IRC log for #koha, 2009-07-20

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

All times shown according to UTC.

Time Nick Message
14:08 martinmorris Hello, is there anybody who can give me a bit of advice about ERROR 114 I'm getting when searching my authorities?
15:19 hello again, won't keep being a pest (honest!) but just trying again in case there's anybody who might be able to help with my ERROR 114 problem when searching authorities
15:31 gmcharlt martinmorris: does it make any difference which sort option you use (none, heading a-z, heading z-a)?
15:31 martinmorris hi gm
15:31 no, no difference at all
15:32 the error is at the actual search stage, and comes up in the log files
15:35 gmcharlt I mean the sort option you select before you execute the search
15:35 martinmorris no, it makes no difference at all
15:35 i also get the error if i execute the search on the command line using yaz-client
15:40 gmcharlt martinmorris: please try this variant in yaz-client
15:40 @attrset Bib-1 @or @attr 1=8001 @attr 7=1 0 @and @attr 5=100 @attr 1=9000 PERSO_NAME @attr 4=6 @attr 5=1 @attr 1=8001 smith
15:40 martinmorris give me two secs
15:40 gmcharlt change 'smith' to your actual search
15:42 martinmorris unknown command: @attrset
15:42 gmcharlt sorry f @attrset...
15:42 martinmorris :)
15:43 works perfectly :)
15:43 gmcharlt ok, let's try another part of the query
15:43 f @attr 5=100 @attr 1=9000 PERSO_NAME
15:44 martinmorris just type that in with my query?
15:44 no, that doesn't work
15:44 gmcharlt no, in yaz-client, as a separate query
15:45 martinmorris right, yes that works
15:45 i do a show all afterwards and both the authorities in there come up
15:46 i'll have to read up more on what all these things mean
15:48 gmcharlt ok, the last bit to try
15:48 f @attr "1=Heading" @attr 7=1 0
15:49 martinmorris that worked with no errors, but came up with 0 hits
15:51 gmcharlt but the full f  @attrset Bib-1 @or @attr "1=Heading" @attr 7=1 0 @and @attr 5=100 @attr "1=authtype" PERSO_NAME @attr 5=1 @attr 4=1 @attr 3=2 @attr "1=Heading" smith
15:52 fails?
15:52 martinmorris hmm, no actually that has just worked
15:52 hold on
15:57 gmcharlt ok, I think I see what's going on - as a temporary measure, in C4/AuthoritiesMarc.pm change instances of '1=Heading-Main' to '1=Heading'
15:58 martinmorris aaaah!
16:00 so i need to change that somehow?
16:00 yes, i can see the difference
16:16 gmcharlt martinmorris: does it work now?
16:17 martinmorris i wasn't quite sure what to change, is there somewhere obvious in that module?
16:18 gmcharlt look around line 223
16:18 21             ##If mainentry search $a tag
16:18 222                 if (@$tags[$i] eq "mainmainentry") {
16:18 223                 $attr =" \@attr 1=Heading-Main ";
16:18 224                 }elsif (@$tags[$i] eq "mainentry") {
16:19 martinmorris and just change that to Heading?
16:19 gmcharlt yes
16:19 martinmorris so is this something that needs fixing in koha then?  or simply that i need a different version of that module?
16:20 gmcharlt it needs to be fixed in Koha - Heading-Main is defined in indexes of UNIMARC authority records but not of MARC21
16:21 martinmorris aaah :)
16:21 ok thanks very much for that, would never have got that on my own
16:21 let me try now
16:22 yep, works wonderfully, thank you so much
16:23 gmcharlt ok - please file a bug on bugs.koha.org so we don't lose track of it
16:23 martinmorris will do, thanks
16:29 do you need to know about whether i'm using DOM etc in the description?
16:34 it's bug 3429
16:34 munin 04Bug http://bugs.koha.org/cgi-bin/b[…]w_bug.cgi?id=3429 major, P5, ---, galen.charlton@liblime.com, NEW, Searching fails for MARC21 authorities with ERROR 114 in zebrasrv log
16:34 martinmorris ah, didn't know that would happen
16:34 gmcharlt that's OK, I want munin to be helpful like that
16:35 martinmorris thanks again for your help
16:35 gmcharlt you're welcome
17:05 hi jwagner, paul_p
17:05 jwagner Hi gmcharlt!
18:56 chris morning
18:59 jdavidb Morning, chris!
18:59 chris hows the weekend goin?
19:00 jdavidb Going well.  I went over to Baltimore yesterday for a big art festival there--was *very* late getting back.
19:02 chris sounds like fun :)
20:46 richard hi
20:46 chris hiya richard
20:46 richard hi chris
20:46 chris get out in the sun in the weekend?
20:47 richard did yesterday. it was good to be able to do it
20:47 chris yeah
20:55 martinmorris hi there, can anybody tell me why biblios I've deleted are still appearing in searches, even after rebuilding my zebra index?
20:55 chris nope
20:55 if you have rebuilt your index they should not be showing
20:55 martinmorris fair enough :)
20:56 when i click on the title, i get a koha error page
20:56 chris have you done a full rebuild (ie with -r)
20:56 martinmorris ah, no i haven't :)
20:56 that'll be the reason, thank you
20:56 chris (whcih deletes and rebuild the index completely)
20:56 martinmorris should i also avoid using the -w flag?
20:56 chris however if you have the cronjob running, it should be removing things as you delete
20:57 martinmorris that's something i need to set up actually, haven't quite got round to it
20:57 chris ahh that'll be the problem
20:57 when you run rebuild_zebra what optiosn are you giving it?
20:57 martinmorris well, i was using -b -w -v
20:57 but i understand now, that's OK
20:57 chris right, so you have 2 options
20:58 you either rebuild in full with -r
20:58 martinmorris yep
20:58 chris or you run with -z
20:58 which tells it to look at what is in the zebraqueue table and process that
20:58 there will be rows in there telling it to delete items
20:58 martinmorris ah ok, and what's in the zebraqueue will be my deleted biblios
20:58 chris *nod*
20:59 martinmorris will that also work for modified biblios, like where i link it to an authority?
20:59 chris thats what the cron job runs with (-z)
20:59 martinmorris (-z that is)
20:59 chris yes
20:59 martinmorris fantastic, thanks
20:59 chris any action that modifies the marc gets a row in the zebraqueue table, which then rebuild_zebra with the -z actions
21:00 martinmorris so in that case a full rebuild with -r -b is pretty rare?
21:00 chris yeah
21:00 id only do that if i did some changes to the db, not using the web interface
21:01 martinmorris i did seem a rather laborious way to work - glad i've got that straight
21:01 i need to get those cron jobs set up
21:01 chris yeah having rebuild_zebra run regularly makes for a more useful koha ;)
21:08 martinmorris now i've just run ./rebuild_zebra.pl -b -z and it hasn't cleared a biblio i've just deleted
21:09 chris you can check the zebraqueue table
21:09 martinmorris that's in the mysql database?
21:09 chris yep
21:09 martinmorris just found it
21:10 chris look for any rows that have done=0
21:10 martinmorris it's got 6207 rows
21:10 no rows
21:10 chris right dod you know the biblionumber of the one you deleted?
21:10 dod=do
21:10 martinmorris i'm just doing another one which is 191
21:11 and still get an empty set
21:11 chris you can then do select * from zebraqueue where biblio_auth_number = 191;
21:11 martinmorris yep, can see it now
21:11 done = 1
21:11 chris what was the operation?
21:12 martinmorris i was deleting any items attached to that biblio, then deleting the biblio itself
21:12 chris yep what was the operation column in zebraqueue?
21:12 martinmorris ah sorry
21:12 chris recordDelete ?
21:13 martinmorris specialUpdate and recordDelete
21:13 chris ok, you could try one with -v
21:13 and see whats happening
21:13 martinmorris nothing at all :)
21:13 chris (you dont have 2 zebrasrvs running or anything like that eh? .. ie we arent indexing one and searchign the other?)
21:14 martinmorris well this does run across two computers but i don't think i am, hold on though
21:15 doesn't look like it
21:16 chris cool
21:16 martinmorris here's my process listing after the two daemons are started up
21:16 22279 daemon --name=koha-zebra-ctl.koha --errlog=/var/log/koha/koha-zebradaemon.err --stdout=/var/log/koha/koha-zebradaemon.log --output=/var/log/koha/koha-zebradaemon-output.log --verbose=1 --respawn --delay=30 --user=koha koha -- /usr/bin/zebrasrv -v none,fatal,warn -f /etc/koha/koha-conf.xml
21:16 22280 /usr/bin/zebrasrv -v none,fatal,warn -f /etc/koha/koha-conf.xml
21:16 22283 daemon --name=koha-zebraqueue-ctl-koha --errlog=/var/log/koha/koha-zebraqueue.err --stdout=/var/log/koha/koha-zebraqueue.log --output=/var/log/koha/koha-zebraqueue-output.log --verbose=1 --respawn --delay=30 --user=koha koha -- perl -I /usr/share/koha/lib /usr/share/koha/bin/zebraqueue_daemon.pl -f /etc/koha/koha-conf.xml
21:16 22284 perl -I /usr/share/koha/lib /usr/share/koha/bin/zebraqueue_daemon.pl -f /etc/koha/koha-conf.xml
21:16 chris just checking
21:16 ohhhh
21:16 martinmorris that's on the sql machine
21:16 nothing on the www machine
21:16 chris you are running the zebraqueue daemon???
21:16 kick that in then head
21:16 you dont wnat to run that + rebuild_zebra
21:17 martinmorris oh right, is that what the daemon is meant to be doing?
21:17 chris yep, its deprecated in favour of rebuild_zebra as a cron job tho
21:18 martinmorris great, that's what i'll do then
21:18 so a -r -b to clear these changes then?  followed by setting up a -z -b cron job and no zebraqueue daemon?
21:18 chris yeah sounds good
21:18 martinmorris cheers
21:18 why wasn't my zebraqueue daemon doing its job i wonder?
21:19 chris yeah i think thats why its been deprecated
21:19 martinmorris LOL - makes sense :)
21:20 you might want to remove it from the INSTALL.debian file
21:20 chris hmm i thought it was, but yep will do
21:22 martinmorris thanks for your help - looks like it's OK now :)
21:22 chris cool
21:56 g'day IrmaCalyx
23:21 heya pianohacker
23:21 pianohacker Hi, chris
23:21 How's it going?
23:23 chris good good
23:23 pianohacker Anything interesting going on at work?
23:24 chris we release the nz stock exchange on  a new platform tomorrow
23:24 thats eating a lot of time
23:24 pianohacker Ahh, cool. Go-live's always fun and scary
23:25 chris *nod*
23:25 mostly scary
23:26 pianohacker Heh. The fun usually comes after the first crazy 48 hours
23:26 chris yep :)
23:39 as long as we dont crash the market ill be happy
23:39 pianohacker Yeah, slightly higher stakes than an ILS rollout (but only slightly, to some directors)
23:40 chris hehe
23:42 gmcharlt pshaw - stock markets are such short term things, barely worthly of consideration when you've got civilization itself to preserve
23:42 chris hehe
23:43 been reading ngc4lib?
23:43 gmcharlt chris: yep
23:43 chris it is pretty entertaining
23:44 gmcharlt I suspect that's why most people who subscribe to it read it
23:45 chris :)
00:01 pianohacker Hmm. Is there really that compelling a reason to have both the opacstylesheet and opaclayoutstylesheet preferences?
00:02 chris i dont think so
00:02 are they both used by the opac templates?
00:03 hmm yes
00:03 pianohacker It's kind of strange. The OPAC templates uses opacstylesheet if it's defined, then opaclayoutstylesheet, then opac.css if neither of them are
00:03 chris yeah
00:03 there there is the color one too
00:03 pianohacker #koha: What usage have you seen of the two preferences?
00:03 chris: Yup, though that's a separate area (thankfully)
00:04 chris and what is opac_css_override ?
00:04 pianohacker more in the realm of per-vhost personalization, I would assume
00:04 chris ah right
00:05 we should ask owen next time we see him
00:05 pianohacker I'll leave that as an unresolved part of the sysprefs editor for now, but yes
00:22 chris man i love git
00:22 pianohacker ?
00:22 chris git checkout 1.26-04
00:22 git merge origin/wrsomething
00:22 git merge origin/wrsomethingelse
00:22 bump the package number in debian/changelog
00:23 git tag -a -s -m 'preprod release' 1.26-05
00:23 ./scripts/deploy.pl -c preprod -v 1.23-05
00:23 release done
00:23 pianohacker Source code management is pretty special
00:24 I've seen someone merge ~100 branches, each a contribution from a separate developer, into a single Koha installation
00:24 Still tough, but actually possible with git
00:24 chris yep
00:26 hi irma
00:36 or not
03:08 atz chris: yeah, opac_css_override is where it pulls the CSS filename from apache SetEnv
03:09 pianohacker atz: Do you know anything re: usage of opacstylesheet/opaclayoutstylesheet?
03:09 chris ohhh
03:10 atz not really... i think there are a couple versions of essentially the same thing on the opac side
03:10 i guess b/c you might have different levels of CSS design...
03:10 pianohacker Presumably
03:10 atz but unnecessary imhoo
03:11 pianohacker I'll harass owen about it
03:11 Never dreamed the syspref editor would involve this much weirdness
03:12 chris yeah, less calls makes the page load faster, especially on IE
03:12 atz evergreen amasses all the CSS and pre-body js and serves it all in-line, w/ server-side minify
03:12 chris we could do that
03:13 as a first easy step
03:13 atz yeah, it seems better in some ways, but intermediate and browser caching is lost
03:13 chris we could minify it all when building a release
03:14 atz well, not the syspref-based stuff, but the truly static ones it would make sense
03:14 chris yeah the general css file
03:15 we could use a minified jquery too
03:15 atz chris: that's already done iirc
03:15 chris cool
03:15 atz we recently upgraded jquery version also
03:15 chris saw that
03:16 part of our build process for stuff is
03:16 cd frontend_templates/js && ./make_site_all.js.sh
03:16 making site-all.js
03:16 cd frontend_templates/js && ./make_min.pl
03:16 which combines all the js into one file, and minifies it
03:18 atz yeah, i think that might be best...
03:18 chris Minified site-all.js (62% original size)
03:18 certainly saves a bunch on file size
03:18 atz go ahead and incur the first load cost... should be cached by every piece along the way then
03:18 chris *nod*
03:19 atz system (or memcahce), network cache, browser, etc.
03:19 chris the other trick we do, is using mod_rewrite
03:19 links to the js css images etc
03:19 are generated in the templates with a epoch timestamp as part of the url
03:19 which changes is the image, or css etc is changed
03:20 mod_rewrite throws that bit away
03:20 which means we just set the expires headers to forever
03:20 and the browser only gets it when the url to it changes (which is when the file changes)
03:21 atz that seems pretty solid
03:22 chris yeah works pretty well
03:22 atz as long as you don't trip over user/application generated filenames that happen to fall in the namespace
03:23 chris *nod* its a directory structure, and we have control over that
03:25 Amit hi #koha good morning
03:25 hi chris
03:25 chris hi Amit
03:25 pianohacker Good night, all
03:26 chris night pianohacker
03:26 Amit hi pianohacker
03:26 pianohacker hibye amit
07:11 CGI566 my login session  is timedout immediately
07:11 how would i change the session timeout
07:32 chris good morning europe
07:34 nicomo good evening in Oceania?
07:34 hi chris
07:43 Amit hi nicomo
07:44 nicomo hi Amit
07:52 CGI566 hi all
07:53 my opac, staff users mailing session is timedout  soon. is there any possibility to extend the session time
07:56 chris there is a system preference that sets the timeout
08:23 hdl_laptop hi
08:27 chris hi hdl_laptop
08:29 hdl_laptop hi chris
08:29 how are you ?
08:30 chris good thanks, and you?
08:33 hdl_laptop good
10:28 Amit hi davi, hdl
10:29 davi hi
10:29 later
10:58 chris @quote random
10:58 munin chris: Quote #14: "joetho: I only looked for missing ccodes- not dysfunctional orphans." (added by wizzyrea at 08:31 PM, July 09, 2009)
11:07 gmcharlt good morning #koha
11:08 nicomo morning gmcharlt
11:08 morning gmcharlt
11:08 oups hit the enter key twice
11:09 gmcharlt :)
11:09 nicomo you're twice welcome then
11:10 chris hehe
11:11 nicomo with that : lunch time for me, brb

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

koha1