IRC log for #koha, 2005-06-03

All times shown according to UTC.

Time S Nick Message
12:01 slef paul: argh!
12:02 slef paul: do you want CVS access to kartouche?
12:02 slef paul: thanks for email
12:20 pate time for my semi-random kibbitz ...
12:20 pate anyone here thought about signing up with Google to make Koha mentoring organization for the summer of code?
12:45 slef pate: no, as I dislike google for being spammers who sponsor the World Economic Forum. :)
13:08 kados pate: i did
13:39 tim What do I use for inventory codes in the barcodes generator?  I thought it would be barcodes, but that's not doing anything.
13:48 paul_away kados => i've commited a new version, that do tag by tag (from 000 to 999). HTH
13:52 tim I had a problem with Koha looking for Utils.pm in /PDF/API2/PDF/.  It's in /PDF/API2/Basic/PDF/.
13:53 tim I tried a symlink and that got rid of the server error it was giving me.  Now I need to get some barcodes out of it.
13:53 paul_away kados => wait a minut, it's buggy...
13:54 tim I'm using 2.2.2b
13:55 paul_away kados => OK, seems it works now.
13:56 paul_away ok, leaving now, going with wife & sons. Have a good day.
14:00 kados paul_away: cool ... I'll give it a shot tonite
14:23 gavin hi all
14:29 tim hi gavin
14:29 gavin just wondering who was about, any developers?
14:30 tim I wish I knew enough.
14:30 gavin only one way ot learn :)
14:30 gavin do you run koha?
14:30 tim Does anyone know how to search a date range in the IRC logs?
14:31 tim Working on it.  We're hopeing we'll be running it this summer.
14:31 gavin where's the page again?
14:32 tim http://koha.org/cgi-bin/logs.pl
14:32 tim I found a mention of a problem I'm having, but I don't know how to get the rest of the conversation.
14:33 gavin just type the date you want in the between .. and boixes
14:33 gavin eg Betwen [ !st February ] and [ 1st Macrh]
14:35 tim Thanks.  I was trying to to just one date.  I added a day for the second one and got it working.
14:35 tim sorta
14:43 gavin now wokring for you?
14:47 tim I did it different and got the day I wanted, but also got a bunch I didn't.
14:48 tim The info I was looking for wasn't there anyway.
15:05 gavin ah. bummer
15:05 gavin are you a librarian?
15:13 GenjiZzZz hiya all.
15:14 Genji Welcome to a brand new day.
15:14 gavin welcome from the twilight of mine...
15:19 slef hello
15:21 gavin hi
15:32 slef http://simms.celt.dias.ie/~gavin/ doesn't look like a home page. What is it?
15:35 slef As in, what is your home page?
15:36 gavin my dev box in work
15:36 gavin my home page is http://fiachra.ucd.ie/~gavin/
15:41 slef Is now the time to start advocating groff?
15:45 Genji cool. figured out how to commit to HEAD (the tip of development) with WinCVS.
15:48 Genji Ahhh! HEAD is not a branch, because it was never created as such.. but rel_2_2 was created as a branch, and thats why it can commit to it without erroring.
15:48 Genji i think.
16:09 gavin does anyone have strong feelings about compatibility with postgres?
16:25 Genji are any coders awake?
16:26 gavin jsut wondering as fulltext indexes are not supported (by default) so if I do up a search using fulltext, postgres usage will get hard
16:26 slef Genji: HEAD is where you are if you aren't on any branch.
16:27 slef gavin: isn't it in contrib?
16:27 gavin yes, but not sure how complex it is to add (and don't want to make koha install even harder!)
16:28 gavin if everyone uses mysql right now then there's no real problem but if it's going to break someone who uses postgres then i'd rather know now
16:38 gavin i'm just trying to get 2.2.2b working.  I keep getting "Connection Failed" from the web pages though
16:46 gavin sorted, forgot to uncomment the listen lines
17:14 gavin Hi guys, does anyone know of some test data for a koha db?
17:18 chris hi gavin
17:18 chris i think joshua (kados) has some
17:18 chris i unfortunately dont have a good set for 2.2
17:20 gavin i tried the one that's in the tarball but it doesn't seem to be right.
17:20 Genji hiyas chris
17:20 chris yeah i think thats old 2.0 or maybe 1.2 data
17:20 chris hiya genji
17:21 chris cool
17:21 gavin there are no stopwords and i can't find anything i search for. will try adding some by hand
17:21 chris good idea
17:42 gavin it's bedtime this side of the world, see you later folks...
18:32 Genji thats wierd.
18:33 Genji nevermind.
18:55 Genji okaay. how do i add a totally new file to the cvs?
19:04 shedges http://sourceforge.net/docman/[…]p_id=1#commandadd
19:10 Genji do i have to like, add it to the module, then tag it HEAD?
19:13 Genji oh... think i've realised something.
20:35 kados Genji: you can export marc records from LibLime's demos
20:45 Genji kados: your marc records include the complex itemtypes.. like JNFB for Junior non fiction book?
20:46 Genji ive imported ISIS, dbf, Excel.. never marc.
23:39 Genji whats a good way to find if a book is out on loan or not, using the old db?
23:39 chris same as using the new db
23:39 chris check the issues table
23:39 chris select * from issues where itemnumber=? and returndate is NULL;
23:40 chris if you get a result, that item is on loan
23:42 Genji hmm... rachels feature request has a checkbox with "show only available." so.. that excludes books on loan, restricted, lost, unknown barcode and  withdrawn, right?
23:42 chris yep
23:42 chris lost u need to check
23:42 chris items.lost
23:43 chris if it =1 or 2 its lost on long overdue
23:43 chris on = or
23:43 Genji getiteminformation($env, 0, $barcode) can give me all i need, yes?
23:43 chris ought to
23:45 Genji bingo. it might return a duedate if its on loan.
23:47 Genji hrm... rewind..... im writing a sql query here... not a perl script.
23:49 Genji so... where items.lost=0 and issues.returndate=0 and .....
23:51 chris not quite
23:51 chris (items.lost=0 or items.lost is NULL) and issues.returndate is NULL
23:51 chris NULL is not equal to 0
23:51 Genji Ah, right.
23:51 chris items.lost will be NULL if its never been lost
23:51 chris but it will be 0 if it was lost, then found again
23:52 Genji and what about issues.returndate?
23:52 Genji gets reset to null each time its taken out?
23:53 chris a new row is made
23:53 chris so when a book is returned the returndate gets set to the days date
23:54 Genji ahh.. right.
23:54 chris if its issued again, a new row is made with a blank returndate
00:07 Genji where (items.itemlost IN (0,NULL) and (items.notforloan IN (0,NULL) and (items.wthdrawn IN (0,NULL) and and issues.returndate is NULL and reserves.borrowernumber = 0
00:07 Genji anything else im missing?
00:08 Genji lost, withdrawn, not for loan, reserved.... oh... out on loan.
00:08 Genji ah.. return date covers that...
00:08 Genji Am i missing anything?
00:24 Genji whats wrong with this query?
00:27 chris ur missing a )
00:28 chris and you have 2 ands
00:28 chris missing lots of ) .. too many (
00:52 Genji hmm? can't see the two ands, and can't see the mismatched ('s.
00:52 Genji you sure your looking at the select statement, and not the original where statement i typed?
00:59 Genji okay, theres something about the issues and reserves tables that makes this statement break down.
00:59 Genji select distinct m1.bibid from biblio,biblioitems,items,issues​,reserves,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 'reincarnation') doesn't work even.
00:59 Genji once i remove issues and reserves from the list of tables... it comes right.
01:01 Genji easy way to allow searching for availabilty.. would be to add items.reserved and items.outonloan bools.
01:02 Genji get issues and reserves to set and unset them.....
01:05 chris that wont work because u have no joining conditions
01:05 chris for items, issues or reserves
01:06 chris youd need something like
01:06 Genji oh ya.... that was a test query, doomed to fail by a mistake....
01:06 chris biblioitems.biblioitemnumb​er=items.biblioitemnumber
01:06 chris plus the conditions for issues and reserves
01:07 chris it would probably be a slow query
01:07 chris lots of joins
01:07 chris probably best to do it with 2 queries
01:07 chris the first one to fetch all the biblioitemnumbers
01:08 chris then u can just use them to find all the items that match
01:08 Genji ahh..
01:33 Genji actually.. i change m1.bibid to biblioitems.biblioitemnumber .....
01:33 Genji so, got a list of biblioitems....
01:37 Genji what exactly is m1.bibid? its a list of biblionumbers, right?
01:42 Genji 'select distinct biblionumber from biblioitems, items, issues, reserves where biblioitemnumber=items.biblioitemnumber and items.itemnumber=issues.itemnumber and reserves.itemnumber=items.itemnumber and items.itemlost IN ('0',NULL) and items.notforloan IN ('0',NULL) and items.wthdrawn IN ('0',NULL) and issues.returndate is NULL and reserves.borrowernumber=0 and biblioitemnumber IN (result of first select) order by biblio.title ASC?
03:21 gavin hi
03:41 slef gavin: installing stuff from contrib is usually not too hard. Note it in the docs that it'll be required if postgres is ever supported ;-)
03:42 Sylvain hi
03:44 slef Sylvain: hi.
03:44 slef What's an iwi?
03:44 Sylvain an iwi ? I don't know
03:55 gavin ah, fair enough
03:56 gavin so,. i've been doing some further work to understand exactly how things are done in Koha
03:56 gavin I see now what stephen said before about marc_word as an indexed table
03:56 slef and you conclude the answer is "slowly"?
03:57 gavin :-) it's not slow now (on a different machine)
03:58 gavin though i'm still unsure why it seemed so slow on the other one, I'll have a look another time
03:58 gavin maybe i did something wrong
03:59 chris iwi probably translates best as tribe (or perhaps nation)
03:59 chris if you are talking about iwi in maori that is
04:01 gavin I've been trying searches via the opac and watching the queries it uses
04:01 gavin i suspect much of it could be simplified by use of match and boolean mode, but I'm unsure if there may be stuff which would not work
04:01 gavin without the marc_word table
04:05 gavin does anyone have any suggestions how best to proceed?
04:08 chris probably how i would do it is the way paul did his recent search improvement to test
04:08 chris take a copy of the search module
04:09 chris then make some changes, and then to use it just change the use C4::SearchMarc; to be use C4::Searchtest; (or something)
04:10 gavin okay
04:10 gavin I'm trying to sort out in my head whether fulltext/boolean is what's required here
04:10 chris right
04:10 chris i think for the keyword search
04:10 chris fulltext might be the one
04:12 gavin do you think not for the others?
04:12 chris maybe on the others too, its tricky
04:12 gavin fair enough
04:13 gavin one issue is stopwords, if you use mysql to do the searching, you have to define the stopwords in mysql too
04:13 chris if i search on 'iwi'  ... should a match in title be ranked higher than a match in author?
04:13 gavin this can be done of course
04:13 chris in the ideal world
04:13 gavin i don't know
04:13 chris we would let the library set the rules
04:14 gavin yes
04:14 chris that would be a pretty neat feature
04:14 gavin you mean to let the library prefer eg author or title?
04:15 chris im just thinking for the keyword search ie
04:15 gavin i worked a little more in the past few days on our in-house one http://simms.celt.dias.ie/~gavin/opac/
04:15 gavin <sorry, please continue>
04:15 chris http://opac.library.org.nz/cgi[…]koha/opac-main.pl
04:15 chris the search box here
04:16 chris it would be cool if we let the librarians choose what that searches
04:17 gavin i see. so you would like libraries to be able to specify greater ranking to matches within author than keyword (for example)? or is this already in place?
04:17 chris yeah, they might say, keyword searches, author, title, subtitle and subject
04:17 gavin or just to restrict what it searches?
04:18 chris maybe both
04:18 chris restricting what it searches definitely
04:18 gavin is this not the purpose of the advanced search though?
04:18 chris yep that allows the user to search what they want
04:19 chris but i just know that for example HLT didnt want keyword search to search authors
04:19 gavin i see.
04:20 gavin so suppose for the sake of argument we had a hidden fields in that page with "author, title, subtitle" which got checked for sanity and plugged into the query in a smart way
04:21 chris was just thinking aloud
04:21 gavin fair enough. i'd never have thought of that
04:22 gavin if you look here http://dev.mysql.com/doc/mysql[…]text-boolean.html
04:22 gavin WHERE MATCH (title,body) AGAINST ('+MySQL -YourSQL' IN BOOLEAN MODE);
04:22 gavin is pretty much what that would need i guess
04:23 gavin off the top of my head, i'd say you'd put a single fulltext index on all columns and another on each one individually for the advanced search
04:23 chris right
04:24 chris you are thinking to search against the biblio table or the marc_word table?
04:24 gavin the biblio
04:24 gavin but i'm not sure if that'll do as much as you need
04:25 chris yeah, perhaps we need a librarian
04:25 gavin i see you also have partial words via 'tower*' type syntax.  
04:26 gavin boolean will do this but it can only trail on the end of a word not within or at the start
04:26 chris right
04:26 chris thats probably the best way for it to work anyway
04:27 chris but i think we probably need to get comments from a librarian ... about exactly what theyd expect a title search say to search
04:28 chris shold it search subtitles also?
04:28 chris that kind of thing
04:28 gavin again i think that would be best defined by librarian onsite
04:28 chris true
04:29 gavin that said, chances are many of them would never do it so a "best default" is imporant
04:29 chris yeah
04:29 gavin the one i did before was for a small academic inst
04:30 chris ok, i have to hand the laptop over to my wife, ill catch you later gavin, good luck
04:30 gavin i got some feedback from scholars like:
04:30 gavin thanks for the chat :)
04:31 gavin a big help.
04:43 slef chris: yes.
04:45 gavin incidentally, i hope i haven't pissed anyone off. i've kind of just walked in and started saying stuff.
04:46 gavin I'm well aware that Koha has been around for some years and is a good working product
05:36 chris for a little while before bed
05:36 chris nope gavin, well you certainly havent pissed me off .. anyone who wants to contribute is more than welcome in my book
05:39 chris slef: its election year here .. so all the nasty raced cards get played
05:40 gavin well, apologies if anyone has been irked. i didn't mean to do that
05:41 gavin slef: where are you from?
05:45 gavin i'm looking at a whole load of hidden vars and sqlUpdate() and such on the opac. is this stuff just used on the librarian's "mega-advanced" search?
05:46 chris i think so
05:47 chris i must confess to not having spent a lot of time on the search stuff in the last year or so
05:47 gavin so ideally, one would want to keep those code bases together (in order that the searches always work the same way)
05:47 chris yep for the most part
05:48 chris with the cavaet that the librarians get to see things the opac users dont
05:48 chris like lost items, or ones marked withdrawn/cancelled
05:49 chris but thats kind of the second level of search
05:49 gavin yes
05:49 gavin how'd you mean?
05:49 chris you search the bibliographical data first
05:49 chris which should be the same for the opac and the librarian
05:49 gavin oh i see, actual copies of books as opposed to a given title
05:50 chris but then you might filter those results differently depending on who you are showing them too
05:50 chris yeah, thats what i meant
05:51 gavin one more thing, i'm working off the installed code (2.2b) as it's convenient and working
05:52 chris right
05:52 gavin am i likely to end up with great trouble mocing into cvs?
05:52 gavin moving
05:52 chris have you worked with cvs much before?
05:52 gavin yeah, a bit. some cvs and some svn.
05:53 gavin can i check out the right dirs direct?
05:53 gavin into the installed location?
05:53 chris ahh theres a good guide to that here
05:53 gavin nice one
05:53 chris http://www.saas.nsw.edu.au/koh[…]mLinkInstallation
05:54 gavin exactly what i need...
05:54 gavin thanks
05:55 chris http://www.saas.nsw.edu.au/koh[…]hp?page=CvsLayout might be useful too
05:55 chris the wiki is a good place to note down any thoughts etc you might have as well
05:56 gavin i put a couple of ideas on the release manager's notes (fulltext indexes mainly)
05:56 chris sweet
05:57 chris whereabouts in ireland do you live?
05:58 gavin dublin.
05:58 gavin are you a rugby fan?
05:59 chris most certainly
05:59 chris lots of rugby this weekend
05:59 gavin so you're looking forward to the upcoming tour
05:59 chris nz maori vs fiji, lions vs bay of plenty, and the jonah comeback game
05:59 chris yeah, it should be a great tour
06:00 gavin have you got tickets?
06:00 chris it depends :)
06:00 chris i have a friend who has a ticket to the wellington vs the lions game
06:00 gavin for you  i mean :)
06:01 chris and he is supposed to be travelling from canberra (in aus) to wellington that day
06:01 chris if he's delayed i get to go
06:01 chris so im hoping for fog :)
06:01 gavin ah.
06:01 gavin that's an interesting position
06:02 chris you had to be in very fast to get a ticket
06:02 chris and i wasnt fast enough
06:02 gavin i must admit i'm a little nervous for the lions
06:02 gavin so i heard
06:02 gavin whatever about the argentina game, they just aren't a strong lions squad
06:03 chris yeah, they arent who I would have picked
06:03 gavin and nz seem very strong right now
06:03 gavin i'm not sure i would've changed selection all that much but there are a few positions where they just don't have good guys to pick from
06:03 gavin NO. 8; Hooker to name two
06:04 gavin Dalaglio is past it, Owen is okay but not great
06:04 gavin this is getting somewhat off-topic :)
06:04 chris :-)
06:04 chris indradg and I talk cricket
06:04 chris (hes from india)
06:05 gavin i'm playing cricket this evening :)
06:05 chris indoor im guessing?
06:05 gavin no, it's not that bad but we do get interrupted a fair bit :)
06:05 chris ahh summer there of course
06:05 gavin as it happens one of the irish international grounds is my club up the road
06:05 chris excellent
06:06 gavin irish weather is wet all year :)
06:06 chris :)
06:06 gavin so i get to play there (it's rather small as international grounds go mind)
06:06 chris well its gone 11 here .. i should be heading to bed .. long weekend coming up (queens birthday) so will be a busy day tomorrow trying to do friday and mondays work
06:07 gavin sound. talk to you later
06:07 chris cya later, good luck at cricket
06:07 gavin thanks for your help...
07:22 gavin is the opac-search.pl broken in CVS right now?
07:28 Genji don't think there should be any errors at this stage.
08:00 slef oof
08:01 slef chris: :-(
08:01 slef gavin: Northamptonshire, England
08:09 gavin hmmm. I'm getting "Internal Server Error" from the opac-search.pl
08:09 gavin slef: just curious :)
08:25 gavin there seems to be some issue with opac-search.pl
08:26 gavin it currently ends:
08:26 gavin    409 else {
08:26 gavin    410 output_html_with_http_headers $query, $cookie, $template->output;
08:27 GenjiZzZz thought i fixed that bug.
08:27 gavin well, i'm not a cvs expert but i moved it out of the way and updated again and got the saem
08:28 gavin is your fix committed?
08:29 GenjiZzZz just made sure of it now.
08:30 GenjiZzZz btw, my new sidebar is implemented totally. only thing that doesn't work is the "only show available items" checkbox.
08:31 gavin very nice.
08:32 GenjiZzZz thanks. write up some feedback and send it to me, if you want. waylon at robertson dot net dot nz. Im being mindful that people can google their way through the logbots logs.
08:32 gavin just having a look through the web based cvs to see what's wrong with my copy
08:32 GenjiZzZz really going to bed now. night
08:33 GenjiZzZz back in around 7 hours.
08:33 gavin the else isn't fixed in cvs
08:33 GenjiZzZz err...... it isn't?
08:33 gavin <sorry.
08:33 gavin http://cvs.sourceforge.net/vie[…]=1.25&view=markup
08:33 GenjiZzZz but i just committed it. ah.. wrong version, thats the old one.
08:33 GenjiZzZz look at 1.26
08:34 GenjiZzZz Wasn't my bug, but i fixed it anyway.
08:34 gavin ah, there must be a deal on the web one
08:34 GenjiZzZz ya.. webcvs is usually slower than the actual cvs server... i think.
08:35 gavin s/deal/delay/
08:35 gavin fair enough :)
09:12 slef Anonymous CVS at sourceforge used to have up to 24 hours delay
09:12 slef it was possible for people to DoS the developers by hammering the webcvs
09:12 slef which wasn't good
09:17 gavin i see.
09:19 slef tim: perl -MCPAN -e shell
09:20 gavin got tired of IrcII
09:23 tim Thanks slef.  It still didn't find anything.
09:23 slef tim: not sure koha is using PDF barcodes instead of barcodes-on-the-fly or GNU barcode.
09:24 tim It was asking for PDF stuff so I figured it needed it.
09:25 tim It's not giving me errors anymore anyway.  It's not doing anything.
09:25 gavin Slef: My DB seems to be old (code complaining of missing tables), any idea where I'd get a dump of a new db?
09:26 slef gavin: I wonder if running misc/updater/updatedatabase would help
09:26 gavin i installed from 2.2.2b
09:38 gavin i gather i'm seeing Genji's new sidebar
09:47 slef no, almost certainly not
09:47 slef irssi and xchat seem popular at the minute
09:47 slef bitchx used to have all sorts of silly defaults that would lead to a runaway client, which blows, which blows
09:48 gavin i hope it's better now now
09:49 gavin may be irrssi
09:52 gavin taht's better
09:57 gavin ?
10:16 hdl gavin : I use kVIRC and am happy with that. ;)
10:39 Sylvain yes, we're just commiting some templates :)
10:39 Sylvain about 80 files to commit, sorry ;)
10:40 slef new design or new language?
10:41 Sylvain readjusment of plug-in
10:41 Sylvain in french
10:41 Sylvain some design changes and some predefined fields modified
10:48 gavin hdl: thanks, i'll give it a look
11:40 pate kados, you around?

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