IRC log for #koha, 2007-02-14

All times shown according to UTC.

Time S Nick Message
13:43 kados paul: you here?
13:43 paul yep
13:43 paul (5:45PM in france)
13:43 kados paul: I was wondering if you could send me the init scripts you have
13:43 kados paul: I will compare them to ours and commit something
13:43 paul ???
13:43 kados paul: for zebra
13:43 kados paul: to daemonize zebra, etc.
13:44 paul ah, ok. I don't have anything related to daemonization of zebra
13:44 paul you should ask sanop...
13:44 paul (although we will need something soon, as IPT will migrate to 3.0 in february probably...
13:44 paul )
13:46 kados anyone from sanop here?
13:47 paul btoumi seems here on #koha-fr, but don't answer
13:47 kados I'll write a note to koha-devel
13:47 paul are u back to US ?
13:47 kados not uet
13:47 kados yet even
13:48 kados I get back on Thursday morning EST
13:50 kados mail sent
13:52 kados paul: I just updated rel_2 and I get:
13:52 kados Can't locate C4/Calendar/ in @INC
13:52 kados paul: is it a new dependency?
13:54 kados paul: it seems it was added to
13:54 kados paul: but there is no C4/Calendar/
13:55 kados :-)
13:55 rch kados: cvs -d ?
13:56 kados hmmm
13:56 kados guess I didn't
13:57 _paul kados :
13:57 _paul toins worked this morning on
13:57 _paul he created the CVS branch & synched everything from 2.2.7 considering only bugfixes.
13:57 _paul he tells me it's done.
13:57 kados great
13:57 _paul so you can give it a try if you want
13:58 _paul 2.2.8 + improvements = 2.2.9
13:59 kados I agree 2.2.8 is better
13:59 kados do we want to try to build translations too?
14:00 kados for 2.2.8? as one of the enhancements?
14:00 _paul that should be easy, as almost nothing has changed in templates
14:00 kados I have asked several translators to test
14:01 kados so I hope soon it will be ready
14:01 kados _paul: there are many languages there now, have a look when you get a chance :-)
14:01 kados I think there is a bug for generating the po file for the intranet templates though
14:01 kados at least the last time I tried it
14:02 _paul is it a 0.2 version of kartouche ?
14:02 _paul or a newer one ?
14:02 _paul because kartouche seems an abandonned project
14:03 kados it's .2 + many improvements
14:03 kados I have contacted the maintainer
14:03 kados he says he no longer works on it but would 'hand over the reigns' to me :-)
14:03 _paul I tried too, but got no answers (because I added some features)
14:03 kados ahh, did you?
14:03 kados maybe you can send me your code and I will incorporate into my version
14:03 _paul yep. like searching a term
14:04 kados then once I have cvs access I will contribute them back
14:04 _paul instead of just entering an line #
14:04 _paul kados, btoumi is here now  ;-)
14:04 btoumi hi kados
14:04 kados hi btoumi
14:04 kados btoumi: did you see my question to koha-devel?
14:05 kados btoumi: maybe you could respond with some details about how you manage the zebra server's process
14:05 kados btoumi: if you have some time?
14:05 btoumi have some time?
14:05 btoumi i have some time
14:05 btoumi just take minute to read ur message
14:06 kados _paul: that sounds like a great feature
14:06 btoumi ok kados u want to know some details to create zebra process
14:07 kados paul: you could just send me tar/gz of whole kartouche, I can dig :-)
14:07 kados btoumi: it would be useful to know what techniques others are using
14:07 kados btoumi: because there are many possible ways :-)
14:08 btoumi ok
14:08 btoumi i create a file who start stop and restart service zebra in mandriva
14:08 kados btoumi: can you send to koha-devel, along with attachment of file?
14:08 btoumi yes
14:09 kados thx, i think it will be useful for other zebra users too
14:09 btoumi i think so
14:12 btoumi kados can i join attcahed file for kohadevel?
14:13 kados btoumi: yep
14:13 btoumi ok i do it
14:27 btoumi kados i do something with small explanation
14:35 btoumi kados it's ok
14:39 btoumi bye all
14:40 kados btoumi++
17:24 cht morning, chris. ru there?
17:59 cht i discovered an interesting bug with 2.2.6 and mysql 5 on debian sarge this week...
17:59 cht (i know this is an unsupported platform)
18:00 cht on the issuing rules page, when you enter something like 1,7,3 in one or more of the "fines" boxes...
18:01 cht and save, what you get back when the page repaints is something like "1.000000,3".
18:04 rch hmm.
18:04 rch just verified this on updated rel_2_2 etch/mysql5.
18:05 cht the simple solution is to edit the template to allow a longer maximum field length for the fines field
18:05 cht this is what i did
18:05 cht in other words, it is an HTML problem
18:06 cht or a number formatting problem in mysql 5
18:42 rch cht: committed fix for rel_2_@
18:43 rch 2_2, even
18:46 cht cool
18:46 cht my first open source bug fix  ;)
18:50 rch :) mysql5 changed the way it stores decimal data types.  so if the column is defined as 6 decimal places, that's your output.
18:50 rch did you format in or just template change/
18:54 cht just the simple template change
18:57 cht if yuo got too carried away with formatting, you would run into currency localization issues for some currencies
18:58 cht perhaps currency localization ought to be a future koha feature?
18:58 cht the six decimal places does look pretty silly for dollars and cents
19:00 rch who uses more than 2 decimal places?
19:00 mason lawyers :)
19:01 rch :)
19:02 rch seriously...  is there a reason to have >2 places?
19:02 mason hmm, i cant thik of any...
21:17 cht i guess there really is not a good reason for more, unless there is some currency somewhere that would require it
23:43 chris sounds like your marc framework is busted to me
23:49 mason try  select * from biblio where title like  '%welyk%';
23:51 rickw question: does MySQL 5 allow 'contains' in where clauses?
23:52 chris no, and neither does 4
23:52 rickw select * from biblio where title contains 'welyk';
23:52 rickw ERROR 1064 (42000): You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'contains 'welyk'' at line 1
23:52 rickw the above where is clause is generated by the koha search catalogue page (!)
23:52 kados select * from biblio where title like "welyk%";
23:52 chris what version of koha are you running?
23:52 rickw plus: koha does not report an eror .. it silently returns an empty serach set
23:52 chris cos ive not seen that
23:52 rickw 2.2.7
23:52 kados that's a weird search ...
23:53 chris yeah, i cant find anywhere in the code that would make that search
23:53 rickw kados: yes, that is the syntax i would have xpected!
23:53 kados wouldn't expect that to be anywhere in koha ... searches are done now on marc_word in  2.2.7
23:53 kados not biblio
23:53 chris yes
23:53 rickw i just did a koha.upgrade
23:53 rickw from 2.0.0 to 2.2.7
23:53 kados ahh
23:53 mason ah, your reading from the web output aye
23:53 rickw successfully, as far as the databse itself goes
23:53 kados me has to run
23:53 kados ciao
23:54 rickw yes i am reading the web annotation
23:54 rickw perhaps it "lies"
23:54 chris i reckon the marc framework
23:54 mason its 'human friendy', ie not real sql
23:54 chris yeah
23:54 rickw btw: none of searches seem to work
23:54 chris probably you have no stopwords definied
23:54 chris and/or have no marc framework
23:54 rickw hmmmm ... so you don't suspect the sql itself. goof.
23:54 chris or no marc data
23:55 chris no, unless that error is in the actual error logs
23:55 rickw irma is reading this right now too
23:55 rickw and we are going "hmmmmmmmmmmmmm?" to ourslves
23:55 chris you have checked the error log?
23:56 rickw Chris it's Irma the mARC data held in a different d/b or attached to those records that have MARC data?
23:56 chris well you wont have any marc data yet
23:57 chris actually i lie you will
23:57 chris its in the same db
23:57 rickw oly few ees bib records have MARC data
23:57 chris they all do
23:58 chris 2.0.0 stores marc data for you
23:58 rickw true
23:58 chris id check the stopwords
23:58 rickw so back to Rick's question of no display
23:58 rickw where?
23:59 rickw in 2.2.7 ,,,in the preferences? are they automated?
23:59 chris go to system administration
23:59 chris there should be a link to stop words
23:59 chris if none are defined, add one ... THE  oughta do
00:00 mason yep
00:01 chris also the best place to look for errors is the error log usually in /usr/local/koha/log/
00:01 dewey Hmm.  No matches for that, chris.
00:01 rickw ok will do but remember Chris this is the unaltered Koha 2.2.7. download - no changes have been made to stop words by us as yet
00:02 chris right but you upgraded it right
00:02 chris from a 2.0.0 so its not a stock 2.2.7 install
00:02 chris as part of the install it makes stop words, im not sure it does as part of an upgrade
00:02 chris so id check
00:03 rickw we have 3 stop words in there already: a, an,the
00:04 chris k, then the next thing is to try a search, and then look at the error log
00:04 chris oh its mysql 5 right, you have made the mysql 5 changes to the search right?
00:04 chris anyway the error log will show if you havent right away :)
00:05 mason i'd turn the mysql logging on too
00:05 rickw rick here
00:05 rickw yes, i hgave made the chanegs to source code (two files Search*.pm) for mysql 5
00:05 rickw h5 just checking the error log now
00:05 chris i wouldnt yet, the error log will show what the error log is easy
00:05 rickw heh?
00:05 chris that was too mason
00:05 rickw too random!
00:06 chris mysql logging is going to just confuse things
00:06 mason Keep it stupid
00:06 rickw i am already stoopid!
00:07 rickw okay
00:07 rickw searched for 'uranium'
00:07 rickw no errors in log file
00:07 rickw no serach results
00:07 rickw :(
00:07 chris what is in the log file?
00:07 rickw well, there ar e hundresd of thousands of records in it ... ;)
00:07 rickw but nothing came out now
00:07 rickw i am doing a tail
00:08 chris on the koha-error_log ?
00:08 mason tail -f ?
00:08 rickw tail -f  yup
00:08 mason cool
00:08 chris and you are trying the search in the librarian interface?
00:08 rickw yup
00:08 rickw well, intranet, catalogue serach page
00:09 chris apache 1.3 or apache 2
00:09 rickw sorry, should have been specifiec!
00:09 rickw apacehe 2
00:09 chris ok, anything in /var/log/apache2/error_log
00:09 rickw my url: http://koha227:8080/cgi-bin/ko[…].pl?type=intranet
00:09 chris sometimes apache2 is weird
00:09 rickw um .. is apache2 writing TWO error logs for theone virtual host?
00:09 rickw i'll check, but i doubt it
00:10 chris no it has its main error log, then one per virtual host
00:10 rickw okay
00:10 chris sometimes errors that you would think would show up in the virtual host one, show up in the main one
00:10 rickw there is debug stuff (Q : select blah bklha) in apache error.log
00:10 chris right
00:10 chris see that oughta be in the koha-error_log
00:10 rickw (that is because the code is calling something like Apache.reuqest.write() or similar)
00:11 rickw "my" variable $thiscount masks earlier declaration in same scope at /usr/local/koha/intranet/cgi-b​in/calyx/ line 183.
00:11 chris i doubt it.. or not intentionaly its not
00:11 rickw is in error.log
00:11 chris right thats just a warn
00:11 mason yep
00:11 chris so the debug has the query, what happens if you run that query in mysql
00:12 rickw mysql> select distinct m1.bibid from biblio,biblioitems,marc_biblio,marc_word as m1 where biblio.biblionumber=marc_biblio.biblionumber and biblio.biblionumber=biblioitems.biblionumber and m1.bibid=marc_biblio.bibid and (m1.word  like 'mining') order by biblio.title ASC;
00:12 rickw Empty set (0.00 sec)
00:12 rickw tell me .. was a different query being used in koha 2.0.0
00:12 chris probably
00:13 rickw the like clause looks sus ... should be like '%mining%' perhaps?
00:13 chris nope
00:13 rickw why use like?
00:13 chris cos you can wildcard it
00:13 chris if you had gone mining*
00:13 chris in your search
00:13 rickw can I in the user inerface, on the serach page
00:13 rickw i.e. use % and _ ???
00:13 chris * you can use
00:13 rickw ah, okay
00:14 rickw i guess most ppl unnerstand *
00:14 chris yep
00:14 chris what does
00:14 rickw so the mystery is: why are our 2.2.7 seraches failing
00:14 chris select * from marc_word where word like 'mining';
00:14 chris return
00:14 chris anything ?
00:14 rickw select count(*) from marc_word;
00:14 rickw +----------+
00:14 rickw | count(*) |
00:14 rickw +----------+
00:14 rickw |        0 |
00:14 rickw +----------+
00:14 rickw hahahaha
00:15 rickw um, should be perhaps run a script to "MARC-size" everything?
00:15 chris theres your problem right there
00:15 chris select count(*) from biblio;
00:15 rickw (I have no idea what MARC is yet, or what this is all about ... i am jsut a humble programmer)
00:15 chris anything there?
00:15 rickw biblio is fine, 16644 records
00:15 chris hmmm
00:15 mason  ?
00:16 chris no mason
00:16 rickw i have never run that.
00:16 rickw should i ??
00:16 chris hang on
00:16 chris do you have a copy of the 2.0.0 database?
00:16 rickw yes i do
00:16 rickw but not in mysql right now
00:16 chris can we just check marc_word is empty in it too
00:16 rickw i could import it in, if required
00:16 chris can just grep it
00:17 rickw hmmm ... i can get back to you iun a few minutes and let you know, if that is alright
00:17 chris k
00:17 rickw yeah, i can vi the file :)
00:17 chris should be a bunch of insert into marc_word lines
00:17 chris if there was data in marc_word
00:18 chris if not, then we can try running
00:18 rickw marc_word not foudn in the entire dataset :(
00:18 rickw i am willoing to try that
00:18 chris k
00:18 rickw (we can always recover this entire test system if need be :)
00:18 chris as a user that can read /etc/koha.conf
00:18 chris prolly root is easiest
00:19 rickw how do you generally run your perl shell scripts?
00:19 rickw (a) ensure they are in $PATH
00:19 rickw (b) perl /full/path/ot/perl/script
00:19 rickw ?
00:19 chris like this usually
00:20 chris perl -I /usr/local/koha/intranet/modules /usr/local/koha/intranet/scripts/misc/
00:20 rickw well, i do have $PERL4LIB set for my sehlls :)
00:20 rickw it is running
00:20 rickw Processing 336
00:20 rickw etc.etc.
00:20 chris cool
00:20 rickw s/PERL4/PERL5/   (oops)
00:20 chris you could look in mysql
00:20 rickw ]as it runs?
00:20 chris and see if marc_word is being populated
00:21 chris yeah, open another shell
00:21 rickw select count(*) from marc_word;
00:21 rickw +----------+
00:21 rickw | count(*) |
00:21 rickw +----------+
00:21 rickw |    18466 |
00:21 rickw +----------+
00:21 rickw yay
00:21 chris sweet
00:21 rickw let me guess
00:21 chris select * from marc_word where word like 'mining';
00:21 rickw it is creating an index of 'words' as found in the biblio tables?!?!?
00:21 chris return anything now?
00:21 chris yeah its creating all the marc, ie taking biblio.title and making that 245a
00:22 chris basically its making skeletal marc records of all your data
00:22 chris plus an index to search over
00:22 rickw yes .. heaps of results
00:22 chris cool
00:22 rickw this is what irma expects, which is great!
00:22 chris so if you try ur search in koha now
00:23 chris do you get results
00:23 rickw yes we do
00:23 rickw thanks!!
00:23 chris cool
00:23 rickw so, is this new to 2,2,7
00:23 rickw ?
00:23 mason nice work lads
00:23 chris 2.0.0 should have had marc .. im puzzled about that
00:23 rickw i.e. is that what you or mason meant about "it now uses a marc search"
00:23 rickw well, as long as I know now
00:24 rickw you see, this is a discardable test system
00:24 chris yep
00:24 rickw and the whole thing will have to be redone when we actually do our production migration
00:24 rickw in a while
00:26 chris yeah all the 2.2.x releases search using the marc_word table
00:26 rickw okay that's great to know now
00:26 rickw perhaps it should be part of the koha.upgrade script?!?!
00:27 rickw i don't recall this particular prob on the Koha mailing list.
00:27 chris well not really, cos 2.0.0 should have marc_word already
00:27 rickw are those librarians telepathic, clairvoyant or what?
00:27 rickw (I know they secretly communiate using channels the FBI and CIA cannot monitor!)
00:27 chris you dont want to rebuild the marc tables
00:28 chris if they already exist as far more is stored in there, and you will end up clobbering it
00:28 chris so the odd thing was that your koha didnt have the marc tables already
00:29 chris ie the upgrade from 1.2.x to 2.0.x
00:29 rickw yes, i am confgused and am asking irma ll about it as I type this
00:29 chris is where the should be run
00:29 chris but if it works thats fine :)
00:30 rickw hmmm .. cannot 'find' an author
00:31 rickw oops,
00:31 rickw that script is till running
00:31 rickw tell me, should i perhaps do a count(*) and all the marc tables?
00:31 rickw or is that script going to populate all the marc-related tables?
00:31 rickw irma now ....
00:33 rickw just clarifying...When a record is imported from another catalogue it's vis MARC protocol - when we have catalogued an item, ne MARC record was created or do I understand that a MARC record is created automatically anyway for each item catalogued?
00:33 mason hiya, its created automatically
00:34 rickw even the content of the notes field? mapped to which tag?
00:35 mason theres a bunch of koha to marc mapping stuff defined by efault
00:36 mason ive u go to sys-params, then to the 'Links Koha - MARC DB' url
00:36 mason s/ive/if
00:37 rickw ok...makes sense ...last configuration of koha for me is over 2 years ago with Alexandre and Katipo's help.
00:37 mason you can set what koha fields you want automatically mapped to which marc tags (and visa versa)
00:38 rickw thanks Mason. signing out. Thanks a lot Irma and Rick
00:39 mason try this link ->  http://koha227:8080/cgi-bin/ko[…]
00:39 mason and this -> http://koha227:8080/cgi-bin/ko[…]pl?frameworkcode=
00:40 rickw thanks will dooooooooo
00:40 mason coolies :)
00:40 rickw we owe you a couple of coolies :)
06:43 btoumi hi all

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