IRC log for #koha, 2005-05-27

All times shown according to UTC.

Time S Nick Message
12:11 kados paul: the z39.50 server I wrote doesn't discard results that the library has ;-)
12:11 kados what would be the point? ;-)
12:12 kados or were you using Koha's z-client to search the z-server?
14:56 slef shaun: reading logs?
15:00 owen Or maybe sawing logs?
15:02 tim at least about to saw.  gotta move around a bit
15:03 shaun slef: yes
15:03 shaun i solved the problem.
16:02 Genji morning guys.
16:25 kados morning Genji
16:36 Genji hiya all.
17:16 kados hey russ
17:16 russ hi
17:49 slef what's FRBR?
17:54 kados Functional Requirements for Bibliographic Records
17:54 kados http://www.ifla.org/VII/s13/frbr/frbr.pdf
17:54 kados http://www.oclc.org/research/p[…]/frbr/default.htm
17:54 kados http://www.loc.gov/cds/FRBR.html
17:54 kados that last one is probably the best intro
19:22 Genji hiya yall. how do you make a book nonfiction or fiction? is it a itemtype thing?
19:23 Genji So, a book, adult non fiction (ANF) would pass through nonfiction filter *NF?
19:24 Genji how would my script know what itemtype/class is nonfiction/fiction?
20:58 rach it is itemtype
20:58 rach but i was trying to agrigate kids NF + Teens NF + Adult NF
21:21 Genji oh hiya rach.
21:21 Genji ya. hmmmm....
21:22 Genji Ahh... its common for libraries to have KNF TNF ANF as their itemtype codes?
21:32 rach it was the idea of "super classes" really
21:32 rach and i think that certainly they have junior non fiction and just non fiction
21:32 Genji ahh. so really.. you want super classes implemented?
21:32 rach in a public library
21:33 rach I wanted a library to tell us if it was a good idea
21:33 Genji hmm.... i can see that it would be cool.
21:33 rach I thought it would be cool, so that's 2 of us :-)
21:33 rach but it might be hard
21:33 Genji nope.
21:33 rach however I think the rest is not so hard
21:33 Genji i can see that it'll be a preferences table i think...
21:34 Genji superclasses -> superclassname (multiple) -> superclassitem (string of class codes)
21:35 chris way ahead of ya :)
21:35 chris take a look here
21:35 Genji yes please.
21:36 chris http://opac.library.org.nz/cgi[…]ha/opac-search.pl
21:36 chris see the classes?
21:36 Genji hmm.. ya.. i been there.
21:36 chris ysql> select * from itemtypesearchgroups;
21:36 chris +-------------------+--------------​---------------------------------+
21:36 chris | groupname         | itemtypes                                     |
21:36 chris +-------------------+--------------​---------------------------------+
21:36 chris | Adult Fiction     | BF|BLF|BSF|BPF|PHB|PPB|BYF                    |
21:36 chris | Adult Non Fiction | BN|BLN|BPN|BSN|BYN|BR|F|BH|BM                 |
21:37 chris | CD Rom            | C|CD|CJ                                       |
21:37 chris | DVD               | DP|DJP|DYP|D|DJ|D
21:37 rach ah nice
21:37 Genji ahh... cool..
21:37 chris which is used in the template
21:37 Genji ill put it into rach's feature request.
21:37 chris <option value="BF|BLF|BSF|BPF|PHB|PPB|BYF">Adult Fiction
21:37 chris <option value="BN|BLN|BPN|BSN|BYN|BR|F|BH|BM">Adult Non Fiction
21:37 rach this is something you'd have to do for each library - because they all use different classes
21:37 Genji yup.
21:37 rach and would want different things
21:38 chris so that needs a nice interface to edit that table
21:38 Genji sounds like it would be easy.
21:38 chris im thinking checkboxes or something
21:39 Genji Would save them the effort of remembering itemtype codes.
21:41 Genji hmmm.. is rach always like this? here for a few minutes, gone again?
21:41 chris yep, cos shes working
21:41 chris like me
21:41 chris so during the day we only glance at #koha
21:42 Genji i have three screens.
21:42 Genji and #koha has been on the right one for the last 24 hours.
21:42 chris i watch #koha but dont usually say much during the day cos im usually too busy
21:47 Genji so, pesudocode for the interface would be, 1. load itemtypes from its table. name and code only. 2. create table with headers. 3. place an add textbox with add command button. 4. on already added groupnames, put all the itemtypes in checkboxes (X) with full name. 4. have Save and Cancel buttons. Save would "update" the groupname table.. and "insert" if add was ever pressed.... then return to system preferences.
21:47 Genji what you think?
21:48 chris sounds ok to me
21:52 Genji okay.. think ill finish off rach's feature request.
21:52 Genji then work on the table and interface. I need to modify updatedatabase.pl to account for the new table, right?
21:52 chris yep
21:54 Genji is there a table to tell systempreferences to display a link to edititemtypegroup.pl?
21:54 chris dunno, u might have to edit the template
21:55 Genji cool.
21:58 Genji could i get your source for opac-search.pl from you?
21:59 chris probably wont help you much thats old 1.2 koha
22:00 Genji hrm.... well.. the source for the group lookup would be good, and how it gets put into the catalogsearch.
22:00 chris ahh thats just fetched from the table and put into a template
22:00 chris ie its stored as
22:00 chris BN|BLN|BPN|BSN|BYN|BR|F|BH|BM
22:01 chris and thats what ends up in the form
22:01 Genji oh.. right.
22:01 chris <option value="BF|BLF|BSF|BPF|PHB|PPB|BYF">
22:01 Genji so it just becomes &marclist=items.itemtype ----- &value=BF BLF BSF etc?
22:02 chris no marclist in 1.2
22:02 Genji ah.
22:02 chris (as i said, it wont help ya much) :)
22:03 Genji hmm....
22:03 Genji ill see what i can come up with.
22:07 Genji out to do stuff. back in... oh.. an hour.
23:17 Genji back
23:18 Genji chris, ya watching?
02:17 Genji ouch. cleaning my room, and my computer plug got pulled out accidently.
03:02 hdl Salut.
03:02 Genji Salut, hdl. I only speak english.
03:02 hdl Hi Genji.
03:04 hdl paul : I have read The latests issue of LinuxPratique Magazine. It has an article about Koha. ;)
03:52 Sylvain salut tout le monde
04:01 Genji ah. drat. wheres the logbots output address?
04:11 Genji found it.. and found what i was looking for. I love the logbot.
04:31 hdl paul : commiting acqui/histsearch modifications : interface design could certainly be improved
04:32 hdl But I have no idea how.
04:34 hdl For detail.pl problem in intranet... Problem on subject search. I think I cannot use Authorites search, and search.march/search.pl either. Am I right ?
05:24 hdl paul : the 'filter' proposed on borrowers w/o issues is not realy a "filter" as such : Indeed, it leads to much more results. For Instance, borowers w/o issues since today...  ;)  But I did it ;)
07:46 paul hdl ?
07:46 paul ton commit sur borrowers_out.pl fonctionne bien ?
07:46 paul parce que pour moi, pas du tout. Tu utilises des requètes imbriquées, qui ne sont supportées que depuis mySQL 4.1
07:47 paul or ce n'est pas encore assez répandu, il faut être compatible avec 4.0
07:47 paul et faire sans requètes imbriquées.
08:43 FrancoisL Hi everyone, hi Paul ! I finally have 10 minutes free (my next appointment isn't started yet !)
08:44 paul alors bienvenue. c'est une heure habituellement calme.
08:44 FrancoisL I was at the BPI (largest public library in Paris, 450 public stations) with the SAN this morning.
08:44 FrancoisL 'twas gorgeous :)
08:44 paul what kind of gorgeous ? You install Koha next week ?
08:44 paul lol
08:44 FrancoisL :) nope - MMediaView...
08:50 kados hu FrancoisL, paul
08:50 kados hi even ;-)
08:50 paul 'morning joshua
09:02 FrancoisL Salam, Joshua :!: How's life ?
09:02 kados pretty good, and you?
09:02 kados weather in athens is beautiful today
09:03 paul in Marseille too.
09:03 FrancoisL Nice :) 31° C here, some clouds...
09:03 paul in Paris, probably raining (for francoisl)
09:03 kados about 30 here today ;-)
09:03 paul ;-)
09:03 kados hehe
09:03 FrancoisL I was in Martinique last week - 35°C, great sun (water at 28°C)
09:03 kados oohh nice! like a bath ;-)
09:04 kados (except for brief times in Greece)
09:04 FrancoisL No see in Athens - what a paradox...
09:04 FrancoisL Must lmeave friends ! Be back in an hour...
09:04 kados I know! although the real Athens is pretty dirty -- the islands are nice for swimming ;-
09:04 kados ciao
09:05 paul joshua/kados, do you have few minuts to speak ?
09:05 paul (it's about "quest for search")
09:12 kados sure
09:12 kados paul: you have an idea?
09:12 paul (on phone)
09:12 kados ok
09:21 kados paul: I'm listening so when you're back let me know your ideas
09:24 slef hi shaun
09:30 paul kados, i'm back
09:30 hdl paul : par rapport à ce que tu me dis sur mon commit : chez moi ça marche, sinon je ne l'aurais pas commité. ;)
09:30 paul et tu es en 4.1
09:30 paul ?
09:30 hdl oui.
09:31 paul donc le pb vient des requètes imbriquées.
09:31 hdl Mais je ne vois pas comment faire autrement que par une requête imbriquée.
09:31 paul en général la réponse est : en 2 requètes
09:31 kados me too
09:31 kados ;-)
09:31 paul i wanted to speak of Search.
09:31 kados right
09:31 paul i've some questions
09:32 kados ok ... shoot
09:32 paul plucene seems not to be what we are looking for isn't it ?
09:33 paul on one hand I agree it's stupid to re-invent the wheel. So if we can find a useful tool, then we should use it.
09:33 paul On the other hand, we are looking for a very specific tool.
09:33 kados well the features are very good with plucene but as far as I can tell it's too slow
09:33 paul it seems.
09:33 kados I agree
09:33 paul we are looking for a very specific tool because we have a complex scheme & dependancies.
09:33 paul for example an "author" can be more than 1 thing.
09:34 kados http://www.namazu.org/doc/tutorial.html#run-mknmz
09:34 paul (author, co-author, authority author...)
09:34 kados http://59.93.161.79/cgi-bin/namazu.cgi (here is an example of Namzu with a Koha database ... sort of --it's indrag's)
09:34 kados right  ... all the relevant marc tags/subfields
09:35 paul (ip unreachable)
09:35 kados too bad
09:35 kados he must have turned off his machine ;-)
09:35 slef is plucene related to lucene?
09:35 kados yep ... a perl port (a slow one)
09:35 paul my goal is to have a Koha not to hard to install
09:35 kados hmmm
09:35 paul (s/to/too)
09:36 slef lucene was never the fasted IME
09:36 kados slef: that explains it ;-)
09:36 indradg hi guys
09:36 paul hi indradg
09:36 slef paul: on that note, can you stop adding undocumented code to the installer, please?
09:36 kados slef: do you know of a fast and comprehensive fulltext indexer/search engine?
09:36 indradg not shut off... IP changed... ADSL
09:36 kados ahh
09:36 paul do you think my idea with marc_words should be definetly discarded ?
09:37 kados not definitely
09:37 slef kados: I've not used searchers much for years. htdig, swish and glimpse were the state of the art back then.
09:37 kados but I don't think it's a long term solution
09:37 paul slef : what are you talking about ? i haven't commited things on installer since months.
09:37 kados we still would have trouble doing boolian searching
09:37 indradg kados : http://59.93.166.247/cgi-bin/namazu.cgi
09:37 slef We may well find that custom searching is needed, but to do that, we probably need to give a search guru a map of the database
09:37 paul kados :not sure, because we can do join, unions... with perl arrays.
09:38 slef paul: auto_install.
09:38 kados if we use a textual database tool that already exists, all we will have to do is get our marc data into a textual format for easy indexing ... it allows us to easily create indexes of anything we way
09:38 kados want
09:38 slef indradg: The requested URL /cgi-bin/whereever you want it to go was not found on this server.
09:38 paul slef, OK
09:38 kados and it allows us to do complex query syntax without having to reinvent the wheel
09:39 kados (and it should be faster)
09:39 paul kados, should I give a try to my idea ?
09:39 kados sure ... if you get a script to create marc_words I will try it on NPL data and we can compare to Lucene, Namzu, etc.
09:39 paul like some code to build marc_words table & do union/join search on 2 items ?
09:39 kados yep
09:39 kados let's try iit
09:39 kados it
09:39 paul ok, i'll give it a try asap.
09:40 paul another question :
09:40 paul about argentinian code.
09:40 paul you've got it isn't it ?
09:40 kados I haven't heard from emiliano since the last two times I emailed him
09:40 kados he hasn't sent me the tar.gz yet
09:40 kados so no ;-)
09:40 kados I'll bug him again
09:40 paul ah, ok. I thought he did it.
09:40 kados you will hear when he does ;-)
09:41 paul OK, thanks ;-)
09:42 kados paul, slef, indradg, anyone else listening: what is a good time for a 'bug squashing session' on #koha?
09:42 paul what do you call a "session" ? a bug squashing like an IRC meeting ?
09:42 kados yes
09:43 kados session is 'period of time'
09:43 kados 'where something is done'
09:43 paul for bug squashing assignment or writing/commiting code in real time ?
09:43 kados both
09:43 kados we'll step through each bug
09:43 kados and decide: hard/easy
09:43 kados easy ones get squashed in real time
09:43 paul most bugs remaining are hard one.
09:43 kados hard ones get assigned
09:44 paul i'll have some times for bug squashing in the next weeks.
09:44 slef kados: during work hours ;-)
09:44 paul thanks to hdl hard work I have some time that i hadn't before.
09:44 kados slef: can you give me a range of hours that would work for you?
09:45 paul the problem being work hours are differents between NZ / Europe / America :-(
09:45 kados right ... well I can work anytime
09:45 kados so NZ/Europe is the big prob ;-)
09:45 slef kados: 09:00-17:00 +0000, some evenings 19:00-20:00 or 22:00-00:00
09:46 paul for me it's 7 => 16GMT
09:46 kados maybe we can rotate the 'session' times
09:46 paul (being GMT+2)
09:46 kados right
09:46 paul i think an irc meeting for bug assignment would be great, at minimum
09:46 slef I can try to do earlier (badly, mostly) but would rather not go late
09:46 kados slef: is that GMT?
09:46 slef yeah, assignment and bouncing
09:46 paul (GTM)
09:46 slef kados: well, +0000
09:47 kados ahh
09:47 slef is that GMT or UTC? ;-)
09:47 kados hehe nice
09:47 kados lucky bastard ;-)
09:47 paul isn't it the same thing ?
09:47 slef (solar or atomic)
09:47 slef no
09:47 slef :)
09:48 kados paul: best times for you (GMT)?
09:48 kados hdl: you too?
09:48 paul as you want.
09:48 paul best between 12 and 16 (GMT)
09:48 hdl ok. for me.
09:48 slef brb, navigating legislation
09:49 kados ok ... so 15:00 GMT would work for slef, paul, and hdl
09:49 kados how about indradg ?
09:49 indradg right now its 0305 UTC
09:49 indradg since I'm +0530 GMT
09:50 indradg UTC == GMT ?
09:51 kados I think so in practice
09:51 indradg that makes to about 7:30 AM IST
09:51 indradg works for me
09:55 kados ok ... I will propose 15:00 GMT for the first meeting
09:55 kados it's 3am in NZ so I doubt we'll get chris to join us
09:55 kados so next meeting we'll have to rotate
09:57 shaun hi slef
09:58 hdl kados : which day ?
09:59 kados http://tinyurl.com/chp5t
09:59 kados Tuesday, May 31 good for all of you?
09:59 indradg kados, makes sense... rotation of meeting time
09:59 indradg 31st May is good
10:00 kados good ... that way, probably only one continent will be sleeping ;-)
10:00 kados for each meeting ;-)
10:00 paul ok for me
10:00 kados slef, hdl paul, is Tue the 31 good for you?
10:00 kados ok good.
10:00 kados that's quorum then ;-)
10:00 paul 14 would be better
10:00 paul (I meanla bibliothèque de sociologie du CNRS
10:00 paul CNRS - UPS 2716
10:00 paul 59/61 rue Pouchet
10:00 paul 75849 PARIS cedex 17
10:01 paul oups, sorry)
10:01 paul I mzn 2PM GMT would be sooner & we would have 3 hours behind
10:01 paul (before dinner)
10:01 kados so 14:00 GMT?
10:01 kados on Tue 31?
10:01 paul if OK for other, then OK for me.
10:01 hdl ok
10:02 kados alright I'll write the email
10:02 indradg ok
10:02 hdl tx. ;)
10:02 shaun kados: under which category does the blogging come? just so I know to keep it out of the discussion tonight ;-)
10:05 kados shaun: I think blogging would be a fine thing to talk about tonight
10:13 slef kados: I think it works.
10:15 kados cool
10:22 shaun just it wasn't mentioned on the agenda, that's all
10:24 kados shaun: I'll add it
10:24 shaun great
10:24 kados shaun: thanks for the catch
10:30 slef any other RSS for the shared blog?
10:31 indradg slef, where is the blog being setup?
10:32 slef I'll tell you once it looks halfway there
10:32 slef could do with more CSS love than I have time for now, but like all good things, rushed CSS doesn't work
10:34 indradg right-o!
10:36 slef hrm, kados's feed is breaking it, I think
10:37 kados slef: hehe
10:38 kados slef: what about my feed is breaking it? (is it timing out?)
10:38 slef it's not detecting your name, for some reason
10:38 slef doesn't quite break it, but it looks ugly ;-)
10:38 kados I've got some special chars there ... that might be the problem
10:39 kados creen chars
10:39 kados greek chars that is ;-)
10:39 slef as entities too
10:39 kados is it the <title> tag causing probs?
10:39 owen Seems to validate okay (mostly): http://feedvalidator.org/check[…]m.cgi%2Findex.rss
10:40 slef well, minor... mostly it goes "uh, I can't parse that" and drops it on the floor
10:40 slef it's at the next pass (copying the latest entries out of the cache) it gets an odd display bug
10:40 kados the <title> or the <language>?
10:41 slef title, pretty sure
10:41 slef can you try unicode instead of entities?
10:41 kados hum ... I think I did and it didn't work well
10:41 slef really, I should update the parser library on that machine ;-)
10:41 slef it's >1 year old
10:42 slef ok, I'll bodge this
10:43 kados it validates now
10:43 slef ok, I'll try before bodging
10:44 slef still broken, bodging
10:45 kados these are utf-8 right?
10:45 kados &#x03BA;&#x03B1;&#x03B4;&#x03BF;&#x03C2;
10:45 kados hehe
10:45 slef damned if I know. Shouldn't matter.
10:46 kados I thought they were unicode
10:46 kados at the time
10:47 kados slef: here's a chart:
10:47 kados http://www.theorem.ca/~mvcorks[…]art=0370&end=03FF
10:48 kados I'm not sure how to 'write' the numbers (instead of entities) in the html ... do I just do U+03BA
10:48 kados ?
10:48 kados erp ... xml
10:49 kados it doesn't actually print the letter when I view it that way -- just the U+03BA :-(
10:55 slef tum tee tum
10:55 slef well, that's not quite what it should be
10:55 slef kados is now known as "unknown"
10:55 kados hehe
10:58 kados slef: with this increased activity, now might be a good time to try to revive your webring
10:58 kados I'm sure chris would join
10:58 kados and maybe we could get owen to start a weblog too ;-)
10:58 owen If he wasn't so damn lazy...
10:58 kados :-)
11:01 slef http://www.otherwayup.org.uk/planets/koha/ is the bodgeup
11:03 owen How can we get it not to show non-Koha posts?
11:03 owen Is it possible?
11:03 kados with my blog it is
11:03 slef non-koha posts?
11:03 kados and chris's too
11:03 slef just shove whatever you want into the blog I use
11:04 owen Maybe it's a question of having each person set up a special feed of only Koha-categoried posts
11:04 slef I have a koha feed
11:04 kados http://kados.org/cgi-bin/blosx[…]og/koha/index.rss
11:05 kados that's my Koha feed
11:05 kados here's chris's:
11:05 kados http://www.bigballofwax.co.nz/[…]om/koha/index.rss
11:06 FrancoisL (ouf!) back from neverending meeting...
11:06 kados hey FrancoisL :-)
11:07 kados looks nice slef
11:09 kados it would be nice if each 'different' blogger had a unique background color for their 'title' div
11:09 slef kados: you can't write html, you notice? ;-)
11:09 kados slef: hehe yep
11:09 slef yeah, underlying code needs a rewrite
11:11 slef I need to add an html parser as well as the xml, as you're not alone
11:14 owen Strange that some links word and some don't
11:15 owen (that is, some are displayed as HTML)
11:15 slef depends if it's valid xhtml
11:15 slef one in the top post has an unescaped &
11:20 slef right, sorted out the categories
11:20 slef I'll add it to the tridaily update
11:23 kados huh ... it's not catching everything
11:26 kados slef: I'm using rss 2.0 and chris is using .91 fyi

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