IRC log for #koha, 2006-06-10

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

All times shown according to UTC.

Time Nick Message
15:52 chris morning
16:30 kados very well
16:31 chris excellent
16:31 kados we should sign with them within a couple weeks if the contract negitiation goes smoothly :-)
16:31 chris all good
16:31 kados they were very impressed with the search
16:31 so staying up till 5am paid off :-)
16:34 chris glad to hear it :-)
16:38 kados the search is really getting there
16:38 I'm hoping to work a bit more on it this weekend
16:38 then I need to tackle the adds/edits in zebra
16:38 deletes too
16:38 chris right, thats the next tricky bit
16:38 kados once I verify it's working, we can release 2.3.0
16:38 chris sweet
16:39 we'll have to do some work on the installer
16:39 kados yea ... on that ...
16:39 I can't really see a way around doing configuration manually
16:39 especially for MARC21 vs UNIMARC
16:39 chris but once things are working we can do a tar ball
16:39 run the installer and work out what we can automate and what we have left to do manually
16:39 kados we almost need a UNIMARC template
16:40 because the attributes for, say, language search, will be different in a unimarc setup
16:40 chris right
16:40 kados as unimarc's fields are in different places
16:40 chris yep
16:40 kados someone with unimarc clients will have to become a zebra expert :-)
16:41 though maybe we could get thd to map between the two
16:41 he seems to have a handle on both
16:41 thd yes
16:41 kados hey thd
16:41 chris basically i think what we will end up having is sets of templates
16:41 kados thd: have you tried the latest instantiation of the zoomopac?
16:41 thd bad bad bad
16:41 chris and sets of system preferences
16:42 that people chose from when they install, to give them a basis, and then tweak from there
16:42 kados yea
16:42 chris we are getting quite a few templates
16:42 sorry system prefs
16:42 thd koha should be multi-MARC but that would break too many things for 3.0
16:42 kados thd: not necessarily
16:43 thd: no reason we can't be multi-MARC still
16:43 it just means we need some UNIMARC person to become an expert on how to index zebra based on unimarc settings
16:43 thd kados: what I mean is that Koha could have a multiple MARC frameworks for multiple MARC flavours and store them all at the same time with conversions between them.
16:44 kados that might work
16:44 but some things are incompatible
16:44 like subjects
16:44 and encodings
16:44 chris i dont see the point
16:44 but i might be missing something
16:44 thd kados: they are not incompatible but that is a secret
16:44 kados ie, there's no way to properly store a non-MARC8 non-UTF8 encoded UNINMARC record
16:45 hehe
16:45 anyway ... I'm far too tired to discuss MARC :-)
16:45 chris :)
16:45 kados only got about two hours last night
16:45 so I'm gonna head out for a drink and turn in early
16:45 thd kados: my z39.50 client does ISO-5226 to UTF-8 conversion
16:46 kados thd: ping me tomorrow about it, I am interested :-)
16:46 thd s/5226/5426/
16:47 kados: just look at it again.  I discovered PHP ISO 2709 class written in French that provided that simple conversion
16:48 kados: Of course Perl has MAB::Encode or whatever it is called
16:51 good night kados
19:24 mason
02:02 paul hello alaurin
02:03 alaurin hi
02:03 dewey privet, alaurin
02:07 paul hi pierrick
02:08 pierrick hi paul
02:08 btoumi hi all
02:08 paul et c'est à nouveau l'envahissement du canal par les français...
02:17 btoumi bon ca va alors je m'en vais si on est trop en plus y a un match
02:17 lol
02:18 paul un match ? de quoi ?
02:18 à cette heure ?
02:33 btoumi non mais faut ce preparer lol
02:55 hi toins
02:55 ToinS salut btoumi
02:56 btoumi toins:ca va bien
02:56 ?
02:56 ToinS très bien !
02:56 et toi ?
02:56 btoumi cav bien
06:02 chris hmmm
06:03 can i stay up till 3.30am to watch the football
06:03 i think the answer is no
06:18 Strait yippee! it works :)
06:18 finally
06:19 wasn't too easy though
06:19 chris ohhh you got it all installed? well done
06:19 Strait yep
06:20 as a regular user without root access to mysql
06:20 chris sweet
06:20 Strait i had to change the installer a bit
06:21 there could be an option "use already created database user... blah... blah... blah... in the installer"
06:22 chris yep, that could be done
06:23 Strait that would only do mysql -u $user koha < koha.mysql -p
06:23 chris yeah, then run the upgrade script
06:23 Strait should be fairly easy to do also :)
06:24 chris updatedatabase i mean
06:24 Strait yep
06:24 chris yep, just needs someone to do it :-)
06:24 Strait i'm not a perl expert, but i could give it a go
06:25 it doesn't look too difficult
06:25 chris cool
06:25 we have to do some work on the installer over the next few days
06:25 so ill take a look then too
06:26 its a bit late in the night for my brain to work good tonight :)
06:26 Strait 14.26 here ;)
06:26 chris 11.26pm friday here
06:26 Strait but i'll have to create users and branches to koha
06:27 and set up issuing rules and all that
06:27 chris right
06:27 btoumi hi all
06:27 chris hiya bruno
06:27 Strait i'll look into it next week
06:27 hello
06:27 chris cool thanks strait
06:37 btoumi hi chris
06:38 hi strait
06:45 chris yep
06:50 night all
06:55 Strait good night
08:04 paul kados around ?
08:08 kados paul: yep
08:08 paul about clonetag&subfield problem
08:08 it's not a html2marc problem
08:08 kados sure
08:08 paul it's a html problem.
08:09 because I even don't get the values in the CGI
08:09 I know.
08:09 i'm copying npl templates to default.
08:09 kados but I do know that thd has done extensive testing on the npl version
08:09 paul but I get a bug with authorities popup
08:09 kados could be ...
08:09 paul where is your openAuth function hidden ?
08:09 kados you'd have to ask owen
08:09 I think it's in a seprate js file
08:10 grep openAuth includes/
08:10 paul mmm... strange, it's nowhere.
08:10 kados hehe
08:10 maybe the problem :-)
08:10 paul do you use authorities with npl marc editor ?
08:11 kados yes
08:11 paul a link to a working marc editor pls ?
08:11 (npl one)
08:11 kados I noticed yesterday that the liblime demo is badly broken :-)
08:11 not the marc editor
08:11 but the whole system :-)
08:36 paul kados / owen : don't bother to port default changes I made to handle mandatory fields/subfields to npl
08:36 I did it ;-)
08:36 (hello owen)
08:36 but I still can't understand why MARC editor works fine with NPL & not with default templates...
08:40 ok, it works. I just have to tweak my css.
08:40 I think there is a minor diff in <div>s, that I missed
08:40 but that are important for cloning fields & subfields
08:41 yes, it has to be cleaned, you're right...
08:42 where are the styles defined in npl theme ?
08:45 owen intranet.css, colors.css, and maybe marc-editor.css I'm not sure if that last one is in use
08:46 paul thx
09:06 kados owen: got a sec?
09:06 owen Yes
09:07 kados owen: I'm wondering if you've been following the Mellon thread on koha list
09:07 owen Not closely, but yes
09:07 kados Brooke just responded to my post this morning
09:08 and I'm not sure she understands how the mellon works ...
09:08 owen I hadn't seen that one yet.
09:08 kados but I could use another set of eyes to confirm my understanding
09:10 owen I don't know about the Mellon rules, but I agree with your point of view that it would be better for a Foundation to manage the grant than a single company
09:10 kados here are the rules:
09:10 they include:
09:10 will recognize not-for-profit organizations that are making substantial contributions of their own resources toward the development of open source software and the fostering of collaborative communities to sustain open source development.
09:11 here are the criteria for the organization:
09:11 http://www.mellon.org/MellonGeneral.htm
09:11 which includes "evidence of suitable classification by the Internal Revenue Service"
09:11 I'm not sure a library fits the bill there, but it might
09:11 if so, maybe NPL could be the recipient
09:11 though I'm not sure you've got a staff member who could devote time to it
09:12 paul isn't owen a staff member :-D
09:12 owen I'd be the only candidate, and I don't know the first thing about grant-writing (and don't really have the time for it)
09:13 paul otherwise, i'm sure that ENSMP will be happy to spend time on this, if it is possible to have an NPO from outside of US
09:13 owen I'd take the job if that was our only option
09:13 kados cool
09:13 I think it's important for an employee of a non-profit to asnwer brooke
09:14 so either ENSMP or Owen :-)
09:15 owen I can write something today expressing my point of view. Essentially: I like the foundation idea, and I don't think NPL has the resources to manage the grant
09:16 kados right
09:16 brooke might not understand that there is no official organization that is 'koha'
09:16 owen In fact, it's the post-grant-writing part that I'd be most concerned about WRT NPL's involvement
09:16 kados yea
09:17 this is why we need the non-profit
09:21 owen The tough job, however, will be getting everyone's participation in the Foundation
09:22 We'll need involvement from everyone in order to make the grant really effective
09:28 kados well ...
09:28 IMO, the foundation will basically just act as a funnel
09:28 if we have some folks who are writing grants and specs for features
09:29 and then asking for koha companies to submit bids to do the work
09:29 I think it'll work quite well
09:29 then, as the foundation begins to establish itself
09:29 it can take on other roles
09:29 maybe even project management eventually
09:29 paul ENSMP has many ideas for OPAC & wanted to fund them this year, so we could ask pascale.
09:29 kados cool
09:30 owen So the foundation receives the grant, and then parcels out money based on requests by Koha users?
09:30 kados paul++ on the one tab!
09:30 owen: basically
09:31 owen: as you well know, requests by Koha users don't really come in 'specification' format
09:31 owen: so the foundation can parse those requests, and write them in software development terms
09:31 paul wow... kados reads mails very very quickly !!!
09:31 owen But who is "the foundation" ?
09:32 Who's doing that writing?
09:33 kados owen: what we're doing ... is forming the organization
09:33 owen: once it's formed, LibLime will make a contribution (probably)
09:33 owen: then the organization can hire a grant writer to write a given grant
09:34 owen: once the process gets started, it will be easy to maintain I think
09:35 owen When I said we'd need involvement from everyone I was just thinking in terms of: How do make the foundation happen without the financial burden being on Liblime
09:35 kados right
09:35 well ... there's another way the org can collect money
09:35 say a bunch of libraries in Ohio want a given feature
09:36 and they need a non-profit to represent them
09:36 as a group
09:36 owen Sure, absolutely.
09:37 kados but really, we're happy to contribute to the foundation, we can deduct it, and it's going to help foster the project as a whole
09:39 paul the indenting owen, just the identing
09:40 almost everything on the same line, with loops & if & unless deeply nested
09:40 i changes nothing at all.
09:41 owen Oh, in addbiblio.tmpl
09:41 paul yep
09:43 hehe... pierrick is publishing pictures of the kohaCon on le-gall.net
09:44 pierrick http://le-gall.net/pierrick/ph[…]php?/category/310
09:52 kados owen: can you change the 'button' for browse to a 'checkbox'?
09:52 owen: and no need to have more than one
09:52 i think
09:53 owen Oh, I see... just have the check box on the first one and they can chage the drop-down if they want to browse by something else.  That makes sense
09:53 kados yea
09:56 I can see expanding how this works in the future, but for right now, I'd just like to see it work period :-)
09:56 and I wrote ID about the query validation question
09:56 I think we could probably announce it to NPL staff on Monday
09:57 as a 'beta'
09:57 owen I thought we might have Ken take a crack at it even before that. What do you think?
09:58 kados definitely
09:58 owen Oh, never mind. He's off on Monday and Tuesday.
09:58 He's working tomorrow, but I don't know if he'll have time for it
09:59 johnb Been away from my desk, just saw the conversation about Mellon.  A few thoughts clarifications:
10:00 1.  Most libraries that are set up outside of government control our non-profit and our recognized by the IRS as 501c3's
10:01 2.  The organization that receives the grant usually takes a percentage of the total grant as an administrative fee usually 5 to 10 percent
10:02 3.  Administrating grants is usually not a full time position unless the grant is very large (7 figures plus)
10:03 4. There are individuals who can help write grants, or you can go to an organization with a proven track record
10:04 kados good points johnb
10:04 johnb: would CCFLS like to be the recipient?
10:04 johnb: you'd have my blessing :-)
10:05 johnb 5.  The important thing to remember about grants is that the grants are usually given to organizations that have a proven track record not necessarily the best written grant,
10:05 CCfls is always looking for funding through grants.  
10:08 We would be glad to help out in anyway
10:09 paul kados : is there a reason why cloneSubfield insert a subfield BEFORE the existing one ?
10:09 ppl usually enter most important values 1st, to a clone would be better AFTER imho
10:10 kados paul: there is a reason
10:10 paul: let me try to remember
10:10 it has to do with how the '+' follows the subfield I believe
10:10 paul maybe because there is no "InsertAfter" and it has to be emulated & taht don't work ?
10:10 kados yea
10:10 well ...
10:11 currently, it will insert the value too
10:11 so if you type:
10:11 200$aSome Author
10:11 and you click on '+'
10:11 you'll get:
10:11 200$aSome Author $aSome Author
10:11 so all that's left is:
10:11 delete the value from the original
10:11 and it works exactly as you want :-)
10:12 (if my memory serves me correctly, I haven't looked at the code in a while)
10:14 btoumi bye all good week end
10:18 osmoze bon week end btoumi
10:19 btoumi :osmoze de meme
10:23 paul bye btoumi
10:25 btoumi bye
10:37 kados owen: browse working
10:37 owen: well ... not fully yet ... but first time it actually returned results :-)
10:37 owen: doesn't distinguish between indexes yet either
10:37 owen I was just about to ask :)
10:38 kados owen: I've got the cql query working even with invalid CQL
10:39 owen: I just basically test if it's valid, and if not, I pass it on enclosed in quotes
10:39 I need users to tell me how good of a job it does
10:39 owen I would suggest 'and's instead of quotes I think.
10:40 kados well ...
10:40 that gets really complicated
10:40 because there might already be ands
10:40 owen right
10:40 kados and adding more will not make it valid CQL
10:40 so you have to submit the query sometime :-)
10:40 can't just keep adding ands and testing if it's valid :-)
10:42 owen: so do you have a sec to talk about how browse should work?
10:42 owen Yeah, I think so
10:43 kados do you like the way worldcat handles it?
10:43 how does spydus do it?
10:43 worldcat has a completely separate search tab for it
10:44 and they let you select an index
10:44 then, depending on which index you selected, it'll do a keyword search on that index
10:44 when you click on the term
10:45 hey ... I just had an idea
10:45 what if all it does, is turn on the 'livesearch'
10:45 so you get a dropdown box with the index
10:46 the livesearch is demoed here:
10:46 http://labs.liblime.com/
10:48 owen Actually, take a look at O.U.'s library search, and try an author or title search.
10:48 They do browse by default
10:49 kados that's not browsing the index
10:49 those are authorites records
10:49 they seem similar
10:49 owen All I can tell you is how I think it should work :)
10:49 kados but really are quite different
10:50 yea ...
10:50 SMFPL is sponsoring that for Koha
10:50 so it will work exactly as it does for OU
10:51 this browse is specifically for the indexes in Zebra
10:51 let me illustrate it
10:51 try:
10:51 'The Handmaid's Tale"
10:52 since by default, it's doing a browse of the 'word' indexes
10:52 it'll only find things with 'the*"
10:52 if we did a phrase browse
10:52 it would pull from the 'phrase' index
10:53 same with the number index
10:53 etc.
10:54 owen keyword browse for 'The Handmai'd Tale' ?
10:54 paul huray for ToinS !
10:54 cloneTag & authority search problem solved !
10:54 (even if the DOM is sometimes strange... :
10:54 kados woot
10:55 paul passing a "this" on a <a> means passing the href, and impossible to reach the parent, while passing this.parentNode works as expected
10:55 )
10:55 kados owen: don't confuse 'indexes' with 'bib1 USE search points'
10:56 owen I'm confused by your telling me not to be confused :)
10:57 kados hehe
10:57 sorry, I'm multitasking
10:57 didn't mean to sounds snippy
10:58 ok ... back to single-tasking
10:58 so ...
10:58 zebra has indexes that it queries when it's searching
10:58 it has one for 'words' one for 'phrases' one for 'numbers' one for 'URLs'
10:59 all 'scan' does, is return values from the index
11:00 does that make sense?
11:02 owen Yes...but I'm not sure that can be turned into a usable search given the examples you've shown me
11:02 The "keyword (count)" results examples seem non very user-friendly to me
11:12 kados I wonder if the worldcat ones are user friendly
11:12 right
11:13 ok ... so if you do a keyword index browse
11:13 for 'the handmaid'
11:13 it works the same way as zebr
11:13 zebra even
11:13 but if you do a 'title phrase'
11:13 it is more useful
11:18 being able to scan an index is, I'm guessing, an advanced feature that mainly librarians and researches would find useful
11:18 it's a way to validate whether a given term or phrase exists in the database
11:18 and see words/terms nearby
11:19 if it's ok with you, I'd like to implement it as another tab
11:19 osmoze bye all
11:19 kados or ... at least as a separate form on the power search
11:20 bye osmoze
11:20 thd bye osmoze
11:20 kados hey thd
11:20 thd yes
11:20 kados thd: I've got some time to talk today when you're ready :-)
11:20 owen how did you do the title phrase search on worldcat?
11:20 kados thd: did you see the new browse feature?
11:20 owen: click on the icon next to each field in the advanced search
11:21 owen: then from the 'indexed in' dropdown select 'title phrase'
11:21 thd browsing is not real browsing if you start by searching but maybe I am also looking at the wrong URL
11:21 kados: what is the URL
11:21 ?
11:21 kados thd: zoomopac.liblime.com
11:21 yea, it's not real browsing
11:21 it's actually 'scanning'
11:22 we should probably call it scanning
11:22 'scan index' or something
11:22 thd: go to the advanced search
11:22 thd: and click on the checkbox for 'browse'
11:24 thd kados: I have a could not be found error from zoompac.liblime.com
11:24 kados hmmm
11:24 thd: http://zoomopac.liblime.com/
11:24 thd: it should work, working for me
11:25 thd kados: I assume it is also working for owen?
11:25 kados yea
11:25 :-)
11:26 paul ok guys, bye bye, see you on monday !
11:26 paul_away (do we have a meeting on monday ?)
11:26 thd kados: maybe it has not propagated fully
11:26 kados thd: you sure it's not misspelled in your browser window?
11:26 thd good evening paul_away
11:27 kados: when is the next meeting?
11:27 kados thd: probably wednesday at 20:00 GMT
11:28 owen thd: I've been using that zoomopac link for several days now
11:29 thd kados: I have http://zoompac.liblime.com/ copied from my URL textbox.  How is it misspelt?
11:29 kados hehe
11:29 strange
11:29 maybe your DNS is not working
11:29 owen http://zoomopac.liblime.com
11:29 zoom opac
11:30 Not zoom pac
11:30 kados ahh
11:30 yea that's it :-)
11:33 thd owen: so if kados is gone for 10 minutes, tell me what advantage this 'browse' feature might have?
11:34 owen I'm still waiting for kados to sell me on the idea. A browse search as I understand it is different from the kind of search he's offering with zebra
11:34 I understand a browse search as one that will let me see my search term in the context of a list of nearby search results
11:34 thd own: yes that was my point so perhaps the name should be modified.
11:35 owen Were you able to try the browse search on zoomopac.liblime.com?
11:37 thd owen: yes, Is this actually meant to return a brief identifier for selecting an actual term to use as the '...' search does in 2.2 now?
11:38 owen That's the way it worked in the WorldCat search Joshua showed me
11:38 ...but I don't know what his plans are for our version.
11:40 thd owen: does WorldCat FirstSearch have a checkbox labelled browse?
11:43 ToinS bye all !!!!
11:43 thd goodbye ToinS
11:54 kados: are you back?

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

koha1