IRC log for #koha, 2006-03-09

All times shown according to UTC.

Time S Nick Message
11:00 paul => nothing found : ok, remove a subfield that maybe wrong => authorities found : ok, i'm almost sure and consider i'm sure
11:00 paul => still nothing found : ok, remove another one => ...
11:00 paul the problem being that searching on "doe" only will return MORE than 1 authority
11:00 paul and in this case, you have a problem (at least your bot has)
11:00 kados right
11:01 kados I already experienced this when running on existing bib data
11:01 paul that's an unsolvable problem.
11:01 kados right
11:01 paul it's a librarian one : "do you prefer : poor authorities or poor biblios ?"
11:01 kados right
11:02 kados finally I think I can explain the skipped question
11:02 paul you know another word than right ;-)
11:02 paul ?
11:02 kados heh
11:02 paul ok, you also know heh. something else ?
11:02 paul lol
11:02 kados in Koha there is only one authtag per framework, right?
11:03 paul you mean, in biblio ?
11:03 paul mmm... no, sorry
11:03 paul now I understand your question.
11:03 kados authtag is in
11:03 kados in MARC21 a single authority record can contain many tags
11:03 kados
11:03 paul in the authority, you have 1 (and only one, at least in UNIMARC) tag that is the "main entry"
11:04 kados yes ...
11:04 kados however
11:04 kados there are only 4 search points
11:04 paul thus, when you build an authority from a biblio, you have to know where (in the authority) you store the values extracted from the biblio
11:04 kados even though there are many more heading types specified
11:04 paul in my previous sample :
11:04 kados they fall into 4 groups
11:04 paul 700$31234$apoulain$bPaul$d1968-
11:05 paul in the biblio
11:05 paul will go in 200$apoulain$bpaul$d1968-
11:05 paul in the authority
11:05 kados so in that case, 200 is authtag
11:05 paul other tags are not handled by build_authorities, as they are NOT in the biblio
11:05 paul yep
11:05 paul (as the biblio contains only the acepted form)
11:06 kados I understand that ... I intend to modify build_authorities to handle external auth file
11:06 kados but the frameworks problem makes this tricky
11:06 kados because I'm not sure whether to have 4 frameworks or 8
11:06 paul but of course, the authority should contain many more things if it's a complete authority
11:06 paul explain a little bit more, I may understand at the end
11:07 kados ok ... so in MARC21 Authorities you have the following:
11:07 kados X00  Personal names
11:07 kados X10 Corporate names
11:07 kados X11 Meeting names
11:07 kados X30 Uniform titles
11:07 kados X48 Chronological terms
11:07 kados X50 Topical terms
11:07 kados X51 Geographic names
11:07 kados X55 Genre/form terms
11:07 kados that's 8 'types' of authorites
11:07 paul shedges just arrives in the middle of a marc21 lesson...
11:08 shedges X55 ???
11:08 kados in fact, in some auth records I suspect you will have values from several of these
11:08 shedges ahh   655, etc.
11:08 kados shedges: meaning 155, 255, 355, etc
11:09 kados so when you're importing an auth file and matching it to a bib file
11:09 kados you want the script to place values from all the relevant tags into the bib record
11:09 kados not just one tag
11:09 kados also ...
11:10 kados though there are 8 types listed above
11:10 kados they are grouped logically into 4 groups:
11:10 kados names, name/title combinations, uniform titles, or subjects.
11:10 kados so a patron doen't have to select 'personal names' or 'corporate names'
11:10 kados they just select 'names'
11:11 paul patron = you mean a borrower, not a librarian right ?
11:11 kados either one
11:11 paul you're speaking of search, not cataloguing ?
11:11 kados a librarian does not expect to do a authority search on 'personal name' either
11:11 kados unless I'm mistaken
11:11 kados correct, for searching
11:11 paul ok.
11:12 paul then, you're probably right. It can't be done unless i'm mistaken
11:12 kados of course, if I'm creating an original auth record
11:12 kados (like I did with the Twain and Lewis examples on the liblime demo)
11:12 kados it seems to work fine
11:12 kados but the question is, how do we import an external auth record
11:14 pierrick paul: why are moves to InnoDB and utf8 commented in updatedatabase ? (sorry to disturb)
11:15 paul (on phone)
11:29 kados hi kis, welcome to #koha
11:29 kls hi kados
11:34 pierrick (it was fast)
11:38 paul back
11:39 paul (not back)
11:41 kados heh
11:45 kados hdl: any success in troubleshooting the serials problem?
11:46 hdl I did.
11:46 hdl Committing
11:48 hdl you can update
11:49 hdl if (scalar(@itemstatusloop)){$data->{"​itemstatusloop"}=\@itemstatusloop;}
11:49 hdl else { $data->{"itemstatusloop"}=[];}
11:49 hdl if (scalar(@itemlocationloop)){$data->{"i​temlocationloop"}=\@itemlocationloop;}
11:49 hdl else {$data->{"itemlocationloop"}=[];}
11:49 hdl
11:49 hdl line 125
11:49 hdl (commited)
11:51 kados woot!
11:52 kados woohoo ... it works!
11:53 kados now I'll update NPL tempaltes and test it
11:53 kados hdl: thanks!
12:05 hdl when one has good information, one can make his way to the solution :D
12:05 hdl little chinese proverb :)))
12:05 pierrick general question : what is the procedure to follow when encountering "premature end of script" ? If I have it without post method form submission, Paul told me to simply execute the script. But with a post method form submitted, how can I hav more informations ?
12:08 hdl warn is your friend ;)
12:08 pierrick I'm testing warn
12:09 hdl you should get the form you send, see the arguments it sends, and then see in the script you are asking for what it becomes.
12:09 hdl But most of the time, /var/log/koha/your_koha_log is a good solution.
12:10 pierrick (nothing interesting in log file)
12:11 pierrick where is the warn message displayed ?
12:11 paul in the log file
12:12 paul I bet koha write in another file than the one you're looking for !
12:13 pierrick Koha write in /usr/local/koha/log/koha-error_log, but only "premature end of script"
12:14 pierrick do you think Koha writes another log file meanwhile ?
12:14 paul the only case I know where it do this is when your file is windows encoded.
12:14 paul and if perl runs fine, then it's not this.
12:19 pierrick OK, I've found the problem : intranet requires OPAC templates to display catalogue search results
12:19 paul ???
12:22 pierrick (I'm verifying before saying something stypid)
12:23 paul either i don't understand, either you're wrong.
12:24 pierrick pierrick@plegall:/usr/local/koha/intranet$ perl -I $KOHAHEAD ./cgi-bin/search.marc/
12:24 pierrick Salut Pierrick at ./cgi-bin/search.marc/ line 36.
12:25 pierrick HTML::Template->new() : Cannot open included file /usr/local/koha/opac/htdocs/opac-tmp​l/default/en/search.marc/search.tmpl
12:29 paul does Koha work on other pages ?
12:29 paul (bip pierrick)
12:30 pierrick of course it works, but I find that too many pages are down
12:30 pierrick (no sound on my computer)
12:30 paul you're with HEAD.
12:30 pierrick I am
12:30 paul so it maybe a code problem.
12:31 paul (if I don't mind, chris worked on this code recently, and it is probably very unstable)
12:31 pierrick I've modified form action in prog/en template
12:31 pierrick I've added ?type=intranet
12:31 pierrick but still have an error 500
12:34 pierrick My question was quite general in fact : is there a beautiful way to find where script ends ?
12:35 paul where script ends ?
12:35 paul a script usually ends at the last line ;-)
12:35 paul lol
12:35 pierrick except when it ends "prematurelly"
12:37 paul in apache log, the line where the error occurs is shown.
12:37 paul in your sample, it's line 36
12:37 paul (or there's something i'm missing in your question)
12:39 pierrick line 36, it my warn('Salut Pierrick');
12:39 paul ah, right.
12:39 paul this one is tricky.
12:40 paul it means that the get_user_and_template failed when opening the template.
12:40 paul there's something wrong in your template.
12:40 paul it's always hard to find what.
12:40 hdl kados : explain.abs is needed ?
12:40 paul in this case, it can't find /usr/... /search.tmpl
12:40 paul it's in opac-tmpl/default, so it can't find it !
12:41 hdl [warn] default.idx [No such file or directory]
12:41 hdl 16:41:53-07/03 zebraidx(31673) [warn] Couldn't open explain.abs [No such file or directory]
12:41 hdl
12:41 pierrick is it normal to use OPAC templates in intranet ?
12:41 paul in get_user_and_template, there's a parameter to say "search in opac template" or "in intranet"
12:41 paul NO, of course.
12:42 paul what is your get_template_and_user look like ?
12:42 pierrick this is why I've added ?type=intranet in the action of the form
12:42 paul do you have :
12:42 paul     type => "intranet",
12:42 pierrick My get_template_and_user is the current HEAD
12:42 paul mmm...
12:43 paul Cannot open included file
12:43 paul could it be an included file that is missing...
12:43 paul hdl, do you have an uptodate head ?
12:43 paul can you check & help pierrick ?
12:43 paul (I haven't, and prefer not to update it atm)
12:45 pierrick in search.marc/, I have "type => $type,", $type being taken from CGI param
12:45 hdl I am too experiencing problems with search
12:45 paul try to hardcode intranet, as there's no reason to use a variable.
12:46 paul (that's an oldies -but not goodies-)
12:47 hdl pierrick: I shall help you as soon as I have finished launching zebra update
12:47 hdl (paul : I am uptodate)
12:54 hdl kados : errors when updating a zebra base from your liblimedata
12:54 hdl I wanted to try if I could get stuf working with english data.
12:55 hdl (Congrats : you have frencf, spanish and chinese characters in your base)
12:55 hdl pierrick ?
12:56 kados hdl: I think that is my fault
12:56 kados hdl: (yes, all three :-)
12:56 pierrick I'm back
12:57 kados hdl: the problem has to do with bad config for isbn searching
12:57 kados hdl: I need to update that ... maybe later today
12:59 hdl So I often had this kind of problems when I mixed interfaces :opac and intranet (asking for opac on intranet port or the other way round) Did you verify that ?
13:00 kados hdl: are you speaking to me?
13:00 hdl sorry. Talking to pierrick
13:00 kados k
13:01 hdl kados: thx for your commit on that page.
13:01 hdl kados: Strange enough that bib1.att is not found
13:01 hdl kados: I had to make a copy of it.
13:02 kados strange
13:02 kados you mean in usmarc dir?
13:03 hdl yes.
13:05 kados I don't have a bib1.att
13:05 kados but I suspect it is in stock zebra
13:06 kados tab directory
13:06 kados unless paul has modified it
13:06 hdl it is.
13:07 hdl But a warning told me it was not found though you asked for it in zebra.cfg
13:07 pierrick As requested by Paul, I've hardcoded $type = 'template' in search.marc/, so I'm sure it doesn't try to load an OPAC template
13:07 pierrick but the problem remains
13:11 kados brb
13:13 pierrick hdl: I've just receive a CVS notification telling me you've modified search.marc/ and its associated template today, I'm checking what you modified and if it can be the origin of my problem
13:14 hdl pierrick: was for rel2_2
13:14 pierrick aren't we suppose to work on "prog" template for intranet on HAD ?
13:14 hdl ;)
13:14 pierrick HEAD even
13:14 hdl so not for the branch you are working on.
13:15 pierrick you're right, sorry to have imagine you could have make a mistake ;-)
13:15 paul hdl never made mistakes ...
13:16 pierrick by the way, how do you manage correction reports from rel-2_2 to HEAD ?
13:16 paul (in fact, he made just 1 in more than one year. but it was a nice one, that our customer really enjoyed...)
13:17 paul I regularly make a cvs update from branch to branch.
13:17 paul then solve all problems manually.
13:17 paul quite a long and boring stuff...
13:18 pierrick (code modification reporting is a lot simplified with Subversion)
13:18 pierrick (but that's another debate)
13:18 hdl (french assembly on DADVSI)
13:18 paul si tu es capable de nous aider à migrer sous subversion, n'hésite pas.
13:19 pierrick Savannah support Subversion ?
13:19 paul yes, I think.
13:19 pierrick I'm checking
13:21 paul (hdl : some real time news from parliament ?)
13:22 hdl Someone speaking about the way the government play.
13:25 hdl article 1 retired and some new amendments have been proposed in these few days.
13:27 pierrick Subversion is not available yet on Savannah
13:28 kados be back later
13:28 kados ciao all
13:29 hdl contrefaçon de la démocratie (Christian Paul)
13:29 paul ciao kados
13:30 paul communiqué de Que choisir aujourd'hui
14:11 paul pierrick : /me votes yes for a specific account on pwg to see your family.
14:32 paul[…]-060307a/#more-70
14:38 pierrick paul: let's do that tonight, I'm leaving office now :-)
14:39 hdl bayrou speaking
18:44 chris so
18:44 chris i have no idea how to delete a record from zebra
18:46 kados heh
18:47 kados first you need the id
18:47 kados then you pass zebra_extended_services the following:
18:48 kados oops
18:48 kados z3950_extended_services('update',set​_service_options('delete'),$record);
18:48 kados that should work
18:48 kados ahh ... we need a way to get the record id in set_service_options
18:49 kados so do something like:
18:49 kados z3950_extended_services('update',set_​service_options('delete',$recordId));
18:50 kados I'll make the necessary change to set_service_options
18:50 kados I'm wrong again I think
18:50 kados I confused service type with service option
18:51 kados turns out you probably need the 'update' service type
18:51 kados I need to check the docs
18:51 chris ahh right
18:51 chris well thats where im up to anyway
18:51 chris add and modify seem sweet
18:51 chris ive extended the search a bit more
18:51 chris so was gonna have a go at deleting
18:52 chris first the full record
18:52 chris then just an item from a record
18:52 kados right
18:52 kados for deleting an item
18:52 chris which i think we load the record, kill the item bit, and update
18:52 kados it is probably 'specialUpdate'
18:52 kados which looks like amodify
18:52 chris right
18:53 chris i think its all just and the zebra files to change from now on, shouldnt need to touch anything else
18:53 kados here we go
18:53 kados The update action. One of specialUpdate, recordInsert, recordReplace, recordDelete, elementUpdate.
18:54 chris yay recordDelete
18:54 chris cool
18:54 kados so I'll fix
18:54 chris excellent
18:54 chris 2 secs ill commit where im up to
18:54 chris and let you have at it
18:54 chris done
18:54 kados k
19:00 kados chris: ok ... try this now:
19:01 kados z3950_extended_services('update',set_service_​options('update','recordDelete',$recordId));
19:02 kados adam didn't give me a straight answer on the difference between recordIdNumber and recordIdOpaque
19:02 kados but I think we want recordIdNumber
19:02 chris so i just have to figure how to get that eh
19:04 kados yea
19:04 chris right
19:04 kados there's a routine in
19:04 kados in fact, deleting an item is a bit more tricky
19:05 chris a routine in to get a recordid?
19:05 kados is'nt recordid just the biblionumber?
19:05 chris dont think so
19:05 kados ahh
19:06 kados I'm pretty sure it is ...
19:06 kados could be wrong though
19:06 chris im pretty sure its not :)
19:06 kados heh
19:06 chris melm 090$c              identifier-standard,identifier-standard:p
19:06 chris its identifier standard
19:06 kados ahh
19:07 chris local-number is att 12
19:08 chris i noticed this when just dumping the xml that zebra hands back
19:08 kados we should have a line like this:
19:08 kados gils.recordId: $type (bib1,Identifier-standard)
19:08 kados somewhere
19:09 kados
19:09 chris ahh
19:09 chris i wonder what file that goes in?
19:09 chris collection.abs?
19:10 kados I bet it's zebra.cfg
19:10 chris yeah sounds right
19:10 chris ill try that out
19:10 kados quoting:
19:10 kados "the sample GILS records that come with the Zebra distribution contain a unique ID in the data tagged Control-Identifier. The data is mapped to the Bib-1 use attribute Identifier-standard (code 1007). To use this field as a record id, specify (bib1,Identifier-standard) as the value of the recordId in the configuration file."
19:11 chris hmm
19:11 chris we arent use GILS records eh
19:11 kados naw
19:11 kados but it should work for ours too
19:11 chris i wonder if just
19:12 kados that's just an example
19:12 chris recordId: $type (bib1,Identifier-standard)
19:12 chris is what we want
19:12 chris without the gils. bit
19:12 kados hmmm
19:12 kados it might be:
19:12 kados xml.recordId: ...
19:12 chris ill try and see what happens
19:13 kados it could even be:
19:14 kados collection.recordId :-)
19:14 chris yeah
19:14 kados or stripping off the gils. bit ...
19:15 kados owen: what's up?
19:15 chris ill just try them until it works :)
19:15 kados hehe
19:15 owen kados: I've been working on the opac dictionary search
19:15 kados sweet
19:16 owen Seems to be working now, but one oddity
19:16 owen[…]ha/
19:16 owen Try one of the ... links to pop up the search window
19:17 owen The opaclayoutstylesheet variable doesn't seem to be coming through
19:17 kados huh
19:19 chris 11:19:16-08/03 zebrasrv(5) [request] EsRequest  ERROR 224 only XML update supported
19:19 chris reckon i need to upgrade zoom, or zebra or something?
19:20 kados hmmm
19:20 kados dunno
19:20 kados when did you get that error?
19:21 chris hmm it may have worked
19:21 kados owen: i think the pop-up must be handled in a separate template->param
19:21 chris 11:19:16-08/03 zebrasrv(5) [log] recordDelete
19:21 kados woot!
19:21 kados chris: what was the right syntax?
19:21 chris oh no its still there
19:21 kados shoot
19:22 kados did you send a commit afterwards?
19:22 chris hmm nope
19:22 kados gotta do that I think
19:22 chris ohh i dont think i have shadow set up
19:22 chris i wonder if thats breaking it
19:22 chris lemme try that
19:22 kados not in zebra.cfg?
19:22 chris shadow shadow:4g
19:23 kados yep ... you need a commit then
19:23 chris do i need to make a shadow dir?
19:23 kados yea
19:23 kados well I may have committed one actually
19:23 kados check your zebra dir
19:23 chris ok ill add a commit and lets try again
19:24 chris z3950_extended_services('commit');  is taht all i need?
19:25 kados yup
19:26 kados I've not tested it without sending two params
19:26 kados but I recon it'll work
19:26 chris the commit worked .. but the record is still there ... its not liking the delete i dont think
19:28 chris hmm
19:28 kados bummer
19:28 kados owen: cvs updated ... try now
19:28 kados I wonder if it's recordId
19:28 kados try this
19:29 kados in
19:29 kados line 240
19:30 kados change $serviceOptions->{'recordIdNumber'} = $recordId; to $serviceOptions->{'recordIdOpaque'} = $recordId;
19:30 chris it could be the wrong line in zebra.cfg
19:30 kados yep, that too
19:30 chris or it could be that i need to add a new one
19:30 chris before i delete it
19:30 kados right ... you probably need to reindex
19:30 kados now that you changed the config
19:30 chris yeah ill try a reindex
19:31 kados owen: don't know if you read the log this morning
19:31 owen No
19:31 kados owen: hdl fixed a bug in serials
19:32 kados owen: now every serial can have an associated item
19:32 kados owen: there are some template modifs to enable that in NPL
19:32 owen in rel_2_2 ?
19:33 hdl yes owen
19:34 hdl hello and good night folks. I see zebra is becoming more and more "tamed".
19:34 kados hdl: night
19:36 owen kados: did you commit changes to opac suggestions to allow anonymous use?
19:37 kados yep
19:38 kados you need to specify an anonymous borrowernumber
19:38 kados in sysprefs
19:39 kados you'll need to updatedatabase to get that modif
19:40 owen I see. No changes to the template?
19:43 kados nope :)
19:44 owen I tried some changes to the template today:[…]
19:44 owen The layout is less ugly but still a little awkward
19:45 kados very big improvement!
19:45 owen does it always show the 'suggested by' name?
19:45 kados good question
19:45 kados I see it and I'm not logged in
19:45 kados so I assume yes
19:46 kados do you have an anonymous patron set up?
19:46 owen No
19:48 kados looks like it always shows the anme
19:48 kados name even
19:48 kados that should be optional as well I'm guessing ... at least on the opac
19:48 kados for security reasons
19:48 kados right?
19:49 kados can't we just omit that column from the template?
19:49 kados seems like an easier solution :-)
19:49 owen Of course
19:49 owen Why are we accepting anonymous suggestions?
19:50 kados stephen wanted too ... at least that was what I got out of the email he sent
19:50 owen There's no way for us to contact the user if their item is purchased
19:50 kados he wanted to create a completely new suggestions form because the current one didn't support anonymous suggestions
19:50 kados you'd have to take it up with him
19:51 kados of course, you dont' _need_ to accept anonymous suggestions
19:51 kados if you don't enable it in the syspref it won't work of course
19:51 owen Come to think of it, why do we have this list of pending suggestions open to the public at all?
19:51 kados dunno
19:51 owen Why not just put up the form and leave it at that?
19:52 kados it would be nice if they could add their name to the list
19:52 kados but Koha doesn't do that yet
19:52 owen Yeah.
19:52 owen I'll have to bring it up with Stephen
19:52 kados of course, in Koha you can manage your own suggestions
19:52 kados ie, delete one if you get the book from somewhere else
19:53 kados so that's one reason to have the list
19:55 owen By the way, I don't see any updates to serials in rel_2_2
19:56 owen least not to the default templates
19:56 owen I see that was updated
19:59 kados hmmm ...
19:59 kados there are definitely some new input boxes in the default
20:00 kados check the 'recieve' section
20:00 kados russ: you around?
20:06 kados chris: here's the note from NBBC:
20:06 kados Look at barcode 254707, it won’t save all three lines from the 942, it drops
20:06 kados off the library code, the dewey or the item type each time we try to correct  
20:06 kados it.  In will only keep two of them, but it atleast cycles through the three.
20:06 kados Each time we look at it again, it is a different one missing – we had a good  
20:06 kados laugh about it all.
20:07 kados also:
20:07 kados We have the book “The Superfluous Manâ€Â? in the reservoir.  When we try to add
20:07 kados things to the bibliographic are of Koha, it erases the MARC record.  We put
20:07 kados things in the first 942, click on “Save Bibliographic â€Â? and it lists the item
20:07 kados number, but not the title.  We go back and look at the record and there is no
20:07 kados title or anything.
20:07 kados so it looks like two issues to me
20:07 kados 1: items aren't saving correctly
20:07 chris hmmm
20:07 kados 2: records retrieved from the reservoir aren't getting in properly
20:07 chris yeah
20:08 chris by items you mean records right?
20:08 kados yea
20:08 kados no ... items I think
20:08 kados the 942 is items right?
20:08 chris dewey and itemtype
20:08 chris nope
20:08 kados ahh ... right
20:08 chris items is just branch, barcode etc
20:08 kados strange as I thought we tested that so many times
20:08 chris yep
20:08 kados I'm gonna try it on their system
20:08 chris cool
20:10 kados weird ... it's working for me :/
20:10 chris im not suprised about the reserviour
20:11 kados yea, I actually forgot completely about it :-)
20:11 chris me too
20:11 chris but im suprised about the 942 thing
20:11 chris hmm maybe its just editing that doesnt work? adding seemed to work fine
20:12 kados heh ...
20:12 kados yea, I just tried it myself
20:12 kados it's definitely happening
20:12 chris when editing one
20:12 kados I'll see if I can reproduce it on liblime's system
20:12 kados yea, editing
20:13 chris right
20:13 kados though I don't know why that would matter
20:13 kados both of them call the exact same routine
20:13 kados (I wonder if the 'last' item is getting dropped)
20:13 kados I bet that's it in fact
20:13 chris i bet it is actually
20:13 chris good thinking
20:15 kados chris: I might need you to do this one :-)
20:15 kados I'm getting googly eyed just looking at it :-)
20:17 chris will do, its our html2xml bit ye?
20:23 kados yea
20:24 chris hmmm
20:24 chris <datafield tag="942" ind1=" " ind2=" ">
20:24 chris <subfield code="c">BOOK</subfield>
20:24 chris <subfield code="k">12345</subfield>
20:24 chris </datafield>
20:24 chris all looks fine in the xml
20:25 chris seemed to stick too
20:26 kados weird
20:26 chris oh that was an add, lets see what happens on edit
20:26 kados you had values in for all three 942 subfields?
20:27 chris yep
20:28 chris <datafield tag="942" ind1=" " ind2=" ">
20:28 chris <subfield code="c">BOOK</subfield>
20:28 chris <subfield code="a">12345</subfield>
20:28 chris <subfield code="k">2</subfield>
20:28 chris </datafield>
20:28 chris can you try an edit, and check the error log
20:28 kados yep
20:28 chris (if you have access
20:28 kados well ... it's on my machine :-)
20:28 chris ahh well easy then :)
20:28 chris see what is coming out in the xml there
20:29 kados aahha
20:29 kados remember this one:
20:29 kados [Tue Mar 07 16:12:50 2006] [error] [client] DBD::mysql::db do failed: You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near '' at line 1 at /home/nbbc/koha/intranet/modules/C4/ line 1445., referer:[…]iblionumber=41044
20:29 kados [Tue Mar 07 16:12:50 2006] [error] [client] DBD::mysql::st execute failed: Column 'biblionumber' cannot be null at /home/nbbc/koha/intranet/modules/C4/ line 1880., referer:[…]iblionumber=41044
20:29 chris ah ha
20:29 chris heres what i did
20:29 chris if you look at in head
20:30 kados looking
20:30 chris line 325
20:30 chris if (!$biblionumber){
20:30 chris    $biblionumber=$input->param('oldbiblionumber');
20:30 chris    }
20:30 kados after my $previous_tag = ''; ?
20:31 chris sorry 352
20:31 kados gotcha
20:31 kados I'll try that
20:31 chris cool
20:34 kados huh ... that script is quite different in rel_2_2
20:34 kados did you commit that to HEAD?
20:34 chris yes
20:34 chris the head one does the zebra stuff
20:35 chris the rel_2_2 one doesnt
20:35 chris im not sure what else the head one does differently
20:36 kados I don't see that change in head's
20:36 chris oh sorry maybe i didnt
20:37 chris committed now
20:38 kados thx
20:38 chris you are adding it to the rel_2_2 one eh?
20:38 chris i assume thats what NBBC is running
20:39 kados yea
20:39 kados check out the rel_2_2 version
20:39 kados it has:
20:39 kados my $oldbiblionumber=$input->param('oldbiblionumber'); # if bib exists, it's a modif, not a new biblio.
20:39 chris right
20:39 kados line 353
20:40 chris but then it uses biblionumber later on instead of $oldbiblionumber eh?
20:40 kados I can't find a $biblionumber
20:40 chris ok, well that bit wont be needed then i guess
20:41 kados seems to just use oldbiblionumber strangely
20:41 chris right so i suspect
20:41 chris its line 367
20:41 kados here's the line 1880
20:42 kados $dbh->prepare(
20:42 kados        "insert into bibliosubtitle set biblionumber = ?, subtitle = ?");
20:42 kados    $sth->execute( $bibnum, $subtitle ) if $subtitle;
20:43 chris hmmm
20:43 kados wonder if it only happens when there's a subtitle
20:43 chris thats stopping it get as far as doing the html2xml bit eh?
20:45 kados could be
20:45 chris ie, your not seeing the xml in the error log?
20:45 kados nope, it's not that
20:46 kados let me check the xml
20:47 kados xml isn't showing up for me
20:47 kados for all three
20:47 kados just two out of the three
20:47 chris how odd
20:47 kados wonder what's different about this install
20:47 chris can u paste it in?
20:47 kados paste in what? the xml?
20:47 chris yeah just then 942 bit
20:47 kados <datafield tag="942" ind1=" " ind2=" ">
20:47 kados <subfield code="a">WN3</subfield>
20:47 kados <subfield code="c">BK</subfield>
20:48 kados </datafield>
20:48 kados after that I get:
20:48 kados <datafield tag="090" ind1="" ind2="">
20:48 kados <subfield code="d">41044</subfield>
20:48 chris right
20:48 kados </datafield>
20:48 chris so next thing to test
20:48 kados </collection>
20:48 chris is it coming across from the form ok
20:48 kados I"ll check
20:50 chris interestingly
20:52 chris i wonder why 090 is after 942
20:52 chris oh no thats right
20:53 chris thats just very odd, i wonder why its only getting 2 of the 3
20:53 kados coming through on the form too
20:53 kados [Tue Mar 07 16:53:54 2006] [error] [client] TAGS: 942 c BK  at /home/nbbc/koha/intranet/modules/C4/ line 1048., referer:[…]iblionumber=41044[Tue Mar 07 16:53:54 2006] [error] [client] TAGS: 942 k 314  at /home/nbbc/koha/intranet/modules/C4/ line 1048., referer:[…]iblionumber=41044[Tue Mar 07 16:5
20:54 kados backwards though oddly enough
20:55 chris right
20:55 kados it gets a different 2 out of three every time too
20:55 chris should thre be 3 tags?
20:55 kados yea
20:55 kados remember the tags come through with the subfields
20:55 chris yeah
20:56 chris but i only see 2 there
20:56 kados that's a warn right after the for that loops through everything
20:56 kados 942 c BK
20:56 kados 942 k 314
20:56 kados 942 a WN3
20:56 kados all three are there
20:56 chris ohh you didnt past the a one?
20:56 kados I see it in my scrollback
20:57 kados maybe it didn't paste properly
20:57 chris weird i dont
20:57 chris hehe never mind
20:57 chris so for some reason its coming through but its not getting into the xml
20:57 chris can you try somethign for me please
20:57 kados so it's definitely in the MARChtml2xml routine
20:57 kados sure
20:57 chris fill in all the subfields of 245
20:57 chris and see if they all end up in the xml
20:58 kados k
20:58 chris wanna know if its perculair to 942 .. or a multiple subfields thing
20:58 kados it works in 245
20:58 chris now thats just mad
20:59 kados yea
20:59 chris ok, for the 942
20:59 chris is it the last one missing?
21:01 kados it seems to change :-)
21:01 kados every time i edit it :-)
21:02 kados can you reproduce it on your box?
21:02 chris nope but i dont have a 2.2 straight install
21:02 chris in the listing of the tags in the error log
21:02 chris does the order there change?
21:03 chris or is it the same order everytime?
21:03 kados the order changes there
21:03 kados huh ... on the liblime server it seems to work
21:03 chris right .. and is it the last one of those
21:03 chris that is left out?
21:03 kados let me check that
21:04 chris weird it works on liblime .. mental
21:04 kados this last time it was the first one that got left off!
21:06 chris always a good plan
21:06 kados aaarg!
21:06 kados same prob
21:07 chris weird
21:07 kados hang on, got a support issue to deal with
21:07 chris and yet it seems to work ok on liblime
21:07 chris ok
21:13 kados can you do me a huge favor
21:13 kados do you have a running mailman you can check something for me?
21:14 kados chris: what's the permissions on /var/spool/mailman/in/
21:14 kados chris: and who owns
21:14 chris ill look
21:14 kados thx
21:15 chris hmmm
21:15 chris what distro
21:16 chris cos on debian
21:16 chris there is no /var/spool/mailman
21:16 kados shoot
21:16 kados well do you have a mailmain/in dir somewhere?
21:17 chris hmm
21:19 chris not that i can see
21:19 chris there is qfiles/in
21:20 chris drwxrws---  2 list list  4096 2006-03-08 09:23 in
21:20 kados ahh
21:20 kados yea, deb must be different
21:20 kados thanks for looking
21:24 chris np
22:29 kados phew!
22:29 chris was it permissions?
22:29 kados that was a killer :-)
22:29 kados yea
22:29 kados those mailman guys have their shit together
22:29 chris yep
22:29 kados I spent all that time checking perms
22:30 kados and here they have a perms_check script!
22:30 kados and it even fixes wrong permissions
22:30 kados rock
22:30 kados so what the heck is going on with this sub on NBBC's site?
22:31 kados I'm going to try adding the exact record they're getting the issue with to liblime's site
22:31 chris plan
22:39 kados bugger
22:39 kados internal server error
22:40 kados [Tue Mar 07 17:37:45 2006] [error] [client] Tag "" is not a valid tag. at /var/www/​tranet/modules/C4/ line 1373, referer:[…]mple/
22:40 kados 942 came through in the xml though
22:40 chris blank tag eh?
22:41 kados I guess ... weird though
22:41 kados I had a repeatable 650
22:42 kados I don't see the blank tag in the xml
22:42 chris its not that part thats grizzling
22:42 chris line 1373 is not in the html2xml bit
22:42 chris its down in NEWnewbiblio
22:43 kados hmm
22:44 chris what is calling that subroutine
22:45 chris its doing a bunch of marc stuff
22:45 kados in fact calls it
22:45 kados ($bibid,$oldbibnum,$oldbibitemnum) = NEWnewbiblio($dbh,$record,$frameworkcode);
22:46 chris right
22:46 chris have a look at it
22:47 chris i wonder what its doing
22:48 kados looks like it calls it if the record is a new item
22:48 kados new biblio I mean
22:48 chris yeah but i wonder whats it doing with the MARC::Field stuff
22:48 chris thats the bit that is failing too
22:49 chris it looks more like a modify to me
22:50 chris my $old_field = $record->field($tagfield1);
22:50 chris                $record->delete_field($old_field);
22:50 chris                $record->add_fields($newfield);
22:50 kados I wonder if maybe $newfield = MARC::Field->new( isn't the right way to add fixed fields
22:51 chris naw i just think that $tagfield1 is < 10
22:51 chris cos $tagfield1 is empty
22:51 kados but why would it ever be empyt?
22:51 kados it's that what we fixed? :-)
22:52 chris we didnt touch this bit
22:52 chris its fetching stuff from a db
22:52 kados yea, but it's getting the $record object from MARC::File::XML isn't it?
22:52 chris yeah
22:52 chris but thats not where tagfield1 comes from
22:52 kados ahh
22:53 chris   $dbh->prepare(
22:53 chris "select tagfield,tagsubfield from marc_subfield_structure where kohafield=?"
22:53 chris    );
22:53 chris    $sth->execute("biblio.biblionumber");
22:53 chris    ( my $tagfield1, my $tagsubfield1 ) = $sth->fetchrow;
22:53 kados right
22:54 kados so maybe first it creates it
22:54 kados MARCmarc2koha
22:54 kados my $olddata = MARCmarc2koha( $dbh, $record,$frameworkcode );
22:54 kados    $oldbibnum = OLDnewbiblio( $dbh, $olddata );
22:54 kados actually
22:54 kados then adds stuff to it once it's created?
22:54 chris if ($tagfield1 && $tagfield1<10)
22:55 chris lets try making line 1373 into that
22:55 kados maybe I don't have one of the koha fields mapped
22:55 kados someone may have screwed up the mapping
22:55 chris could be
22:55 chris thats a good thing to check
22:56 kados heh
22:56 kados biblionumber's not mapped
22:56 chris ll
22:56 chris +o
22:56 chris no wonder
22:56 kados prolly sirsi
22:56 kados :-)
22:57 chris :)
22:58 kados yep, that was it
22:58 kados I've got to find a way to lock that down :-)
22:58 chris :)
22:59 kados well ... it worked
22:59 chris so it works on liblime
22:59 kados so that's not much helright
22:59 chris yeah thats weird
22:59 chris i wonder whats different
23:00 kados the only thing I can think of
23:00 kados is maybe they imported it originally from the reservoir
23:00 kados but I don't think they did
23:01 kados based on the email I got
23:01 chris yeah
23:01 chris too weird
23:01 chris i thought it might be templates, but it did seem to be coming thru from the form ok
23:01 kados We have the book “The Superfluous Manâ€Â? in the reservoir.  When we try to add
23:01 kados things to the bibliographic are of Koha, it erases the MARC record.  We put
23:01 kados things in the first 942, click on “Save Bibliographic â€Â? and it lists the item
23:01 kados number, but not the title.  We go back and look at the record and there is no
23:01 kados title or anything.
23:02 kados so it sounds like it's happening with more than just 942
23:02 chris yeah that sounds quite seperate
23:05 kados [Tue Mar 07 17:28:37 2006] [error] [client] DBD::mysql::db do failed: You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near '' at line 22 at /home/nbbc/koha/intranet/modules/C4/ line 1802., referer:[…]rameworkcode=EDIT
23:05 chris i dont think that has anything to do with it
23:05 kados but that can't be good
23:05 chris no
23:05 kados a query failing
23:06 chris i have no idea why its not making it into the xml thats just really od
23:06 chris d
23:09 kados I wonder
23:09 kados paul's old routines
23:10 kados were creating blank values in subfields
23:10 kados anything that was linked to a koha table
23:10 kados would probably register as a blank value rather than now as a NULL
23:10 chris right
23:11 kados which could be causing that query above to fail
23:11 chris could be
23:11 kados [Tue Mar 07 19:09:40 2006] [error] [client] DBD::mysql::db do failed: You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near '' at line 1 at /home/nbbc/koha/intranet/modules/C4/ line 1445., referer:[…]iblionumber=41044
23:11 kados [Tue Mar 07 19:09:40 2006] [error] [client] DBD::mysql::st execute failed: Column 'biblionumber' cannot be null at /home/nbbc/koha/intranet/modules/C4/ line 1882., referer:[…]iblionumber=41044
23:11 kados is the full error
23:12 kados it's two errors really
23:12 kados shoot, this is nuts
23:12 chris i wouldnt worry about this one for now
23:12 chris i think its a red herring
23:13 kados ok
23:13 chris at least for the 942 problem
23:13 kados yea, that makes sense I think
23:13 kados unless MARChtml2xml is being called twice
23:14 kados but from the log I don't think it is
23:14 kados cause then I'd see the transaction twice and I don't
23:14 chris yeah
23:14 kados interesting
23:14 kados this last time I saved that record
23:14 kados I didn't fill out all three values
23:15 kados and now it only has 1 :-)
23:15 chris hmm so its always missing one?
23:15 kados I wonder if the 090 is overwriting it
23:15 kados since the 090 seems to be added last for some reason
23:16 chris so what is the difference between liblime and nbbc
23:16 chris thats gotta be the key
23:16 kados I can diff and
23:17 chris templates would be the only other thing i can think of
23:18 kados no dif with the module
23:18 kados "no diff on the script
23:19 kados checking the template
23:19 chris can we switch them both to default templates?
23:19 chris actually scrap that idea
23:23 kados swapped out the template
23:23 kados with the latest cvs
23:23 kados same prob
23:24 kados wait
23:24 kados I lie
23:24 kados sec
23:28 kados weird
23:29 kados so it loses whatever one was listed first on the form
23:30 kados in the 942
23:30 kados addbiblio.tmpl,, and are all stock CVS same as liblime
23:31 chris thats just mental
23:31 kados I just noticed something
23:31 kados 942 has an authorized value
23:31 kados I'm going to get rid of that in the framework
23:31 kados then see if tha fixes it
23:32 chris ahhh that could well be it
23:32 chris least it gives us a hint what to look at fixing
23:34 kados nope it's not :(
23:34 kados sigh
23:34 kados I think I tried creating a blank record
23:34 kados and I didn't have the problem
23:35 chris so if you create one, then edit it, its fine
23:35 chris but if you edit this one, it doesnt work?
23:35 kados let me double check that
23:35 chris ta
23:40 kados huh
23:41 kados so I created the exact same record
23:41 kados confirmed it wasn't a duplicate
23:41 chris right
23:41 kados and it took out _two_ of the 942s :-)
23:41 chris what the hell
23:41 kados I think it's laughing at us :-)
23:41 chris theres nothing after 942 eh?
23:42 chris can we add another tag on tab 5 in the framework
23:43 kados yea
23:43 chris my latest theory is its broken for the last tag
23:43 chris it mine be as wrong as all the other theories ive had about this problem though :)
23:45 kados so I created a 970 tag
23:45 kados with a subfield $a
23:45 kados and edited the same biblio
23:45 kados not only did it delete one of the 942s
23:45 kados it dropped the 970$a too
23:46 kados ok I'm gonna wipe out the logs
23:46 kados restart apache
23:46 kados and put some fresh warns in there
23:46 chris good idea
23:46 kados double check everything again
23:46 kados chris: is there a way to flush the logs in apache?
23:46 kados chris: without restarting?
23:47 chris dont think so
23:48 kados ok ... here we go
23:54 kados
23:54 kados so when it arrives at MARChtml2xml
23:54 kados everything's there
23:55 kados though for some reason, 090 is last
23:55 chris yep i think its the last hidden input on the form?
23:55 kados might be
23:56 kados the value has to be generated by koha
23:56 kados not sure if that's done before or after the form is loaded
23:56 kados probably before
23:56 kados anyway
23:56 chris so a,k,c was the order
23:56 kados at some point the first one in 942 disappears
23:56 chris and it only got k and c
23:56 chris yeah
23:57 kados spooky
23:57 kados also, 970 went away
23:57 kados that's got to be significant
23:57 kados it's like everything after a certain point gets dumped
23:58 chris yeah, what happens if you put in a 970b
00:00 kados interesting
00:00 kados the value for the $b was retained
00:00 kados but it wasn't linked to a subfield
00:00 chris hmmm
00:01 kados[…]
00:01 kados well ... I guess it is on the OPAC view
00:01 kados anyway, $b shows up
00:02 kados so in the 9XXs Koha's dropping the first subfield
00:02 kados but just on NBBC's server :-)
00:02 chris yeah
00:02 chris what version of perl?
00:02 chris are they both the same?
00:03 kados his is perl, v5.8.4 built for i386-linux-thread-multi
00:03 kados This is perl, v5.8.5 built for i386-linux-thread-multi
00:03 kados no :-)
00:03 kados the NBBC box is debian
00:03 kados is fedora
00:03 chris i wouldnt expect that to be a problem
00:03 chris since i havent had the 942 dissapearing on my box
00:04 kados so NBBC has 5.8.4
00:04 chris which is 5.8.4
00:04 kados right
00:04 chris problems like this are so frustrating
00:04 kados the thing is
00:04 kados it _has_ to be happening _inside_ of MARChtml2xml
00:04 kados so what's going on in there
00:04 chris yes thats the only thing that creates the xml file
00:05 kados I suppose I should check versions of MARC::File::XML
00:05 chris naw
00:05 chris it doesnt even use that
00:05 kados right
00:05 kados what else could it possibly be?
00:05 chris around line 1079
00:06 chris we have
00:06 chris if (@$values[$i] eq "") {
00:06 chris                }
00:06 chris                else {
00:06 kados yea
00:07 chris lets put
00:07 chris }
00:07 chris elsif (@$tags[$i] eq '942'){
00:07 chris actually scrap that
00:08 chris aftere the else {
00:08 chris and before the if ($first){
00:08 chris lets try
00:08 chris if (@$tags[$i] eq '942'){
00:09 chris warn "DEBUG tag=942 subfield=@$subfields[$i]";
00:09 chris }
00:10 chris then at about line 1067 where we have
00:10 chris else {
00:10 chris                        my $ind1 = substr(@$indicator[$j],0,1);
00:10 chris after the else {
00:10 chris put the same if in
00:10 chris maybe DEBUG 2
00:10 kados only gets called twice
00:10 chris for one
00:10 kados (the first one)
00:11 chris so we can see which one isnt getting called
00:11 kados k
00:12 chris one should get called once, and one should get called twice
00:13 kados [Tue Mar 07 20:13:41 2006] [error] [client] DEBUG tag=942 subfield=k at /home/nbbc/koha/intranet/modules/C4/ line 1086., referer:[…]iblionumber=41044
00:13 kados [Tue Mar 07 20:13:41 2006] [error] [client] TAGS ETC:942 c PERI at /home/nbbc/koha/intranet/modules/C4/ line 1048., referer:[…]iblionumber=41044
00:13 kados [Tue Mar 07 20:13:41 2006] [error] [client] DEBUG tag=942 subfield=c at /home/nbbc/koha/intranet/modules/C4/ line 1086., referer:[…]iblionumber=41044
00:13 chris ah ha
00:13 kados that's all we've got
00:13 kados the other one doesn't get called
00:13 chris so its the first one back up by 1067 thats not getting called eh
00:14 kados hmmm
00:14 kados strangely, before I put that one in I got:
00:14 kados [Tue Mar 07 20:11:00 2006] [error] [client] DEBUG tag=942 subfield=a at /home/nbbc/koha/intranet/modules/C4/ line 1083., referer:[…]iblionumber=41044
00:14 kados [Tue Mar 07 20:11:00 2006] [error] [client] TAGS ETC:942 k 315 at /home/nbbc/koha/intranet/modules/C4/ line 1048., referer:[…]iblionumber=41044
00:14 kados [Tue Mar 07 20:11:00 2006] [error] [client] DEBUG tag=942 subfield=k at /home/nbbc/koha/intranet/modules/C4/ line 1083., referer:[…]iblionumber=41044
00:14 kados sorry I missread you
00:14 kados yes, 1067 isn't called
00:15 chris ah ha
00:16 kados is 9XX < 10?
00:16 kados what'd you find?
00:16 chris if it was that wed get a controlfield
00:16 chris whats just before 942?
00:17 kados 856
00:17 kados $3: Table of contents
00:17 kados $u:[…]8/2005005430.html
00:17 chris my theory is for some reason $first isnt set
00:17 kados for any of the first 900s
00:17 kados cause it happens in 970 too
00:20 chris ok line 1054
00:21 chris $first=1;
00:21 chris how bout we do the if in here
00:21 chris and see if thats what happening
00:22 kados only called once
00:22 kados [Tue Mar 07 20:23:03 2006] [error] [client] DEBUG FIRST tag=942 subfield=k at /home/nbbc/koha/intranet/modules/C4/ line 1058., referer:[…]iblionumber=41044
00:22 kados $k was the first one
00:23 chris ok so now we know whats happening
00:25 chris interesting
00:27 kados what's that?
00:27 chris can u try something for me
00:27 kados sure
00:27 chris line 1056
00:27 chris else {
00:27 chris comment this out
00:28 kados 1056 for me is $first=1;
00:28 chris ok the else after that
00:28 chris must be 1058
00:28 chris then down about 1075 (or 1077)
00:28 chris               }
00:29 chris comment that out too
00:29 kados k
00:29 chris and then give it a whirl and see how mental the xml looks now
00:29 chris :)
00:30 kados so ...
00:30 kados that worked except that it merged our 650s into one :-)
00:31 chris darn
00:32 chris so we onl get one 650 section now?
00:32 kados yea
00:32 kados lemme double check that
00:33 kados yea ... it merged them
00:33 kados[…]
00:33 chris and if we uncomment them we get the 650's
00:33 chris but we lose the first tag of the 900 ones
00:34 kados :-)
00:34 kados but only on this machine :-)
00:35 chris ahh it ignores the blank ones that we put in to sep the 650's
00:36 chris if we comment it out
00:36 chris so sees it all as one tag
00:36 kados right
00:38 chris im still puzzled as to whats different on here nd on liblem
00:38 chris liblime even
00:38 kados yea
00:38 kados is it possible it's perl?
00:39 chris doubt it very highly
00:39 chris oh hey
00:39 chris
00:39 chris look at the 856
00:40 kados in the xml?
00:40 chris ahh scratch that, the other subfields are empty thats ok
00:40 chris what happens if you put a value in for 856x ?
00:41 kados here's something stranger
00:41 kados if I create a minimal record
00:41 kados with just 100$a and 245$a and three values in 942
00:41 kados it works fine
00:41 kados I'll try 856x
00:41 chris right
00:42 kados (I reverted back to before the comments)
00:42 chris righto
00:43 kados no change
00:43 kados[…]
00:43 kados 856 $x works fine
00:43 kados could there be a difference in the include files in the template?
00:43 chris so the 942 works .. if 856 x is set?
00:43 kados naw ... guess that woudl show up on the way in
00:44 kados no
00:44 kados 942 still strips one value out
00:44 chris but 100, 245 and 942 works
00:44 chris if you have nothing in 856
00:44 chris does it work?
00:45 kados no
00:45 kados it's _just_ the 9XX fields
00:45 chris but
00:45 kados I tried a warn inside the <10 ... nothing
00:46 chris if its 100a 245a and 942 it works
00:46 kados yep
00:46 kados true
00:46 chris can we slowly remove tags
00:46 chris until it works
00:46 chris hehe
00:46 kados hehe
00:47 chris ok, lets take a break, and ill try looking at it with fresh eyes this evening
00:47 chris theres got to be something we are missing
00:48 kados huh
00:48 kados I deleted 8XX and the second set of 65X
00:48 kados and now it works
00:48 chris and now it works
00:48 chris mental
00:49 kados I"m gonna readd and see what happens
00:49 chris good idea
00:51 kados ha!
00:51 kados so i added the second set of 650s
00:51 kados worked fine
00:51 kados then i added the 856
00:51 kados dropped
00:52 kados must be something about the 856 then
00:52 chris looks that way
00:52 kados ok ... I'll tell NBBC not to edit the catalog tomorrow
00:52 kados that we're still working on it
00:52 chris could it be the 856$3
00:52 kados I'll try that
00:53 kados nope
00:53 kados lemme try the $u
00:54 kados nope
00:54 kados with any value in 856
00:55 kados either field or both
00:55 kados it drops the first from 942
00:55 chris right
00:55 kados soon as I get rid of those it's fine
00:56 kados I'm gonna check the framework
00:56 chris k
00:57 kados but really, that woundn't have anything to do with it
00:57 kados it's still happening inside MARChtml2xml
00:57 chris yes
00:57 kados lemme modify my little record and put a 856 in there
00:57 chris and we know its matching                 if (!$first){
00:57 chris                    $xml.="</datafield>\n";
00:57 chris when it should be the first
00:58 kados yep, that did it
00:58 chris so its def 856
00:58 kados dropped one from 942
00:58 kados yep
00:58 chris but on liblime, 856 works ok
00:58 kados lemme check that
00:59 kados yea
00:59 kados I added the exact record in fact
00:59 kados and it worked fine
00:59 chris right
00:59 chris no sir i dont get it
00:59 kados eep
01:00 kados someone musta messed with liblime's frameworks again
01:00 chris no biblionumber agin?
01:00 kados naw ... looks ok
01:01 kados ok ... I give up
01:01 kados tomorrow's another day
01:01 kados thanks for the help chris
01:01 kados can't wait to get this bill :-)
01:01 chris np im sure we'll spot it with fresh eyes
01:02 chris once i get these deletes working
01:02 chris i think we are ready fro serious testing on the plugin
01:04 kados sweet
01:05 kados meadville should be working on getting it going this week
01:05 chris cool
01:05 kados well I've got to head to bed
01:05 chris sleep well
01:05 kados read you tomorrow
01:05 chris dream of 856
01:05 chris :)
05:56 hdl_away hi
05:56 hdl always trying to understand how to search.
05:57 hdl I entered a book.
05:57 hdl I try to search for this book.
05:57 hdl No way.
05:58 hdl Does anyone have a clue ? chris, paul, kados
05:59 hdl should I recompile the whole stuff (ZOOM, zebra, yaz, reverse order) with locale utf-8 ?
06:08 pierrick_ Hi koha team :-)
06:18 pierrick_ bonjour Paul
06:18 paul hello
06:20 pierrick_ Are ppl using CGI::Carp sometimes on Koha ?
06:20 paul pas que je sache.
06:20 paul (et je sais même pas ce que c'est 8-) )
06:21 pierrick_ je vais essayer aujourd'hui et je ferai un mail d'explication sur koha-devel
06:21 pierrick_ c'est un module de gestion d'erreur CGI
06:21 pierrick_ parce que lorsqu'on a un erreur 500, on est limite limite sur la détection d'erreur
06:22 pierrick_ (enfin moi en tout cas, parce que je ne maîtrise pas encore grand chose)
06:37 chris pierrick CGI::Carp is really good to use when developing
06:37 chris especially fatalsToBrowser
06:38 pierrick_ yes, that's what I read in Perl Cookbook
06:38 paul hello chris.
06:38 paul have a good night !
06:39 chris hello paul
06:45 chris search still not working hdl?
06:46 |hdl| no
06:47 chris right, you have zebrasrv running ?
06:47 chris have you tried with yaz-client?
06:47 chris yaz-client host:port/database
06:47 chris eg for me
06:47 chris yaz-client localhost:2100/koha2
06:50 |hdl| I did.
06:50 |hdl| I tried find test
06:50 |hdl| or find montest
06:50 |hdl| (two biblio I just created.)
06:50 |hdl| No results.
06:51 chris hmmm
06:52 chris can you see them in koha?
06:52 chris eg
06:52 chris kohaurl/cgi-bin/koha/
06:53 |hdl| creation gave me this in zebra :
06:55 chris ahh you know whats happened
06:55 chris i think it hasnt done a commit
06:55 chris ie kados has got it using shadow registers now
06:56 chris but after the record is added its not committing
06:57 chris so its in the shadow cache
06:59 |hdl| and it works better.
06:59 |hdl| But not well :
07:00 |hdl|
07:01 |hdl| no mapping found problem
07:01 chris hmm
07:01 chris the framework is all ok right?
07:02 |hdl| should be.
07:02 chris hmm im not sure what that error is
07:02 chris it looks like its MARC::Charset not liking the utf-8
07:03 chris i might have to ask kados about that
07:03 chris at least its returning 13 records now :)
07:03 |hdl| some weird characters in labels, but that is not fwk, more like MARC::charset
07:04 |hdl| Yes but ending on internal server error :5
07:04 chris yeah
07:04 |hdl| :5
07:04 chris i dont know much about MARC::Charset im afraid
07:05 |hdl| So commit is quite critical when it comes to shadow register management
07:06 chris yes
07:06 chris we need to make sure that after every update we commit
07:06 chris cool, ill go to sleep :)
07:06 chris and ill look into it tomorrow
07:06 |hdl| good night.
07:07 |hdl| Thanks
07:07 chris good night
07:07 |hdl| Koha developped 24h a day... :D
07:09 |hdl| paul : tu as ce genre de problème
08:00 pierrick_ did you notice how fast emails are arriving on koha-cvs ? :-)
08:01 paul yep
08:01 paul savannah eated something illegal I suppose.
10:14 osmoze hello
10:14 paul hello JS
10:14 paul tu as lu le mail d'hdl sur la liste ?
10:14 paul on attend que tu te portes candidat !
10:15 osmoze je viens de le lire a l instant
10:16 osmoze bien sur que je me porte candidat, la seule question reste a savoir comment rendre cela possible dans un double environnement et etre reelement efficace
10:17 paul en fait, on souhaite juste vérifier que la doc est complète et qu'on obtient un Koha fonctionnel.
10:17 paul pas la peine de migrer tes données !!!
10:18 osmoze ok
10:18 paul en faisant le tout comme un vrai newbie !
10:18 osmoze huhu ^^
10:18 osmoze |hdl|>  envoi moi le lien de telechargement
10:18 osmoze a moins que j utilise la version cvs
10:20 osmoze paul > il sera toujours temps de migrer les donner pour tester en environnement reel...Par contre, l upgrade marche t elle depuis la 2.0 ?
10:20 osmoze aussi, es ce la version qui utilise zebra ?
10:21 paul nope.
10:21 paul C'est juste une version officielle.
10:21 paul pour permettre à un newbie de tout installer (perl, mySQL, apache...) et configurer
10:21 paul l'idée c'est de mettre un truc complet et à jour pour les petites structures qui veulent un truc en monoposte.
10:22 paul et ne pas avoir besoinde renvoyer trop de monde vers PMB ;-)
10:22 osmoze ok, j ai un pc sous windows qui ne sert pas, je serais heureux de vous amener mes critiques ( bonnes ou mauvaises ;))
10:25 paul (il va avoir un paquet sur sa machine de dispo)

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