IRC log for #koha, 2006-02-17

All times shown according to UTC.

Time S Nick Message
11:05 kados morning folks
11:15 _hdl_ morning kados ?
11:15 kados _hdl_: guess afternoon for you :-)
11:16 _hdl_ yep :)
11:16 _hdl_ I looked for a use of purchaseordernumber but couldnot get it.
11:16 _hdl_ Do you know hwat it is used for  ?
11:22 kados yes
11:22 kados a purchase order is used by a library to allocate funds for a specific purpose
11:22 kados before actually paying
11:22 kados it also acts as a reference point between a contract and an invoice
11:22 kados and a check
11:23 kados so someone wonders 'what is this check for made out to LibLime'
11:23 kados they can see the purchase order
11:23 kados then check the contract and invoices with the same purchase order
11:23 kados does that make sense?
11:44 _hdl_ Yes but in Koha, there is no use of this field.
11:45 _hdl_ As I read you, this field would act as an internal field for basket research.
11:45 _hdl_ Do I understand correctly ?
12:13 kados hmmm
12:13 kados I think it is mainly for budget-base acquisitions
12:13 kados s/base/based/
12:14 kados so when a library makes a purchase for books
12:14 kados they create a purchase order
12:14 kados with a purchase order number
12:14 kados which is used by the library to track the purcahse
12:15 kados if you write a mail to koha-devel I'm sure Stephen can answer your question better as he served as NPL's clerk for some time
12:35 _hdl_ shedges : I am wondering how and where purchaseordernumber is used for in Koha.
12:35 _hdl_ Seems never used in my code.
12:35 _hdl_ But what was it used for ?
12:35 _hdl_ Kados seemed to say you helped Katipo with Acquisitions ?
12:36 _hdl_ (see logs today)
12:36 shedges most libraries here have to have a "purchase order" before they can buy things.
12:36 shedges It's just a paper form, usually signed by the treasurer and the director
12:37 shedges it says that the purchase has been approved, that the library has enough money to pay for it, that the administration is aware of the order.
12:38 shedges Auditors look to be sure each invoice was pre-approved by a purchase order.
12:39 shedges So as orders arrive, they are usually matched with the corresponding purchase order.
12:39 _hdl_ But is invoice linked to a parcel you receive or to an order ?
12:39 _hdl_ (SHOULD be both. ;) in ideal world.)
12:40 _hdl_ I am asking it because you say purchase order is a way to approve invoice BEFORE paying it.
12:48 _hdl_ kados shedges any comment ?
12:50 shedges the invoice is linked to each shipment
12:51 shedges the purchase order is linked to each order
12:51 shedges one purchase order can have several invoices
12:51 shedges for example, I order 10 books, using one purchase order
12:51 shedges seven of the books are shipped (with an invoice) but three are backordered (out of stock)
12:52 shedges the three come later, on a separate invoice
12:52 shedges so I have one purchase order, one order, two shipments and two invoices
13:06 _hdl_ I understand.
13:08 _hdl_ But I am wondering if it is possible to manage all these things at the moment, and keep a history of parcels arrived.
13:09 _hdl_ Becaus if many parcels comes for one order, they will overwrite the with the latest input.
13:10 shedges ahh
13:10 _hdl_ aqorderdelivery was maybe desinged for this purpose.
13:10 _hdl_ But I wonder.
13:25 paul hello all
13:25 paul kados around ?
13:40 shedges hi paul
13:40 paul hello stephen.
13:40 shedges I think he is on a conference call right now
13:40 shedges (should be done soon)
13:40 paul ok, i'll leave in around 30mn
13:40 paul hoping i'll catch him
13:43 kados paul: you still here?
13:43 paul yes
13:43 paul I have some bad news, posted a few minuts ago on koha-devel...
13:43 kados opactest.liblime.com is running head and searches work!
13:44 paul it seems that DBI::mysql and utf8 are NOT friends at all.
13:44 kados hasn't come through yet
13:44 paul we have a major problem...
13:44 kados yikes!
13:44 paul copying some links :
13:44 kados do we need to move to postgres?
13:44 kados :-)
13:44 paul http://www.cpanforum.com/threads/654
13:44 paul http://lists.mysql.com/perl/3714
13:44 paul http://marc.theaimsgroup.com/?[…]1970179409036&w=2
13:44 paul http://lists.mysql.com/perl/3006?f=plain
13:45 paul in fact, they CAN be friend, but through 2 changes :
13:45 paul * bug DBD::mysql maintainer to have the fix included. Would require an official release, an upgrade of DBD::mysql, but it's a better solution in the long term.
13:45 paul * modify EVERY SQL query to Encode::decode_utf8() every value. A quite huge task !
13:45 kados hmmm
13:45 kados why do we need utf-8 in the mysql db?
13:45 paul (unless i'm missing something, & believe me, I really expect it !)
13:46 kados don't we just need it for the bibliographic data?
13:46 paul of course, if we want to be utf8 compliants
13:46 kados I don't think we need it in mysql
13:46 kados I think we can just encode as utf-8 before reaching the template
13:46 kados meaning all the browser sees is utf-8
13:47 kados ahh ...
13:47 paul that would be a poor solution I think. For example LDAP is utf-8 by default. Thus, for libraries using ldap & koha, the borrowers names have a strange look.
13:47 kados I see ... borrower names is a problem
13:47 kados forgot about those
13:47 kados hmmm
13:48 paul EMN has the problem & accept it, but only because I told "3.0 will be unicode" !
13:48 kados is there a fix for DBD::mysql?
13:48 paul (one of the 4 links)
13:48 kados (can we patch our local installations until it's released)
13:48 paul except it seems the patch is quite old, and was never released :-(
13:48 kados hmmm
13:49 kados I am happy to send an email to the maintainer
13:49 kados explaining our situation and asking him for advice
13:49 kados if he's not able/willing to fix it
13:49 kados it looks like we'll have to modify our SQL
13:49 kados if you describe to me how to do this I'll be happy to be the grunt man
13:50 kados so ... I'll await your email on koha-devel
13:50 kados you can rest easy tonight :-)
13:50 thd paul: have you looked at my SQL and quoting proposal?
13:50 paul it's quite simple : every time you fetch something, add a Encode::decode_utf8()
13:50 kados thd: good point this might be a good opportunity to apply that
13:50 paul thd : I saw it, but could not read it completly yet.
13:51 paul but still in "unread" status.
13:51 kados paul: unless you object, I think I'm going to have stephen add it to the coding guidelines
13:51 kados paul: thd's proposal that is
13:52 paul kados : if the maintainer can't include it, then we will have to change coding guidelines. Except that such a fix will enforce mySQL-only copatibility
13:52 thd paul: the identifier quoting part is the most important aspect for the short term
13:52 paul because other DBD::XXX don't suffer this problem it seem.
13:52 paul thd : I think i'll agree with your proposal.
13:52 kados paul: yep
13:52 paul thus the best solution from far would be to have the fix integrated !
13:52 kados paul: I'm actually seriously considering moving LibLime's future dbs to postgres
13:53 paul imho, that would :
13:53 kados paul: which is based on ingres, one of the best SQL dbs ever created
13:53 paul * delay koha 3.0
13:53 paul * be a problem for libraries using Koha 2.2 when they will migrate.
13:54 kados (I don't mean that mysql will be excluded, only that I want to create support for postgres)
13:54 thd paul: my proposal is only for placeholder code as SQL code is newly created or modified.
13:54 kados (and probably not for 3.0, I'm thinking 3.1 or 3.2)
13:55 kados (many large libraries would never go with a solution based on mysql ... with good reason i think)
13:55 thd paul: this would make it much easier to add flavour specific code later.
13:56 thd paul: the quoting aspect is to use My SQL specific back quotes for the default MySQL code.
13:58 thd paul: aside from identifier quoting needed already, marketing should be the significant motivator.
14:00 thd _hdl_: are you creating a parallel order package management system to that of normal acquisitions?
14:00 paul I hope not !!!
14:01 thd paul: what is he doing then?
14:01 paul I asked him to provide some tools to improve recievings, not rewritte it !
14:01 paul i'm afraid he is doing too much.
14:02 _hdl_ I am not rewriting it.
14:02 thd paul: every task tends to become too much after a closer look at the problem :)
14:02 paul hdl : ouf ;-)
14:03 paul I just wanted him to provide some facilities to see all recieves.
14:03 _hdl_ Just tryJust trying to well understand what has been done.
14:03 paul hdl : I know that we have a problem when a recieve is partial.
14:03 thd _hdl_: are you writing this 2.X also?
14:03 _hdl_ yes.
14:03 paul but there is no simple solution to this problem
14:03 _hdl_ First for 2.2.X
14:04 kados bbiab
14:04 paul we have to create a new table.
14:04 paul bye kados. i wont be here tomorrow.
14:04 paul but friday yes
14:04 thd _hdl_: I am happy about the first and would want to give you as much assistance as I can for that.
14:05 thd _hdl_: you do know that normal acquisitions is still currently broken.
14:07 thd _hdl_: chris ran out of time to fix it all for 2.X and made some uncommitted New Zealand specific shortcut fixes for a Katipo 2.X customer.
14:12 thd _hdl_: I am also interested in the issue of checking in multiple items for automatically adding them to the database with pregenerated barcodes rather than manually manipulating the item editor each time.
14:13 thd s/checking in/adding to the items for the biblio upon receipt/
14:13 thd s/upon receipt/upon unpacking/
14:14 _hdl_ Just typing in barcodes with a space.
14:14 thd _hdl_: I did not know.  Is that how it is meant to work?
14:15 thd good evening _hdl_ and paul.
14:15 paul I leave too.
14:15 thd _hdl_: will you be about tomorrow?
14:16 paul_away (I think he should)
14:16 _hdl_ yes
14:18 thd hdl_away: I will  try to sleep earlier today and rise earlier tomorrow or stay awake forever.
18:59 kados paul_away: you around?
19:47 russ jo you about6?
19:47 russ oops wrong channel
22:07 russ hi jo
22:07 russ yep
22:07 pez hello
22:07 pez : )
22:07 russ hi pez
22:18 audrey yes
22:24 kados _d1099 at C4/Biblio.pm line 158.
22:24 kados ZOOM error 224 "ES: immediate execution failed" (addinfo: "update_record failed") from diag-set 'Bib-1'
23:04 chris actually koha might handle 13 now, lemme check
23:05 chris yep we handle up to 14
23:15 kados er?
23:15 kados chris: is that a joke?
23:16 chris sorry was talking in hlt
23:16 chris and then changed channels without noticing :)
23:17 chris 14 in the isbn field
23:17 chris (13 char isbns are coming out soon)
23:17 chris was jsut checking how big 2.2 would handle
23:21 kados ahh
23:22 kados so the thing is ... I'm getting this error when I try to import records
23:22 kados this is line 158 of biblio.pm:
23:22 kados warn "zebra_create : $biblionumber =".$record->as_formatted;
23:22 chris didnt we fix that the other day?
23:22 kados no
23:23 chris im sure we were importing ok werent we?
23:23 kados well ... it died on me
23:23 kados and I assumed that was because it had finished
23:23 chris ahh i imported 3000 ok .. maybe i better check again
23:23 kados but I don't think that was why it died
23:24 kados I can make a 135M MARC file available to you if you want it
23:24 kados it's from a dynix system ... one of LibLime's new clients
23:24 chris ooh
23:25 chris yeah pop it up somewhere and drop me an email with a url
23:25 kados k
23:25 chris so when im home i can take a look
23:26 kados k
23:28 kados sent
23:28 chris cool
23:29 chris so whats the actual error its throwing?
23:30 chris i think its scrolled out of my scroll buffer
23:37 kados I think I found it
23:38 kados this client sent me two marc files
23:38 kados the first one had some errors in it
23:38 kados and that's the one that is thowing errors for us
23:38 kados I'll send you the link to the one that's been indexing for a bout a minute now
23:39 chris sweet
23:39 chris yeah we need nicer error handling for zebra ... that traps the error and reports abck nicely
23:40 kados yep
23:40 kados ok ... same link as before
23:40 chris cool
23:41 kados I wonder if there's an equiv to select count(*) in zebra
23:41 chris it might actually be MARC::File
23:41 chris that dies
23:41 chris when it tries to render the marc
23:41 chris either way, nicer error reports are needed
23:41 kados yep
23:41 chris hmm bound to be something like that
23:42 kados whatever it was, it required manual editing of the MARC
23:42 chris you can do do a search that will find everything, and then count the results
23:42 kados wonder what search that would be
23:42 chris  my $rs = $Zconn->search($q);
23:42 chris        my $numresults=$rs->size();
23:43 kados shoot
23:43 kados it just borked again
23:43 chris but yeah, i wonder what $q needs to be :)
23:43 chris ok, it should report and skip
23:43 chris i reckon
23:44 chris report a bad record, and then continue on
23:44 chris or say
23:44 chris bad record, skip and continue y/n
23:44 kados that makes sense
23:45 chris fatal errors should only be things like zebra dying
23:45 kados opac-detail is 500ing on me
23:45 kados is there stuff you haven't committed?
23:46 chris switch marc off
23:46 chris in system pref
23:46 chris and it should work
23:46 chris marc display isnt finished yet
23:47 chris tho why they opac tries to display marc im not sure :)
23:49 kados it's actually set to 'normal'
23:49 chris the marc on or off?
23:49 chris can you set it to normal?
23:49 chris i thought it was just on or off
23:50 kados log has:
23:50 kados [Wed Feb 15 20:08:07 2006] [error] [client 70.106.188.196] Undefined subroutine &C4::Biblio::MARCfind_MARC​bibid_from_oldbiblionumber called at /home/koha/testing/koha/op​ac/cgi-bin/opac-detail.pl line 82., referer: http://opactest.liblime.com/cg[…]ntains&value=test
23:50 chris ah yep, marc is still on
23:50 kados preferences say:
23:50 kados Define the default view of a biblio. Can be either normal, marc or isbd
23:50 chris naw not that one
23:50 chris that doesnt work :)
23:51 kados gotcha
23:51 kados hey, stuff's even displaying
23:51 chris :)
23:51 kados http://opactest.liblime.com/cg[…]etail.pl?bib=1570
23:51 chris sweet
23:51 kados nice job chris!
23:52 chris yeah it should jsut drop in for 2.2.5 too ... paul did most of it
23:52 chris havent got there yet
23:52 kados right
23:52 chris its fetching from koha tables at the mo
23:53 kados there's still data in the koha tables?
23:55 chris well there will need to be at least skeletons in items, with at least barcode and itemnumber
23:55 chris there is nothing in the marc_word table though
23:55 chris thats all in zebra
23:56 chris currently SearchMarc.pm is a drop in replacement
23:57 chris as Search.pm gets bigger more stuff dissapears from the db
23:57 chris make sense?
23:57 kados yea makes sense
23:57 chris so it will always be workingish .. just slowly working more and more from zebra
23:58 kados I'm not clear whether my contractual obligations are fulfilled yet :-)
23:58 kados well, after items is done
23:58 chris once i have a nice, i have a barcode
23:58 chris give me all the info i need
23:58 chris routine in Search.pm then 99% of the biblio info in koha will no longer be needed
23:59 kados gotcha
23:59 chris but if i take it out now, circ will break
23:59 kados yep
23:59 chris the good news is
00:00 chris the stuff on the first search results page is fetched from zebra
00:00 chris the detail.pl touchses koha tables
00:01 chris but i think ill leave that, get my search-test.pl going
00:01 chris and then get my get_record() going
00:01 kados sweet
00:01 chris and then opac-detail.pl should be easy
00:02 chris it seems to be clicking into placce
00:02 kados excellent ...
00:02 kados so I can hack on the import problem
00:03 chris cool
00:03 kados any ideas for how to make it 'skip' a record?
00:03 kados rather than just dying?
00:03 chris wrap the bit that is dying
00:03 chris in an eval { };
00:03 chris then do an
00:03 kados (it looks like it's dying with the ->as_formatted method
00:03 chris if ($@){
00:03 kados k
00:03 chris print out some error message
00:03 chris }
00:03 chris and it will keep on running
00:04 chris but at least grizzle about the error
00:04 kados k
00:04 chris it might just die on the next bit :)
00:05 chris i think its probably dying because $record doesnt exist
00:05 chris so maybe before the ->as_formatted
00:05 chris do an if ($record){
00:05 chris all the stuff it does with record
00:05 chris } else {
00:06 chris print "record wasnt set .. error woop woop etc";
00:06 chris }
00:06 kados actually ... I should probably write to a file
00:06 chris i think marc::record couldnt make a record, and then we try to use it, but its doesnt exist
00:06 kados cause the output's going by way to fast to catch that
00:07 chris yeah, or pipe to a file
00:07 chris when you run
00:08 kados piping's easier :-)
00:08 chris :)
00:08 kados ok ... running now
00:08 chris ideally
00:09 chris we should hand the error back
00:09 chris so that the calling program can deal with it
00:09 chris rather than printing it in biblio.pm
00:09 chris then bulkmarcimport.pl could print it out with the count, so you know what marc record is duff
00:10 chris and acquisitions could pass it to a template etc
00:10 kados yep
00:10 kados makes sense
00:11 chris ok, gonna go sort out some more bugs from my bugzilla list, catchya later
00:11 kados later
00:22 thd kados: are you still there?
00:28 kados thd: yep
00:29 thd kados: I committed a new version of SearchMarc.pm to HEAD.
00:30 thd kados: I have a separate rel_2_2 checkout and tried to commit a rel_2_2 version.
00:31 thd kados: I aborted that commit when the log was showing me that I was about to commit a modification of every file in the koha branch.
00:32 thd kados: the file dates of my rel_2_2 checkout seem right so what could be wrong.
00:32 thd ?
00:33 thd kados: I had tested the rel_2_2 cvs commit with my local rysync archive of the Koha CVS tree.
00:34 kados ?
00:34 kados what did the new version contain?
00:35 thd kados: Quick fix for functional bug in getMARCsubjects to avoid returning values that
00:35 thd vary between different uses of the same authorised subject heading causing
00:35 thd linked subject searches from the detail view to fail.  Other presentation fixes
00:35 thd within getMARCsubjects.
00:38 audrey kados: know much MARC tag designation?
00:39 audrey what is difference between 852 subfield h and 852 subfield k?
00:39 kados audrey: thd would be a better one to answer that
00:39 audrey ok. thd?
00:40 audrey do you know what the finer differences are?
00:40 audrey 852 is location according to http://www.loc.gov/marc/biblio[…]hold.html#mrcb852
00:40 kados chris: whenever you get back, it still died
00:40 kados chris: even with the eval
00:41 audrey thd: but what is difference between subfields h and k?
00:41 thd audrey: $k is a prefix like JUV for juvenile; $h is the base call number without the cutter number or possible location  suffix.
00:42 audrey and m is the location suffix or the cutter?
00:44 thd $i is the cutter; $m is the uncommon suffix.
00:44 audrey thanks! really helps my understanding.
00:45 audrey thd: what does a suffix look like?
00:45 audrey the $m one?
00:46 thd audrey: http://lwww.loc.gov/marc/ has much documentation but little explanation.
00:46 audrey i am seeing that
00:47 audrey do you know a better, more explanitory website?
00:49 thd audrey: 852
00:49 thd ##$aDLC$bc-G & M$hG3820 1687$i.H62$mVault
00:50 audrey so the specific room or branch described right in the call number?
00:50 audrey not used my many libraries, eh?
00:50 thd audrey: try a cataloguing textbook.  There is a good one by Chen.
00:50 audrey right, but I am not near any cataloguing testbooks right now.
00:51 thd audrey: $m is seldom ever used in my experience.
00:51 audrey ok, so not to worry too much about it.  cool.
00:51 audrey good.  now i know what it is, can explain it to others, and know not to stress about it.
00:52 audrey thanks:)
00:52 thd audrey: There is a good brief guild to MARC on the LC website but it does not have the detail for which you are looking.
00:52 kados thd: following up on our discussion about possible writing opportunities for you
00:52 kados thd: would you be interested in preparing 'fact sheets' for Koha?
00:53 thd audrey: Years ago I did find the notes for a library science class online but I have no idea now.
00:54 audrey thd: that's cool. may try to find other marc sites later.
00:54 thd kados: what would a 'fact sheet' as distinct from a 'lies sheet' contain?
00:54 kados thd: :-)
00:56 thd kados: I had thought about an approach for writing for the first issue of the LibLime newsletter.
00:57 kados thd: yea?
00:57 kados thd: I'm all ears
00:58 thd kados: I was late for my dentist appointment thinking about it
01:00 kados thd: :-)
01:00 kados thd: so ... what's your idea?
01:00 thd kados: tentative title, "Take back control of your library systems"
01:02 thd kados: The principles of free software freedom and there application to libraries with special consideration of the issue that there are very few free software projects for large complete systems as opposed to tools for building them.
01:02 thd s/there/their/
01:03 thd s/the first one
01:05 thd actually that would only match the first one :)
01:07 thd kados: what do you mean by Koha 'fact sheets'?
01:09 kados thd: well ... detailed descriptions of features maybe?
01:11 thd kados: Why did "cvs -z3 -d thd@cvs.savannah.nongnu.org:/sources/koha commit $Log -r rel_2_2 koha
01:11 thd " want to commit everything in my rel_2_2 checkout when I only modified one file?
01:11 kados what you wanted to do
01:11 kados was just do
01:11 kados cvs commit filename
01:11 kados you don't need to commit the whole tree
01:11 kados just the file you changed
01:12 thd kados: cvs commits only the modified files automatically.
01:14 thd kados: that code worked fine to my local copy of the Koha CVS tree downloaded with rsync.
01:15 kados rsync is your problem
01:15 kados I bet
01:16 thd kados: my CVS checkout used CVS directly from savannah I hope :)
01:17 kados thd: looks like your changes killed my opac search :-)
01:17 kados http://opactest.liblime.com/cg[…]ha/opac-search.pl
01:17 thd kados: Well, I will check that again.
01:17 kados wait
01:18 kados yea, something weird happening
01:18 thd kados: Head does not work on Koha 2.X.  You need my rel_2_2 file.
01:18 kados [Wed Feb 15 21:34:52 2006] [error] [client 70.106.188.196] ZOOM error 10012 "CQL transformation error" (addinfo: "Illegal or unsupported index (addinfo=\xb0\x9e\xdf\blastic)") from diag-set 'ZOOM', referer: http://opactest.liblime.com/cg[…]ha/opac-search.pl
01:18 kados ahh ... you changed 2.2
01:18 kados so this problem is unrelated then
01:19 kados i see
01:19 kados it's a CQL parsing error
01:19 kados can't handle multiple subject search terms yet
01:20 kados well any multiple terms it seems
01:20 thd kados: yes, I did not break anything that was not already broken.  I only fixed a few things that will give you a different appearance in every detail view.
01:21 kados thd: could you be more specific?
01:22 thd kados: I dealt with the problem you had with Sears in 650 $2.
01:23 kados I thought we already had a fix for that?
01:23 thd kados: Also, there was a line missing after the while loop that prevented the trailing " -- " from being removed from the last 6XX.
01:25 kados thd: but searches were working fine now
01:25 thd kados: And other improper extra " --  ".  I just realised that I may have not caught all possible cases, but certainly all that I have seen.
01:25 thd kados: that fix was never committed.
01:27 thd kados: the other issues were presentation problems that did not affect searching but looked bad on every detail view to have a trailing " -- " after the end of the last subject heading.
01:27 kados thd: I see
01:28 thd s/looked bad/had a poor appearance/
01:28 kados OK ... I'll try those out on a working system sometime soon
01:36 thd kados: that did not work, I killed my connection to be extra certain of not corrupting the whole Koha CVS tree.
01:37 thd kados: how do I commit a single file or small group of files?  More importantly, what is the problem rsync is not involved this time for certain.
01:38 thd s/problem/problem?/
01:41 thd chris: not too take you from your important work but why would CVS insist that I modified the whole rel_2_2 checkout when I only changed one file?
01:41 kados to commit a file just type:
01:41 kados cvs commit filename
01:41 kados if you want to commit a small group of files inside a directory:
01:41 kados cvs commit *
01:41 kados from within the directory
01:41 kados but don't do atomic commits
01:41 kados like cvs commit koha
01:41 kados :-)
01:42 thd kados: How will CVS know where to put the file?
01:42 thd kados: What are atomic commits?
01:42 kados cvs commit koha
01:42 kados is an atomic commit
01:42 kados thd: cvs just knows
01:43 kados thd: based on where it is in your repo
01:43 kados thd: all that info is stored in the CVS directory
01:43 kados thd: inside every repo directory
01:44 thd kados: am I supposed to remove the unmodified files.  CVS is just supposed to be able to just know the deference.
01:44 kados no don't need to remove anything
01:45 kados you can read some docs on how to use cvs on the savannah site
01:45 kados if you have usage questions
01:47 thd kados: I have read all the CVS docs and while I do not have extensive practise of usage I studied the docs from several CVS systems.  I am using just CVS and it is supposed to know what is modified and what is not.
01:49 thd kados: I had failed to understand a point about log messages a couple of months ago before my first commit.  chris put me straight about automatic logging.
01:52 thd kados: You are saying that I have to preserve the directory tree within which a single file or maybe even more than one file with the same name in different locations would be are you not?
01:54 thd kados: If that question bores you and you are still awake, who is meant to read the 'fact sheets'?
03:14 audrey Chris, hi
03:18 chris hi audrey
03:18 audrey have a moment
03:18 audrey ?
03:19 chris sure
04:52 |hdl| hi
04:53 thd hello |hdl|
04:53 hdl could you sleep a little ?
04:54 hdl (non slleping work is no good.
04:54 thd hdl: I hope to sleep but I have been testing a problem committing to rel_2_2
04:55 hdl Search.marc ?
04:55 thd yes
04:55 hdl (I saw you committed sthg.
04:55 hdl what kind of problem ?
04:57 thd I have the same code modifying rel_2_2 but when I go to commit CVS presumes that I have modified every file in rel_2_2
04:58 thd hdl: I have tested my CVS arguments with no problem using a local copy of the source tree.
04:58 hdl maybe you have a recursive commit on your directory ? ?
04:59 hdl This IS indeed a problem ?
04:59 chris you can just commit one file thd
04:59 chris cvs commit path/filename
04:59 thd hdl: Yes, that is the default but CVS is supposed to know the difference between what I have modified and what is the same.
04:59 chris eg cvs commit C4/SearchMarc.pm
05:00 chris you can go cvs diff filename to see what cvs thinks is different too
05:01 thd chris would that be actually cvs commit koha/C4/SearchMarc.pm
05:01 thd ?
05:01 chris well im normally in koha/
05:02 chris where koha is my cvs checkout
05:02 chris so cd koha
05:02 chris cvs commit C4/SearchMarc.pm
05:03 chris you could go cvs diff someotherfile .. to see why cvs thinks it is different
05:04 thd chris: I will try checking the diff return for all the files in rel_2_2 now :)
05:11 osmoze hello
05:13 hdl hi cris
05:13 hdl chris
05:13 hdl Hav you seen the logs ?
05:14 hdl Trying to dive into acquisitions, one more time :)
05:15 chris ahh, ill probably be doing some work on acquisitions for 3.0 in the next few weeks
05:15 chris on the full acquistions anyway
05:54 hdl chris : what are you planning to do ?
06:03 osmoze do you know if it's possible to transcript unimarc to usmarc ?
06:04 thd hdl: I assume chris intends to fix what is still broken.  I am curious about what aspects exactly chris knows to remain broken.
06:06 thd osmoze: I would be very pleased if you could find the correct files for doing that.
06:07 thd osmoze: http://www.bl.uk/services/bibl[…]ic/usemarcon.html
06:08 thd osmoze: what is your purpose for conversion exactly, particularly why do you want to convert into USMARC?
06:10 thd osmoze: paul wrote a rough conversion for MARC 21 to UNIMARC.
06:11 osmoze in fact, we have to send our "catalogue" to departemental librairy, but we are in unimarc and they are in USMARC. My question is that possible to send our "catalogue"
06:11 osmoze (excuse my poor english ^^)
06:11 thd osmoze: paul was not using USEMARCON.
06:13 thd osmoze: Do you really want to convert into UNIMARC from USMARC?  USMARC -> UNIMARC ?
06:15 osmoze no, it's the opposite
06:15 osmoze UNIMARC --> USMARC
06:15 thd osmoze: some libraries have the configuration files for USEMARCON.  If you find a library that has them, please let me know.
06:15 osmoze k
06:16 osmoze in fact it's to have an interoptability KOHA/ Multilys
06:17 thd osmoze: That is a big part of the holy grail.
06:17 osmoze lol :)
06:18 thd osmoze: contact some people at BNF or other large libraries about configuration files for USEMARCON to do that.
06:20 thd osmoze: Someone has the files somewhere already.  You could make your own but that is a nontrivial task with the special syntax for USEMARCON.
06:21 hdl osmoze : we could do that but on twodays work.
06:21 osmoze ok
06:22 osmoze hdl : c est a dire ?
06:22 thd osmoze: There is or was originally a proprietary version of USEMARCON that had configuration files not contained in the free version.  Find someone who is willing to share them or contact the original company about what leads they may have.
06:24 thd hdl: only two days?
06:24 osmoze thd>  french are good :p
06:25 hdl thd: I know His catalogue.
06:25 hdl :)
06:26 thd hdl: you mean two days for the values he has in his catalogue.
06:28 thd hdl: I want a tool to convert all BNF etc. records as completely as possible into MARC 21 and the other way from LC etc.
07:14 |hdl| thd: sorry : still working remotely on my computer.
07:14 |hdl| you said you wanted a tool to convert BNF biblio to LIC.
07:14 |hdl| s/LIC/LOC/
07:14 |hdl| This is not two days owk.
07:15 |hdl| s/owk/work/
07:15 |hdl| They are by far much more complicated biblio.
07:16 |hdl| And UNIMARC to MARC-21 is not a mathematical bijection.
07:17 |hdl| So There must be choicesto be done. Generally it is done on a target-display basis.
07:18 |hdl| We want this kind of information int this field.
07:18 |hdl| When you have a well defined kind of biblio, it is wuite easy.
07:18 |hdl| But with BNF it is truly a great deal.
07:18 |hdl| thd: ???

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