IRC log for #koha, 2006-10-25

← Previous day | Today | Next day → | Search | Index

All times shown according to UTC.

Time Nick Message
11:00 kados paul: querytype ccl2rpn
11:00 paul yo... it works...
11:00 kados paul: that will force it to use the file you specified in the -c paramater
11:00 paul & i've found a bug in yaz-doc...
11:01 because, in chapter 5, page 47, it says "queryttype type the following is supported : ... ccl for CCL search"
11:01 oups, no.
11:01 kados ccl yes, and ccl2rpn also
11:01 paul they also say "ccl2rpn for ccl2rpn conversion"
11:01 kados because you can override zebra's knowledge of ccl
11:01 for example
11:01 you can start zebra telling it where a ccl file is
11:01 paul so, zebra supports rpn only & we need to say yaz "ok soft, I give you ccl, you translate it to rpn for the server"
11:01 kados and querytype ccl uses it
11:02 and/or, you can say 'transform on the fly using ccl2rpn'
11:02 one is client-side (ccl2rpn) and one is server-side (ccl)
11:02 make sense?
11:03 paul: and guess what, CCL search = (Type-2 Query)
11:04 paul yep. but I thought zebra was speaking ccl directly. am I wrong, or it speaks only RPN & we need to "translate" ccl to rpn ?
11:05 kados it does both/either
11:05 (well, internally it only speaks RPN)
11:05 I will explain again
11:06 paul no need i think i've got it.
11:06 kados ok
11:06 so querytype ccl2rpn works for you?
11:06 paul we have to translate on client side or on server side. We have setup translation on client side
11:06 yep
11:06 kados yes
12:14 paul kados : just reintroducing facets & adding unimarc support (commit soon) in rel_3_0 : I really LOVE that feature !
12:15 kados w00t
12:15 it is a significant performance hit
12:15 and it only pulls from the first page of results
12:15 but I spoke with ID about incorporating it into Zebra
12:16 and they look for sponsorship of that feature
12:16 which may come in 2007 from me :-)
12:16 paul a quick question : authors, in unimarc, can be in 200$f, 200$g, 700$a, 701a & 702a.
12:16 is there a way to explain the fact i have different subfields ?
12:16 kados as currently implemented you can only use one subfieldcode
12:16 paul ok, that was what it seems to me.
12:16 kados I did have a version that could specify more than one
12:17 but it was buggy
12:17 so I reverted
12:17 paul I'll see what I can do on this topic
12:17 kados ok, great!
12:17 it's all about having enough time :-)
12:55 paul: have you or hdl looked at tumer's code for translating the facets labels?
12:55 but I think it's in HEAD
13:09 another cool NZ project: http://kupu.oscom.org/
13:09 I've used this editor and it's swank
15:35 cm kados: bulkmarcimport is choking because zebra isn't running:
15:36 Converting record 1 encoding from MARC8 to UTF-8
15:36 problem with Zebra at /usr/local/koha/intranet/modules/C4/Context.pm line 490.
15:36 ZOOM error 10000 "Connect failed" (addinfo: "tcp:@:9900") from diag-set 'ZOOM'root@koha:/usr/local/koha_cvs/dev_week/koha​-new-upload/koha/misc/migration_tools
15:37 it also chokes with zebra after processing a couple of records.  I can include the output of that if you want to see it.
15:37 kados hmmm
15:38 cm yup.
15:38 kados I've a feeling it's the new context.pm connection object
15:38 zconn
15:38 I have to run bulkmarcimport tomorrow for a new client
15:38 so if you can wait until then it'll be fixed for sure ;-)
15:38 cm yeah.  I tried commenting out the section that checks to see if zebra is running, but I got further errors and gave up.  :P
15:39 yeah, I can wait another day.  :)
15:39 kados cool
15:39 sorry about that, I figured you would still be going rel_2_2->dev_week
15:39 cm are you using it with or without zebra?
15:39 kados so I haven't tested bulkmarcimport
15:39 I'm using it with zebra
15:39 hey chris
15:40 how was fiji?
15:40 chris really really busy
15:40 but nice to be warm
15:40 cm i'd like to be warm right now.  our library's furnace is broken at the moment.  :/
15:40 chris stink one
15:41 cm yes, very stinky.
15:41 chris the koha server is on its way to the national library in the cook islands now .. with trained librarians, so fiji was a success :)
15:41 kados nice
15:41 cm cool.
15:53 chris hows things been in koha land?
15:57 cm hi chris.  things are going well enough.  we're aiming to migrate our first library in mid-december.
15:57 that is, if we don't move it back again.  ;)
15:57 chris woo excellent
15:58 that'll be to 2.4 eh?
15:58 cm yeah.  :)  it's been fun.  not looking forward to the training part.
15:58 yeah, 2.4
15:58 i had it running last week, but I broke it again.  :)
15:58 chris ill do the training for ya :) (thats what i just was doing in fiji hehe)
15:59 cm :)  that would be awesome.
16:00 i don't think the CCFLS would be willing to ship you here, though.  :(
16:00 chris heh yeah, probably not
16:00 cm yeah.  so how's the baby?
16:01 chris 2 months away from birth, but very kicky
16:01 cm ah, a feisty one.  :)
16:01 chris hopefully that means he/she will be tired and sleep lots once they are born .. but i suspect not
16:01 cm he he.  :)
16:36 rhariram hi kados
16:38 kados hi rhariram
16:39 rhariram did you see my latest comment on the bug
16:40 kados rhariram: yes ... quite strange
16:40 I'm not sure what's going on
16:40 rhariram: have you tried to repeat the bug on the Koha demo: http://koha.liblime.com
16:40 rhariram ok. will do that right away
16:41 was trying to browse the code
16:42 where does the code sting recbranch of for a new rec and exi
16:43 kados addbiblio is very complicated
16:43 too complicated IMO
16:43 rhariram sorry. prev should read as: where does the code branch off for a new rec and existing rec
16:43 kados I believe there's a flag for 'modify
16:44 that is set if ther ewas an oldbiblionumber passed in
16:44 but I'd have to check to be sure
16:55 rhariram yes, same issue at http://koha.liblime.com
16:56 that is: record looks good after creation but is corrupted after editing
16:58 and again corruption only in the fields defined in biblioitems table
17:07 kados very interesting
17:08 rhariram: could you append a note to the bug report and include some examples of characters that get mangled?
17:09 rhariram: that way i can do some testing to try to figure out when they are getting mangled
17:09 rhariram ok. what kind of example would you like to see.
17:10 shall i cut-paste the mysql query output ?
17:11 or shall I add a screenshot of the good and corrupt data
17:26 kados rhariram: what would be most useful would be an actual raw marc record I can import to reproduce the problem
17:27 rhariram: or if that's not possible, a text file with the properly encoded characters that get re-encoded imporperly
17:29 rhariram shall I get you the raw marc file by exporting from my koha installation ?
17:32 kados sure
17:40 rhariram ok. just attached a 1 rec mrc file with the good data
18:09 on a bit unrelated note, is there any script to convert dumpmarc.pl's output (after some manipulation)back to marc ?
18:13 is there a reverse of dumpmarc.pl ?
09:21 btoumi kados: are u around?
10:21 kados btoumi: I am now
10:22 btoumi hi kados i had some probleme with updatedatabase from rel 3 0 and i have found the solutions
10:23 kados: ty ;=)
10:25 kados : are u always around?
10:26 cm it seems like he is.  :)
10:31 hey kados, how did your bulkmarcimport go?
10:37 kados cm: well, it's happening today, and I haven't started yet ... in about an hour or so it begins ;-)
10:38 cm good luck.  let me know how it goes.
10:38 kados will do
10:38 cm thanks. :)
10:38 hdl kados: in ccl queries, can we use Material-type:BOOK ?
10:39 kados hdl: if you have an index called material-type and it contains data like 'book' :-)
10:39 or are you asking about the ':'?
10:39 hdl or does the "title" require no hyphen, a limitted number of characters
10:40 No.
10:40 kados ':' is illegal in ccl
10:40 but I parse it and turn it into '='
10:40 you can use a hyphen and I don't know of any limit on number of characters in a attribute type
10:41 less html to load on the page

← Previous day | Today | Next day → | Search | Index

koha1