IRC log for #koha, 2006-06-14

All times shown according to UTC.

Time S Nick Message
13:10 kados hey tumer
13:11 kados tumer: did you see the 'scan' option now?
13:11 tumer hi kados
13:11 tumer i didnot
13:11 kados tumer:
13:11 kados tumer: go to 'power search'
13:11 kados tumer: type in a search and click on the 'scan the index' checkbox
13:11 kados well ... 'scan indexes' actually
13:12 kados you can specify whether to show a complete subfield with the completeness attribute
13:12 tumer it sets a local ip number i cannot connet to zoomopac
13:12 kados and you can limit by field, etc.
13:12 kados ?
13:12 kados what sets a local IP?
13:12 tumer
13:13 tumer
13:13 kados tumer:
13:13 kados that IP will work too
13:13 tumer not connected yet if it dows i'll give it a try
13:14 kados k
13:14 tumer it seems you exelled in this zebra i cannot catch up with you
13:14 kados there's more to come :-)
13:14 kados I'm looking into doing asynchronous OPAC queries
13:15 tumer why?
13:15 thd kados: I verified that attributes are not specified to function any differently with scan or search
13:15 tumer check zebraop in it dows asynch
13:16 kados tumer: cool ...
13:16 tumer we need async for Z3950quee procedures
13:16 tumer you commtted something for that i believe
13:17 kados all I committed (to rel_2_2) was an update of processz3950queue to work with ZOOM
13:17 kados it needs to be completely re-written to use ZOOM properly
13:17 tumer my internet connection from proxy is broken i cannot see anything yet
13:18 tumer kados:we decided to remove everythýng from koha regarding old db. MARC and zebra only from now on
13:19 kados tumer: who's we?
13:19 tumer NEU
13:19 kados wow ...
13:19 kados that's great
13:20 tumer is it good for you as well?
13:20 kados well, I think so
13:20 kados not for dev-week
13:20 kados but I think that makes sense for 3.0
13:20 kados will you still store the marc in mysql?
13:20 tumer yes and only the marc record
13:21 kados cool, I think that's the best approach
13:21 kados and I think we should separate holdings from biblios
13:21 tumer currently its working like that except items
13:21 kados have two separate zebra dbs for them
13:21 kados so we'd have:
13:21 kados biblios
13:21 kados items
13:21 kados authorities
13:21 dewey i heard authorities was authority records
13:21 kados reservoir
13:21 dewey i think reservoir is identical to biblios but not the actual catalog
13:22 kados those four would be the zebra databases
13:22 tumer reservoir is good as it is
13:22 kados I disagree
13:22 kados because it is impossible to find anything
13:22 kados unless you know the exact title
13:22 kados better to put it in zebra
13:23 tumer authority i committed like that anyway
13:23 tumer k for reservoir
13:23 kados yep ... though I haven't tried it yet
13:23 kados I've also been talking to ID about the default 'simple search'
13:23 tumer any news
13:23 kados and they are going to expose the CCL query stuff in Yaz
13:24 tumer ccl or CQL
13:24 kados which will be a more user friendly default
13:24 kados CCL
13:24 kados it will accept:
13:24 kados harry potter
13:24 kados whereas in CQL you have to go:
13:24 kados "harry potter"
13:24 kados or
13:24 kados harry and potter
13:24 kados CQL will still be an option
13:24 kados I'd like to expose all the supported query languages
13:25 tumer cool
13:25 kados and let libraries (and patrons) choose which they want to use
13:25 kados anyway, as soon as the CCL stuff is done (next day or so)
13:25 kados I need to double-check that adds/edits are working
13:25 kados (think I need to modif this db a bit to enable them)
13:25 kados and I'd like to test authorities after that
13:26 kados once it's stable
13:26 kados I'll release 2.3.0
13:26 tumer by the way i noticed you keep changing cql to cqlrpn query all the time any reason?
13:26 kados heh ... sorry :-)
13:26 kados i was experimenting with trying to defive the RPN from the CQL2RPN call
13:26 kados no luck yet
13:26 tumer they are exactly the same search you know
13:27 tumer the only differnce is:
13:27 kados yea ... but the difference is that my interface needs to have access the the RPN
13:27 kados so we can pass it on for merging queries, resorting, etc.
13:27 kados another thing I need to ask ID about :-)
13:27 tumer they both return the same result except:
13:27 kados anyway ... I'm swamped today
13:28 kados go ahead
13:28 tumer if i want to reach your zebra but use my own i use cqlrpn
13:28 tumer if i wqnt to leave it as you defined it i us e cql
13:29 tumer so with cqlrpn i can define my own
13:29 tumer with me?
13:30 tumer so if i do cql search server looks to find the automatically from <cql2rpn> definition
13:31 tumer otherwise i have to tell where the is during conntection and us e cqlrpn
13:31 tumer internally oth turned into pqf
13:31 tumer thats it
13:31 kados I see
13:32 tumer and found some info about ports
13:32 tumer 210 for z3950
13:32 tumer 346 for zebra
13:32 tumer universally accepted
13:33 tumer its just a note
13:33 kados cool
13:37 tumer do you know any good reading about holdings marc apaprt from LC web site?
13:38 tumer how to implement it?
13:38 kados yes
13:38 kados hang on a sec
13:38 kados
13:38 kados click on 'Holdings'
13:38 tumer thanks
13:38 kados that site also has good info on Authorities
13:41 tumer
13:42 tumer  is working now wher do i look
13:44 kados go to power search
13:44 kados and use the little checkbox for 'Scan Indexes'
13:45 kados experiment with completeness attributes and with phrase vs word
13:45 tumer very nice
13:45 kados scanning the indexes might be an interesting way to provide minimal authorities headings for people without authorties
13:46 tumer clicking on count can take us to biblios of those word probably
13:46 kados in fact, it already works :-)
13:47 tumer so it does excellent
13:47 kados yup
13:48 kados and it's committed already, so you can play with it :-)
13:48 kados I don't have time to code today ... working on another project
13:48 kados but I'm eager to get back to it :-)
13:48 tumer great work and you say you didn't know programming?
13:48 kados it's been a lot of fun to work on once I got the basics of Zebra down
13:48 kados yea, no formal programming :-)
13:49 kados I just picked it up along the way :-)
13:50 tumer hear hear
13:50 tumer and some keep learning without ever lraning
13:51 tumer anyway since internet is back i better start downloading the new commits
13:52 tumer bye all
13:52 kados have fun :-)
13:52 kados bye tumer
15:12 rach howdy
15:28 thd hello rach
15:31 thd kados: I have an item that cannot be deleted because it could not have been created.  A phantom item.
17:54 thd kados: are you any less busy now?
17:54 thd kados: I found the origin of the bug I had been reporting
17:56 thd kados: if you duplicate a field the duplicated record loses the a in 000/09.
17:57 thd s/duplicated/reloaded/
02:10 btoumi hi all
02:16 osmoze hello
02:16 btoumi hi osmoze
02:35 btoumi :paul are u here?
02:37 paul yep.
02:37 paul salut btoumi
02:41 btoumi :paul salut
03:13 thd good morning paul
03:15 thd paul: are you there?
03:15 paul (on phone)
03:16 thd paul: let me know when you are off phone, I have two simple questions
03:22 paul thd : i'm back
03:23 thd paul: maybe 3 questions
03:23 paul I couldn't ask you if your trip to spain was nice, after the devWeek. So, feel free to answer this question too ;-)
03:24 thd paul: first I committed my expanded MARC 21 bibliographic framework
03:24 paul yes, I saw them, thx
03:25 thd yes my trip to Spain was very good and there may be a possibility for Koha there, I am investigating
03:26 thd paul: I did not use the verbose column naming insert convention but used the column sequence only convention
03:26 thd paul: the same convention I had used for editing the framework
03:27 thd paul: this is a little less reliable but would there be any problem in this case?
03:27 paul I don't think so, except with mySQL 5.x
03:27 paul but there already are many problems with mySQL 5, so ...
03:28 paul mmm...
03:28 paul no, sorry, i misread your message.
03:28 paul there's no problem exporting without complete sql insert, except when there is a database change.
03:28 paul but in 2.2.x there is none.
03:28 Comete hello
03:28 paul that may be a problem for 3.x
03:28 paul hi Comete a
03:29 btoumi hi comete
03:29 thd paul: I could have substituted the more verbose form with a regex but I feared a mistake in my regex might do more harm
03:30 Comete is there a way to keep the first enrolled date of a borrower AND the last one ?
03:30 thd paul: 3.0 is the same so far and future revisions can always add columns
03:31 paul right.
03:31 paul Comete: not in koha 2, ask btoumi for koha 3
03:31 thd paul: how would this be a problem for MySQL 5.X specifically?
03:32 Comete btoumi: is there a way to keep the first enrolled date of a borrower AND the last one ? in koha 3
03:32 paul I misread your question, so there is nont
03:32 paul none
03:32 btoumi :Comete Yes
03:32 Comete youpi
03:32 Comete :)
03:33 btoumi ;=)
03:34 thd paul: I have another SQL script for applying the MARC 21 default framework to an existing installation.  It is not a very clever script so it should not be run automatically.
03:34 paul put it in misc, with a proper documentation.
03:34 paul if it can be used, then it can be published. even if it is with a lot of care !
03:35 thd paul: It works fine but it would be very rude to run it automatically on an unsuspecting user.
03:36 thd paul: It has many comments for documentation
03:36 paul to solve this problem I usually :
03:36 paul - add a -c flag to "confirm that I want to run this script"
03:36 Comete does anyone read my post on the wiki fr ? there are no comments at the moment...
03:37 paul - just show help when the script is called without any parameter
03:37 Comete i would like to know if these features already exist
03:37 thd paul: can I do that for an SQL script?
03:37 paul Comete: you should ask once again on infos@ mailing list ;-)
03:37 paul (& i will answer wherever I find some time)
03:38 paul mmm... no, sorry. If it's pure SQL, you can't
03:38 Comete paul: ok thank you
03:39 thd paul: It does have a very large warning with a big stop now menage in a comment about what it will do.
03:39 thd s/menage/message
03:39 paul that should be enough
03:41 thd paul: perhaps your recent fix to the MARC record editor has prevented fields from being duplicated in the NPL templates?
03:41 paul I can't test everything, sorry :\
03:41 thd paul: I leave that for owen and kados to sort out :)
03:42 thd paul: at least it no longer corrupts the MARC 21 records when the field is duplicated
03:43 thd paul: I did notice that in the default template it duplicated the field content along with the field.  Is that a bug or a feature :) ?
03:43 paul good question...
03:44 paul some think this is a bug, other this is a feature
03:45 thd paul: as long as the record is not corrupted the duplicated content can always be overwritten.
03:47 thd paul: I discovered yesterday that the former problem I was having with MARC 21 records was that every time I duplicated a field leader position 9 would change from Unicode to MARC 8 and then MARC::XML would mange the record upon saving.
03:48 thd paul: thanks for fixing that.
08:48 kyle_away /nick kyle
10:25 thd kados: are you any less busy today?
10:26 btoumi by all
10:29 thd owen: are you there?
10:29 owen yes
10:29 thd owen: did you see my newly reported bug?
10:30 owen bug 1104?
10:31 thd yes
10:32 thd owen: I think this may be in consequence of what paul had done to fix other problems with the editor.
10:33 thd owen: so the editor may work ok now for the intranet templates but something needs to be done for the NPL templates
10:34 thd s/intranet/default/
10:34 owen So the duplicate field function works in the default templates but not npl's?
10:34 thd yes
10:36 thd owen: please note an earlier bug where leader position 9 would change from an 'a' to blank on field duplication.  That resulted in record corruption after saving.
10:36 owen Any idea of the cause of that?
10:37 thd owen: do you mean the change in the leader position 9
10:37 thd ?
10:37 owen Yes
10:39 thd owen: Maybe MARC::XML was not recognising the presence of the 'a' for Unicode and resetting it to blank for MARC 8.
10:39 thd owen: the consequences would be bad for scrambling the subfields where there were more than two subfields in a field.
10:40 thd owen: The scrambling was evident after saving.
10:41 thd owen: that is the unreported bug referenced in bug # 1104
10:41 owen Hi shedges!
10:41 shedges hey owen
10:42 thd owen: with paul's fix and certainly with no field duplication that bug is no longer manifesting.
10:44 thd owen: without field duplication, proper records cannot be created in most instances.
10:49 thd owen: are you still there?
10:50 owen yeah, but I'm bouncing between a few different tasks
10:53 thd owen: although it is much less important, It would seem very useful to have the preference system read the directories and show the available CSS options as it does with the available template options instead of forcing the user to hunt for them.
10:53 owen I agree.  kados and I have talked about it, but haven't had time to make it happen
10:54 thd owen: what is kados doing now? This editor bug is very serious.
10:57 thd owen: he does now have customers using the editor, unless they have adopted something else in the face of multiple bugs.
10:58 thd tumer: are you there?
10:58 tumer hi thd
10:58 owen thd: I am hardly ever in direct contact with kados. We communicate here.
10:58 owen So I don't know where he is either
10:59 thd owen: I know kados is very busy, I suspect5t with whatever he was busy with since midday yesterday.
10:59 tumer thd & owen: i use npl templates with a different fix
10:59 thd tumer: really do you have your own fix for the record editor?
11:00 tumer before pauls commits i changed the java in npl templatws and it seems to work
11:00 tumer just comment everything regarding trigger and it seems to work foe me
11:01 thd tumer: maybe you could even commit those changes
11:01 tumer i did not commit them as i was still testing them but pauls commits came so i stopped
11:02 tumer i cannot reproduce the bug you are talking about though
11:03 tumer my problem was that i could not repaet a field more than once, now i can
11:03 thd tumer: paul fixed some problems but now repeatable fields cannot be added for the NPL templates only.
11:03 thd tumer: I cannot repeat a field any times now with the NPL templates
11:04 thd tumer: this is now bug # 1104
11:04 tumer i dont know what paul has fixed but i did not implement it as i did  not have a bug
11:05 thd tumer: are you running a forked bug free Koha or do you not use the record editor vigorously enough to notice the bugs?
11:05 tumer probably rather than doing an official commit i'll pass the changes to owen so he can test them
11:06 thd tumer: whatever solves the problem quickly.
11:07 tumer we use the editor vigorously but a stable version of koha
11:07 thd tumer: paul is preparing to issue a release while the NPL templates are not working for the record editor now.
11:08 tumer i'll have a quick look
11:08 thd tumer: you miss out on all the excitement by using a stable version.
11:09 tumer stabilised tumers version
11:12 owen When I look at the fresh-from-cvs copy of, I don't see any tabs.
11:13 thd owen: you mean tabs for groups of fields?
11:13 owen Yeah
11:14 tumer thd :just committed the fixed template
11:14 tumer test it whether it works
11:14 thd owen: well you only have whatever tabs your framework defines
11:14 owen My framework hasn't changed, as far as I know.
11:15 thd owen: do you mean that all fields appear in one long form at the same time?
11:15 owen The source shows all the right form fields for all tags, but the tabs don't appear
11:15 owen thd: should they?
11:16 tumer owen: paul has put them in a tabloop now i did not test it yet though
11:17 thd owen: one long form had not been historic behaviour but I had thought that it might be faster for me to use.  Yet I never suggested it to anyone.
11:17 owen I thought he was just referring to how the script passed the data to the templates...
11:17 tumer he also changed the default.tmpl it has to read a loop now (i think)
11:19 owen I see the tabs in the default template, but they don't work :)
11:19 thd tumer: does your record editor use one long form or are tabs needed for navigating to different field groups?
11:19 tumer tabs are there
11:19 tumer i did not update to pauls and
11:20 thd tumer: does your record editor use one long form with all fields appearing simultaneously with the tabs as an additional navigation feature.
11:20 thd ?
11:21 tumer editor has tabs, field appear acoording to tabs
11:22 thd owen: so tumer did not knowingly introduce such an innovation in your checkout
11:23 thd owen: my problem with the tabs is that there is poor or no keyboard only navigation to the tabs.
11:24 owen I'm not fond of the tabs myself. When I first designed the npl templates I took them out. Then the catalogers asked for them back. Now they're asking for them to go away again :)
11:24 thd owen: every time I need to touch the mouse keys I am doing something much slower than necessary.
11:25 thd owen: I think the tabs are fine if they had a good keyboard shortcut.
11:25 tumer thd: tabs and field duplication are 2 different issues. Paul did a commit that changed both features. I did not use it. All i am
11:26 thd owen: I did not think you had any cataloguers using the record editor for anything other than updating holdings.
11:26 tumer saying with old NPL templates my fix works for duplication for me
11:27 thd tumer: I am aware that they are different.  Tabs are a relatively unimportant user interface issue compare to adding repeated fields.
11:27 thd s/compare/compared/
11:27 tumer sure
11:30 thd tumer: I want to send you a URL without having the whole world bombard my poor machine without enough memory or bandwidth, etc. to accommodate them.
11:30 tumer
11:30 thd tumer: do private channel IRC messages not work on your IRC client?
11:31 tumer should do i think never used it
11:52 thd tumer: is there something frustrating about asynchronous mode?
11:52 tumer frustating for you on a dial up
11:53 tumer to update your Dbian i mean
11:53 thd tumer: oh yes, I live in ancient building with ancient wiring.
11:54 thd tumer: I update very infrequently for that reason.
11:54 tumer thd have you heard of kados today
11:54 tumer my utf8 problem still persists with this new XML as well
11:54 thd tumer: kados seems to have been completely unreachable since you communication with him yesterday,
11:55 tumer i believe he is deep in ZEBRA
11:55 thd tumer: To what UTF-8 problem are you referring?
11:56 tumer with this new MARC-XML my characters (template only) get scrumbled
11:56 tumer i get wid-character-in print warnings from perl
11:56 tumer s/wid/wide
11:57 thd tumer: That error message can be avoided by informing Perl that you will be using wide characters
11:58 tumer yes but on screen they are scrumbled as well?
11:59 thd tumer: unless they are really scrambled that is a font problem in the CSS for which I fell asleep trying to prepare commits a couple of months ago.

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