen No, I just talked about the idea
10:18 owen I don't know how to implement it
10:19 paul otherwise, I can handle it easily, creating a 1 line hash & warning the user.
10:19 paul ok, it's for me.
10:19 kados 1001: Issuing Rules Error
10:19 kados this one comes with a patch ;-)
10:20 paul (on phone)
10:20 hdl Is this patch to be acknowledged by everyone ?
10:20 kados I don't know that code very well
10:20 kados so someone who does should take a look
10:20 kados hdl: are you familiar with it?
10:21 kados hdl: no rules ;-)
10:21 hdl I tested some rules.
10:21 hdl issuing rules.
10:21 paul (back)
10:21 hdl I can try and cope with it.
10:22 kados hdl: great ... can you reassign it to yourself?
10:22 kados 992: Suggestions doesn't check wether a biblio exists
10:22 hdl okµ
10:23 kados really more of a feature request I think
10:23 kados I haven't really played very much with the suggestions functions
10:23 kados paul: what do you think?
10:24 paul half a bug & half a feature request imho
10:24 paul I reassign it to me.
10:24 kados 585: Using 'document.forms[0]' notation prevents use of other for
10:25 kados an old one
10:25 paul right.
10:25 paul should be fixed during template 3.0 rewritting imho
10:25 kados right
10:25 kados maybe assign it to owen?
10:25 kados owen: how do you feel about that ;-)
10:26 paul steve tonnesen won't fix it ;-)
10:26 kados heh
10:26 kados 925: Currency search doesn't appear to do anything
10:26 paul (did anyone know what he's doing now ? as he never said goodbye if i don't mind)
10:26 owen yeah, assign it to me and I'll call for help if needed
10:26 kados (no idea ... i wrote him a few months ago but didn't get a response)
10:27 paul 925 : ok, i remove the search box immediatly
10:27 kados 943: pb accentuations LDAP <=> KOHA
10:27 kados assigned to paul already
10:27 paul yep.
10:28 kados 958: Field "notes" becomes empty at receipt
10:28 paul it's a EMN specific one. not critical
10:28 kados 958 is partially in French ;-)
10:29 kados so automatically gets assigned to Paul ;-)
10:29 paul it has been reported by EMN too
10:29 paul still some investigations to do. not sure to understand the bug completly
10:30 kados 933: Cancel process results in untemplated dialogue
10:30 kados paul: (NPL doesn't use ordering so 958 is hard for owen or i to diagnose ... maybe we should ask katipo as they have fixed some stuff lately but haven't committed to CVS)
10:31 kados I can look at 933
10:31 paul ok for both.
10:31 paul hdl commited some stuff on this subject too
10:31 paul (acquistion)
10:31 kados 849: Error when generating barcodes
10:32 paul seems stephen has found the culprit : the version of PDF::API2
10:33 kados yep
10:33 kados it's a real problem
10:33 kados caused by use of debian woody ;-)
10:33 paul I have an old version, so it works. I should probably upgrade :-D
10:33 paul any news from argentina about the missing module in argentina code ?
10:33 kados not yet
10:33 kados I will bug them today
10:34 kados I'll update 849 with new info
10:34 kados 1009: Not all items mapped to MARC
10:35 kados paul: I noticed that running rebuildnonmarc does not work correctly
10:35 kados paul: is this the same bug?
10:35 paul probably not.
10:35 paul this bugreport is 100% right
10:36 kados paul: ok ... I'll report the problem with rebuildnonmarc also
10:36 paul I use show columns where the old api hardcoded the fields. I'll fix it too.
10:36 kados ok
10:36 kados 1002: Error in bulkmarcimport indicates possible permissions probl
10:36 kados assigned to MJ
10:37 kados paul: could you look at 1002?
10:38 paul I don't know what to do with it.
10:38 kados paul: there are a few other installation bugs I haven't reported yet
10:38 kados paul: related to updatedatabase
10:39 kados paul: I'll report them later today
10:39 paul ok
10:39 kados 1005: mouse's click doesn't work in submit button
10:39 kados owen: can you take care of this?
10:39 owen I'll take a look.  I've never seen this problem before.
10:40 kados 680: Intranet doesn't allow linking of subject headings
10:40 paul mmm...
10:41 paul i've fixed it on my computer today.
10:41 paul commiting & changing status.
10:41 kados 185: Barcode not auto-calculated when autobarcode is on
10:41 kados very old ;-)
10:41 thd paul: 1002 the kohaadmin user should be created with more than just use permissions on the Koha DB in the DB creation script.
10:42 paul thd : it's a mysql version problem.
10:42 paul on 4.0, it is created correctly.
10:42 paul on 4.1 not.
10:42 kados 185 comes with a fix ... I wonder if Chris committed the fix
10:42 kados paul: right
10:43 kados paul: so the install should at least warn of the possible error
10:43 kados paul: best case scenerio it would check version and add the password correctly
10:43 paul ok
10:43 kados 984: Bug in installer appears to not correctly escape password
10:44 paul (note that the fix is buggy in theory)
10:44 paul (as you don't lock the database, 2 cataloguers could get the same number)
10:44 kados paul: are you talking about 185?
10:44 paul (but only one will be able to save, as the 2nd one will get a "barcode already exist")
10:44 paul yep
10:44 kados paul: ok
10:45 paul in old code, there is a lot of multiple access like this one.
10:45 paul I try to get rid of them when I meet them.
10:45 paul (using lock table or auto_increment where applicable)
10:45 kados paul: could you send me an example of where you fixed it ... I will try to emulate your work
10:46 paul 984 is for mj
10:46 kados paul: ok ... I'll reassign it to him
10:46 kados 836: $auto_install error
10:47 kados I've not seen that error
10:47 kados lately anyway ;-)
10:47 kados so it is probably fixed eh?
10:48 paul you are the reporter aren't you ;-)
10:48 kados yes ;-)
10:48 kados 682: Move HTML out of
10:48 paul I think it's un-fixable.
10:48 paul almost.
10:49 kados I'll take a look
10:49 paul (those templates are already complex. if we add a logic to handle every type of fields possible, then the template will become a real pain...)
10:49 paul (to write & to debug)
10:49 kados true
10:49 paul so, my opinion is that it should be kept as it.
10:50 kados ok
10:50 kados can you mark as WONTFIX?
10:50 paul unless owen says no, we must find a solution to this
10:50 paul owen ?
10:50 kados 1008: Long MARC breaks KOHA
10:50 kados (in the meantime)
10:51 paul good bugreport.
10:51 paul and the solution is OK for me. no reason to have a tinyint.
10:52 kados great ... will you commit the changes (to updatedatabase?? and install??)
10:52 paul yes. I assigne the bug to myself, with a comment
10:52 kados ok
10:53 kados 961: Error message, when deleting an authority
10:53 thd paul: why is 682 unfixable?
10:54 paul thd :
10:54 paul in the MARC editor, any html field (=MARC subfield) can have at least 7 forms :
10:54 paul * standard input
10:54 paul * authorized value list
10:54 paul * itemtype list
10:54 paul * branch list
10:54 paul * plugin
10:54 paul * authority link
10:55 paul * textarea, for long fields
10:55 hdl kados : I assign 961 to me.
10:55 kados hdl: great ... thanks
10:55 kados 426: Reserving a book (with a reserve fee) results in a hostile m
10:55 paul if you want to handle all those 7 forms completly in the template, then you must pass a lot of parameters to it, and get a too complex template & code.
10:55 paul kados : about 961
10:55 paul could you take care of
10:55 kados thd: it creates some performance issues as well (I think)
10:55 paul Is it possible to create  "authority" in the components on the Koha Bug Tracker ?
10:56 kados paul: I've already noted it and will do it after the meeting
10:56 paul patrons are a pain ;-)
10:56 kados heh
10:56 paul ok
10:57 owen about bug 682: I agree with paul.  Too much trouble to fix.  Maybe down the road when it's time for a re-write of the whole script we can tackle it then
10:57 thd Of course, that is clear.
10:57 kados at NPL too
10:58 kados I'll assign it to chris
10:58 paul hehe...
10:58 kados :-)
10:58 paul in french we say "missing persons are always wrong"
10:58 kados 989: Can't add authority (javascript error)
10:58 paul ;-)
10:58 kados hehe
10:58 kados owen: still a problem?
10:58 paul 989 is fixed.
10:58 kados cool
10:59 kados paul close it out?
10:59 paul or someone would have said it
10:59 kados 864: Subject search results not limited, show all results
11:00 owen No more subject search, eh?
11:00 kados 864 brings up problems with subject searching in general
11:00 paul 864 is a MARC=off bug i think
11:00 kados the old Koha used to keep track of the subject headings
11:00 paul so it's for chris ;-)
11:00 owen Oh yeah, I see now
11:00 kados so when you did a search for 'dogs' it would give you categories to choose from to refine the search
11:01 kados paul: is such functionality planned for Koha 3.0?
11:01 kados paul: (I mean, is that the way your libraries want it to work as well?)
11:01 paul kados : ask the release manager ;-)
11:01 kados paul: hehe
11:01 kados paul: ok ... so that's what I would like to see happen
11:02 kados paul: but I'm not sure if everyone agrees that that's the right way to do it
11:02 paul atm, there is the ... on the right of the subject search.
11:02 paul that shows all subjects containing "dog".
11:03 thd there is much thought that should go into subject searching.
11:03 kados ok
11:03 kados thd: agreed
11:03 kados getting back to 864
11:03 hdl you can use dictionary to get all dog related subject ;)
11:04 paul owen :
11:04 paul[…]ha/
11:04 kados actually, I think we could nab LOC subject headings
11:04 paul click on the ... after subject line
11:04 paul you can do a "dictionnary" search here.
11:04 owen Oh, right.
11:05 owen I was confused by the "..." in your sentance.
11:05 kados paul: wow ... I didn't even know that existed
11:05 owen It's fairly new
11:05 paul it's something hdl added in 2.2.2 iirc
11:05 paul (or in 2.2.3 ?)
11:05 thd paul: when will 6XX $a $x $etc. link to the appropriate match in instead of just linking 6XX $a?
11:05 kados very nice
11:05 hdl 2.2.2b
11:06 paul thd : not soon. some pitfalls i don't know yet how to solve
11:06 kados ok ... so now getting back to 864 :-)
11:07 kados someone needs to go through and fix all the marc=off bugs
11:07 thd paul: It is something every good ILS does easily now.
11:07 kados thd: Koha can already do this
11:07 kados thd: using the 'seealso' fields
11:08 paul thd : not sure I was speaking of the same thing you were.
11:09 kados paul: about marc=off ... I think I saw a fix in the MARCURLS code you recently fixed ;-)
11:09 kados paul: am I correct?
11:09 paul yes, but they work only when MARC=ON, in non marc view.
11:09 paul in non marc view, when MARC=off, there should be no changes
11:10 paul (& in fact i was just improving a little some of your code)
11:10 kados (thanks :-))
11:10 kados (if you want to really improve it you will get the 'titles' to show up as the anchor for the url :-))
11:11 kados so what are we doing about 864?
11:11 paul ???
11:11 kados (the urls have titles: 'Table of Contents' and 'Summary' in the MARC record)
11:11 kados (I discovered them after I added MARCURLS subroutine)
11:12 paul we let 864 open, for sure. and wait for someone to fix it.
11:12 paul maybe i could give it a try, but not sure, as i'm not sure to be able to reproduce it.
11:12 thd paul, kados: I had been speaking of linking to what you had referred to as 'builded' subjects in the OPAC: so that subjecta--subsujectx links to subsubjecta-subsubjectx and not merely displaying as subsubjecta only in the OPAC.
11:12 kados (each URL is actually a set 'title' and 'url' ... I can't remember the subfields now but they are linked somehow in the record ... at the time I couldn't figure out a way to have Koha link them)
11:13 kados thd: lets talk about this after the bug meeting
11:13 kados thd: I'm interested to know what you mean
11:13 kados paul: how much longer can we work on bugs today?
11:13 paul 15 minuts.
11:13 thd s/subsubjecta only/subjecta only/
11:13 kados paul: ok ... great
11:14 kados 810: Cannot set renewal limit, renewals always set to 0
11:14 paul thd : that works in MARC view. lot of problems in non MARC view. And I was speaking of the same thing as you were.
11:15 paul mmm... 810 seems fixed to me.
11:15 kados owen: do you agree?
11:16 owen As far as I know.
11:16 paul ok, joshua, you can FIXED if ;-)
11:16 kados paul: ok ;-)
11:17 kados 995: Group Maintenace Looks like it should do something but it do
11:17 paul => chris
11:17 paul no hesitation ;-)
11:17 kados :-)
11:17 kados 885: Non-ISO8859-1 charset support broken in CSS theme in 2.2
11:18 kados paul: is 885 valid?
11:18 paul probably
11:19 paul acli is the author & bugreporter
11:19 paul so he knows what he is speaking of ;-)
11:19 kados ok ... but 885 was quite a while ago
11:19 kados I was wondering if it has been fixed since then
11:19 owen Nothing has changed to affect that bug, I think
11:19 paul not that I no.
11:19 owen I'm not even sure what would have to be chagned.
11:19 paul (know)
11:19 kados owen: do you know of a way to fix it?
11:20 kados ahh
11:20 paul tmpl_process3 must be modified to handle charset management.
11:20 kados so acli needs to do that I suppose
11:20 paul and the only change done recently has been done by me, and it was related to copying css & non translatable files.
11:21 paul yes. but we didn't hear from him since a long time
11:21 kados owen: can you look into it?
11:21 paul maybe a mail could be a good idea.
11:21 owen I can try
11:21 paul ok, guy, 1 or 2 bugs, & i'll have to leave.
11:21 kados right
11:22 kados 892: data in a closed order appear differently than data before c
11:22 paul I have their DB, i'll look.
11:22 kados great
11:22 kados let's close the bug meeting
11:22 kados I've got some stuff to do too ;-)
11:23 paul ok, have nice day americans.
11:23 thd good day paul
11:23 paul frenchies here will try to have a good night ;-)
11:23 paul (6:30PM for me thd)
11:24 kados paul_away: read you soon
11:24 owen I don't suppose anyone can confirm bug 1005?  I don't see any problem here.
11:25 thd kados: did you understand what I was referring to about subject linking in the details view?
11:26 kados thd: no ... could you explain it?
11:26 thd kados: the library you had just migrated works around this problem with nonstandard data :)
11:27 kados thd: could you explain the problem first?
11:27 kados thd: I'm not sure I grok it
11:27 kados owen: I'd mark it as WORKSFORME
11:27 kados owen: if it works for you ;-)
11:28 thd kados: If a subject is divided with subdivisions only the first subdivision appears and is lined in the detail view.
11:28 kados thd: I need an example
11:28 thd kados: example methodology--science
11:30 thd kados: Koha shows Methodology only in the detail view and is linked to all subjects with methodology including Methodology--sociology
11:31 kados thd: so in MARC how is this data stored?
11:32 thd kados: 650 $aMethodology$xScience
11:32 thd kados: In your latest migration they have 650 $aMethodology--Science
11:33 kados thd: I need to look at the MARC standard for subjects ... in fact, it was on my list for this past weekend because I wanted to make sure my mappings for my client were correct
11:33 kados thd: but I didn't get to it
11:33 thd kados: So they have MARC wrong but the linking works right because of that
11:33 thd in Koha
11:35 thd kados: there are circ 250,00 LC subject headings but you can get more exact headings by compounding them.
11:35 kados thd: what do you think about importing LC subjects into Koha as an 'authority'
11:36 kados thd: and then mapping subject headings for records that are imported
11:36 thd kados: circa 250,000 becomes many more and very precise.
11:37 thd kados: That is the way it should be done but that is for 3.0 perhaps
11:37 kados thd: of course
11:37 kados thd: I'm not sure how to do it ... but I think it's the right way to proceed
11:37 kados thd: maybe even have DDC and LCC as options for import
11:37 kados thd: and update them monthly
11:38 thd kados: I had a suggestion for paul about this but he had to leave.
11:38 kados thd: I still don't quite understand the problem you're mentioning (in the meantime)
11:38 kados thd: could you give me a real word example from NPL's database?
11:38 kados
11:38 thd ok. one moment
11:42 hdl paul_away kados I had powersupply failure.
11:42 hdl Sorry.
11:43 kados hdl: no problem ... the bugfix meeting is over now :-)
11:43 kados hdl: paul has gone to dinner with his lovely family ;-)
11:43 hdl Did I miss anything ?
11:43 kados hdl: no
11:43 kados hdl: we closed shortly after you left
11:44 hdl is there an agenda for Website meeting
11:44 kados hdl: not yet
11:45 kados hdl: I'll post one soon
11:45 kados hdl: (I'm a bit behind ;-))
11:46 hdl no problem.
11:46 hdl Going to take care of my lovely chid.
11:46 kados hdl: :-)
11:47 thd kados: I am having difficulty finding an example at NPL suggesteing a deeper problem with records there
11:48 thd kados: I will compare two sources for the same record to compare the subject headings
11:55 thd kados: http://search.athenscounty.lib[…]
11:56 kados thd: ok ... I'm looking at it
11:57 kados thd: so what's the problem?
11:57 thd kados: look what happens to Science--History--Greece and Science--History--Rome in the detail view
11:57 kados thd: it becomes Science and Science, ancient
11:57 kados thd: is that what you mean?
11:57 kados thd: the History. and Greece are missing?
11:58 thd kados: yes, and every OPAC that I know gets this right.
11:58 kados thd: that should be an easy fix
11:58 kados thd: all we have to do is tell the template to concatenate those entries
11:58 thd kados: paul said that he did not know how.
11:59 thd kasos: In searching for an example I noticed something odd.

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