IRC log for #koha, 2008-08-09

All times shown according to UTC.

Time S Nick Message
12:22 hdl kados ?
12:32 kados hi hdl
12:32 kados hdl: what's up?
13:06 gmcharlt greetings #koha
13:07 owen Hi gmcharlt, everyone
13:12 hdl kados : sorri.
13:14 hdl tried to find the same problem on a USMARC with similar items on.
13:14 hdl But no way.
13:17 cnighs g'morning
13:25 hdl kados : i sent you a link for bug 2410 so that you see the bug.
13:26 hdl Can't reproduce it on a USMARC install. (don't know why)
13:26 kados hdl: OK, thanks ...
13:27 hdl sorry for delay.
13:27 hdl It is quite hard for me to find good sample data for that.
13:35 hdl kados : was Paul's proposed patch for bug 2440 accepted and pushed ?
13:39 owen atz?
13:39 atz ?
13:42 owen Hi atz, I'm looking at your tabindex bug report
13:42 owen I'm not getting any validation errors from addbiblio. I wonder if it depends on the framework?
13:43 atz did you do a search?
13:44 owen Ah, just did. I see them. I thought you'd turned them up by validating.
13:45 atz i think they should probably just be tabindex=0
13:45 atz meaning, don't stop here on tab
13:48 hdl owen : why is language selection list not sticked to bottom of the page but going up and down ?
13:48 hdl http://o14.hdlaurent.paulpoula[…]koha/
13:49 hdl seems there is a missing div
13:49 hdl Do you have the same behaviour ?
13:52 owen hdl, we changed the way the language-selection worked in the opac
13:52 owen It no longer "sticks"
13:52 hdl was it a demand ?
13:53 owen It was a request from the release manager :)
13:53 hdl Or was it because there was problems on some persons ?
13:53 hdl pages
13:54 hdl ok so rm is right. But it looks strange to me.
14:44 hdl kados or anyone on : It seems to me that usage of HomeOrHoldingBranch has not been propagated to all the functions.
14:48 pianohacker hdl: Which one in specific?
14:50 hdl line 1443
14:50 hdl in
14:52 hdl AddRenewal too.
14:52 hdl since it takes item homebranch and not CircControl
14:52 hdl branch
14:55 hdl Maybe we should add an internal function Getissuebranch and use it through the code. same for itemissuingbranch.
14:57 ryan hdl: where's the script that updates bibs with modified authorities ?
14:58 hdl ./misc/migration_tools/
15:00 ryan hdl: that looks like it requires two authority records ?
15:00 hdl ryan: not in batch mode.
15:00 hdl It takes the authorities recorded in local_file repository
15:04 ryan hdl: did this script get lost?
15:04 ryan hdl: the version i'm seeing is very 2.2
15:06 ryan afaik, we no longer have a marc_subfield_table.
15:06 ryan kados: , gmcharlt , anyone else ?
15:08 gmcharlt ryan: ack, it does look like it has not been updated for 3.0
15:09 gmcharlt use of MARCdelsubfield is antiquated as well
15:10 hdl ryan : I was looking if I could propose a proper version for it.
15:10 hdl Unfortunately I can't.
15:11 ryan hrmm.
15:11 ryan Anyone volunteer to rewrite merge_authorities today ? :)
15:11 hdl ryan: I can update it.
15:15 ryan hdl: do you want me to file a bug for it?
15:15 hdl pls do
15:15 ryan ok, will do.
16:53 kados hdl: you here?
16:53 hdl yes
16:53 kados hdl: do you know the status of bugs 2440 and 2351 ?
16:54 owen I've got one for you kados, I noticed you changed the status of Bug 2344 from patch-sent to p3. Did my patch not solve the problem?
16:54 kados hdl: ryan submitted a patch and a note about paul's patch ... I want to make sure it resolves your usage
16:55 kados owen: ahh, does your patch provide a new patron button or a link when yui and js are missing? if so, I missed it
16:55 kados owen: if it can be closed out by all means mark it FIXED :-)
16:56 owen Yeah, I added a form for non-js users that gets removed when the js toolbar runs
16:56 kados cool
16:56 owen I just wanted to make sure there wasn't a problem.
16:57 kados I thought Joe had set that patch-sent status
16:57 kados and since the last comment indicated there were still issues I figured it should be reverted
16:57 kados owen: just FYI, I rely on status patch-sent + a comment saying what the patch does ... some bugs have more than one patch sent throughout their life
16:57 hdl kados : paul reported he sent  a patch correcting this bug.
16:58 hdl Was there a problem with it ?
16:58 kados hdl: yes, but it conflicted with a prior patch from ryan, and ryan sent paul a message on patches asking for clarification
16:58 owen Okay kados I'll make sure I do that
17:11 hdl MMM... Don't know what paul answered and what he thinks about that.
17:11 hdl Will test.
17:40 kados hdl: so far paul didn't answer
17:40 hdl OK.
17:41 kados hdl: thanks
17:44 cnighs kados: iirc, paul mentioned he would not have internet access until he returned at the end of August
17:48 kados cnighs: *nod*
17:48 kados anyone tell me what the meaning of this regex is:
17:48 kados /(\P{IsAlpha}$_\P{IsAlpha}|^$​_\P{IsAlpha}|\P{IsAlpha}$_$)/
17:51 kados the context: $_ is a stopword, and the value being tested is the search string
17:52 kados it seems to fail where $_ == 'the' and the value being tested == 'the'
17:55 kados ' the ' rather
18:06 cnighs kados: it appears that the regexp is looking for $_ with non-alpha chars either on both sides, or following or preceding (I think)
18:06 atz kados: if unquoted, that might be good
18:07 atz it's looking to see if it has alpha characters on both sides, or on the front (next to line start ^), or on the back (next to line end $)
18:09 atz from perlreref,  
18:09 atz \pP     Match P-named (Unicode) property
18:09 cnighs atz: so \p{IsAlpha} would be looking for non-alpha?
18:10 atz cnighs: no, you were right on that
18:10 atz \P is the inverted version
18:11 atz i think our search parsing is not very articulate though
18:12 atz at least it is Unicode aware though
18:24 kados /(\P{IsAlpha}$_\P{IsAlpha}|^$_\P​{IsAlpha}|\P{IsAlpha}$_$|^$_$)/
18:24 kados that fixes the problem
18:24 kados by testing for the exact stopword
18:39 mc hello world
18:40 mc is it possible to find repeatable subfields
18:40 mc ?
18:41 mc something like 600$aFrancoise$aFrederique$aAnne$bRigal
18:41 mc ?
18:43 gmcharlt mc: find in zebra?
18:47 mc oops
18:47 mc no; sorry
18:47 mc in a real word notice
18:47 mc i mean: can this case appears
18:48 gmcharlt that paricularly case, not in the MARC21, but repeated subfields are common in general
18:49 mc ok
18:49 mc thanks
18:49 liz-nekls you know, it sure would be nice if the design preferences were separate from the operational preferences
18:50 gmcharlt liz-nekls: in the system preferences interface?
18:50 liz-nekls yea
18:50 atz good point
18:50 mc and can it be ordered with other subfields? ( i mean $a$b$a$b don't mean the same than $a$a$b$b
18:51 mc )
18:51 mc ?
18:51 owen What can we say, liz-nekls, they're better-organized than they used to be :)
18:51 liz-nekls hehe
18:51 liz-nekls oh I'm not really complaining
18:51 liz-nekls it was just a thought :D
18:53 gmcharlt mc: yes.  in MARC21, it happens a lot in subject headings, where the main heading in $a can be followed by multiple subdivisions coded $x, $v, $y, and $z and various orders
18:53 mc in fact: can i store [ 700 => [ a => [qw( Indiana Junior )], b => 'Jones' ]] or something like that
18:53 mc arghh
18:53 mc thx for reply, gmcharlt
19:01 mc 'night koha people
19:01 mc (have a good week end too)
19:02 gmcharlt g'night, mc
19:42 liz-nekls anyone here know anything about CKLS?
19:43 liz-nekls like, which version of Koha they are running?
19:47 kados liz-nekls: 3.0
19:48 liz-nekls kados: tyvm
20:15 hdl kados : about the bugfix 2485 I sent, can you detail ?
20:16 gmcharlt hdl: there are references to $item->{itemtype} that would fail
20:16 hdl :[
20:17 hdl mc is right when he is asking for abstraction levels.
20:18 hdl I guess that Having itype and itemtype, will be a nightmare for maintainance.
20:18 kados hdl: for 2351, I have a patch that prevents duplicate check for empty barcodes
20:18 kados hdl: does that resolve it adequately for you?
20:19 kados oops, I mean 2440
20:19 gmcharlt hdl: fully agree.   item-type-qua-issuing-rule-type needs to be distinguished from item-type-qua-material-or-format-type
20:20 kados hdl: i tried applying paul's patch but I still don't see a way to add no barcode and no item ... you have to create an item afaikt, but now you can have empty barcodes, does that make sense?
20:21 hdl adding fake items makes more sense than adding none ?
20:21 hdl Do you really want me to agree on that ?
20:21 kados hdl: it's paul's patch, I'm asking you if you think it resolves the issue
20:23 hdl when was it sent ?
20:23 kados hdl: or maybe I don't understand how paul's patch resolves it
20:23 kados originally on Jul 31
20:23 kados fix for #2440 : acquisition recieve & item creation
20:24 gmcharlt bbiab
20:25 kados hdl: it appears acqui receive still works, even multi-receive, so I'm prepared to submit this patch from paul
20:26 hdl kados : could bc be defaulted to space if no input ?
20:27 hdl I think this would really close the problem
20:27 kados hmmm
20:27 kados could you clarify what you mean?
20:28 hdl oh sorry.
20:28 kados with paul's patch, if barcode is empty, no item is added
20:28 hdl OK for your patch.
20:28 kados if it's filled, an item is added
20:28 kados but without my pach you can't submit the form with empty barcodes for more than one order
20:29 hdl I misunderstood.
20:29 kados Ok, I've pushed it up
20:29 kados I'll ahve nicole update the docs
20:49 hdl gmcharlt: I was saying that maybe we should find and design a kind of abstraction for Biblio/Work and biblioitem/Book. So that with one name, we call the entity required at the moment.
20:50 hdl I think that tumer did a job on that.
20:51 hdl He wrote a schema (and also designed some zebra files) that could fit.
20:51 hdl Maybe I am day-dreaming.
20:53 kados hdl: when do you plan to have that ready?
20:53 kados hdl: I want to release this evening if possible
20:53 kados but I'd be willing to delay until Monday if necessary
20:53 hdl I try to do as fast as I can.
20:54 hdl BUT I want to do some Unit test.
21:47 pianohacker What is items.stack?
21:48 pianohacker And, for that matter, .coded_location_qualifier ?
23:11 hdl gmcharlt: Does mysql with Innodb do automatic locking on tables ??
00:30 gmcharlt hdl:: it can
00:30 gmcharlt hdl: just saw your comments on abstractions - I'd be interested in seeing tumer's schema
00:31 hdl Maybe he has ctweaked it since.
00:39 hdl has been there  :
00:42 hdl was on the logs of 11/08/2007
00:42 hdl 2006.08.11 rather
00:43 gmcharlt yeah
00:43 gmcharlt internet archive doesn't have a copy of that page, alas
00:49 hdl three files sent.
00:49 hdl I had them on my disk.
00:53 gmcharlt hdl: thanks
00:54 gmcharlt hdl: similar to some ideas I have rattling around my head
00:54 hdl similar to ideas many have in head.
00:54 hdl But time runs short.
00:55 gmcharlt indeed
00:55 hdl (it is so old... 2 years old.
00:55 gmcharlt still, I'm hoping that we can move in some of these directions for 3.2
00:55 hdl It really is a pity)
00:55 gmcharlt yes - schema would have to be revamped
00:55 gmcharlt e.g., no need to redo MARC21slim
00:56 gmcharlt and native Dublin Core and MODS support, maybe even EAD, would be considerations now
00:56 hdl Yes but Unimarc Support is still a musthave  for us.
00:57 hdl We could do a transform from koharecord to UNIMARC.
00:57 hdl though.
00:57 hdl It could be less painful.
00:57 hdl and more sustainable.
00:57 gmcharlt I meant MARC21slim the XML schema, which works fine for any ISO2709-based format AFAIK
00:59 hdl[…]/unimarcslim.html
01:02 gmcharlt hdl: (PDF)[…]s/UNIMARC_XML.pdf
01:02 hdl But Could be.
01:02 gmcharlt I see what you mean
01:03 gmcharlt *sigh*
01:03 gmcharlt how often are embedded fields used in practice in France?
01:04 hdl We, in Koha NEVER use them.
01:04 gmcharlt and outside of Koha?
01:04 hdl there is still a way to translate to non embedded field.
01:27 hdl gmcharlt: 3.21 AM here.
01:27 gmcharlt hdl: go to bed! :)
01:27 hdl I have something which works on the fly.
01:27 gmcharlt ah,
01:27 hdl But as a commandline script fails.
01:28 hdl In fact it does not update nor delete authorities.
01:28 hdl Can I send the patch to you ?
01:28 gmcharlt sure, I'll take a look
01:30 hdl doing this atm
01:34 hdl ent
01:34 hdl sent.
01:34 gmcharlt good night and thanks, hdl
01:44 hdl gmcharlt: something I realized was that we would need a script for all the previously edited authorities.
01:45 hdl Because they have not been logged as they were in koha2.2 (in files)
01:46 gmcharlt hdl: good point
01:46 hdl That script I can work later on.
01:46 hdl I think.
01:47 hdl unless you can cope with it.
01:47 hdl 'nite now.
01:47 gmcharlt g'night, hdl

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