IRC log for #koha, 2005-02-19

All times shown according to UTC.

Time S Nick Message
12:33 owen Wow, paul, lots of new stuff!
12:33 paul hi owen.
12:34 paul nope, mostly french & italian translations
12:34 owen Are you working on Italian?
12:51 paul nope, but i've installed kartouche for translation to various iso-8859-1 languages
12:54 owen Very interesting.  I hadn't seen anything like that before.
13:04 paul chris, rach, & other katipo folks :
13:04 paul http://katipo.co.nz/solutions/products/koha.html
13:04 paul "the page does not exist :-("
13:24 owen That's not good!
13:24 owen Paul: the BiblioDefaultView preference--is that just for the OPAC?
13:25 paul yes
13:25 paul it defines the default view for OPAC.
13:25 paul if you think it's useful for librarian too, it could be done probably
13:26 paul (although a little bit tougher, due to the number of pages)
13:26 owen It would be useful to us--I'm going to be changing the link  in the intranet to point to detail.pl instead of MARCdetail.pl
13:26 paul you also have the "opacstylesheet" that can be used to define a stylesheet external to the templates
13:26 paul (being kept after upgrades)
13:26 kados well we may need 'normal view' because MARCdetail isn't working for us :-(
13:27 kados morning paul
13:27 paul (no, evening ;-) g'morning to you josh)
13:27 paul but you must find why the MARCdetail don't work for you
13:27 paul it may be a hidden database problem !
13:27 kados :-)
13:28 kados yes ... working on that now ... I reimported the data and running updatedatabase on it now
13:28 kados I hope not!
13:28 paul (henri damien laurent is going back home. Will commit the 1st report stuf tomorrow morning
13:28 paul )
13:28 kados great!
13:28 paul (was with me today, usually working through webcam)
13:29 kados paul: here are some errors
13:29 kados DBD::mysql::st execute failed: Duplicate column name 'authtypecode' at updatedatabase line 1243.
13:30 kados DBD::mysql::db do failed: All parts of a PRIMARY KEY must be NOT NULL;  If you need NULL in a key, use UNIQUE instead at updatedatabase line 1262.
13:30 kados DBD::mysql::db do failed: All parts of a PRIMARY KEY must be NOT NULL;  If you need NULL in a key, use UNIQUE instead at updatedatabase line 1265.
13:30 kados Modifying marc_word (concat on tag and subfield for better perfs)
13:30 kados DBD::mysql::db do failed: Duplicate column name 'tagsubfield' at updatedatabase line 1280.
13:30 kados updatatedatabase is giving me
13:31 kados this is because I added these tables earlier (before 2.2 was ready) so our opac would have marc searching
13:41 paul the 1st (1243) error should mean you have both thesaurus_category & authtypecode columns
13:42 paul the 2nd (1262 & 1265) are not problems, but should be solved for better perfs
13:42 paul the last (1280) needs some checkings
13:42 paul do you have a tagsubfield column in marc_word table ?
13:42 kados yes ... added to allow marc searching
13:43 paul does it contain tag+subfield values
13:43 paul ?
13:43 kados yes
13:43 paul so, the updatedatabase is not the culprit.
13:43 kados :-)
13:43 kados right ... I added these tables earlier so our opac would have marc searching
13:44 paul in MARCgetbiblio, you should add a warn "".$record->as_formatted just before the return
13:44 paul thus, you can see if the marc record is correctly build
13:44 kados ok
13:44 paul (if yes, the pb is after the marc building)
13:49 paul kados : 2.2.1 will be released tommorow afternoon or monday
13:50 paul (the package is ready, just some more tests & it's here)
13:52 kados great!
13:52 kados ok ... I added that line and the marc record is showing up
13:53 kados in the log
13:53 paul good news, it's not a MARC problem.
13:53 kados whew!
13:53 paul so, it's a perl or template one.
13:54 paul it does not appear in opac or in librarian interface ?
13:54 kados hmmm, strange since we checked out fresh CVS copies of everything
13:54 kados yes
13:54 kados does not appear in either
13:54 owen And it's the same in default/css and npl templates
13:54 paul what is shown if you try to add a new biblio ?
13:55 paul do you see something ? can you add a biblio & then see it ?
13:55 kados checking
13:56 kados strange ... the addbiblio.pl screen is blank
13:56 kados the Add Bibliographic Record link doesn't do anything
13:56 kados (updatedatabase is still running though)
13:56 kados (so that may be why)
13:57 paul select tab,count(*) from marc_subfield_structure group by tab
13:57 paul gives what ?
13:57 paul should give entries with -1,0,1,2,...,10
13:58 kados |   -1 |     1034 |
13:58 kados |    0 |       14 |
13:58 kados |    1 |       27 |
13:58 kados |    2 |        9 |
13:58 kados |    3 |       23 |
13:58 kados |    4 |       17 |
13:58 kados |    5 |        3 |
13:58 kados |   10 |        6 |
14:04 paul ok, so this table is correct too
14:04 paul I must say it's really a mistery...
14:04 kados we're baffled too
14:04 paul you are speaking of MARCdeail.pl ?
14:04 kados yes and opac-MARCdetail.pl
14:07 paul you should try to set some warn "" in the loops
14:08 paul to see if they are reached
14:08 paul mmm... another idea
14:08 paul what is marc_subfield_structure for you ?
14:08 paul (field list)
14:09 kados | tagfield           | char(3)      |      |     |         |       |
14:09 kados | tagsubfield        | char(1)      |      |     |         |       |
14:09 kados | liblibrarian       | varchar(255) |      |     |         |       |
14:09 kados | libopac            | varchar(255) |      |     |         |       |
14:09 kados | repeatable         | tinyint(4)   |      |     | 0       |       |
14:09 kados | mandatory          | tinyint(4)   |      |     | 0       |       |
14:09 kados | kohafield          | varchar(40)  | YES  |     | NULL    |       |
14:09 kados | tab                | tinyint(1)   | YES  |     | NULL    |       |
14:09 kados | authorised_value   | varchar(10)  | YES  |     | NULL    |       |
14:09 kados | thesaurus_category | varchar(10)  | YES  |     | NULL    |       |
14:09 kados | value_builder      | varchar(80)  | YES  |     | NULL    |       |
14:09 kados | seealso            | varchar(255) | YES  |     | NULL    |       |
14:09 kados | authtypecode       | varchar(10)  | YES  |     | NULL    |       |
14:09 kados | hidden             | tinyint(1)   | YES  |     | NULL    |       |
14:09 kados | isurl              | tinyint(1)   | YES  |     | NULL    |       |
14:09 kados | frameworkcode      | varchar(4)   | YES  | MUL | NULL    |       |
14:09 kados | link               | varchar(80)  | YES  |     | NULL    |       |
14:13 paul the updatedatabase is stille running i bet ?
14:13 kados yes
14:14 paul the thesaurus_category must be dropped (probably manually) after the updatedatabase
14:14 paul but that should not be the origin of your problem
14:14 paul I go back to the suggestion of putting "warn" into the loops
14:15 kados ok
14:16 kados which loops would be the best to check?
14:18 paul for (my $x_i=0;$x_i<=$#fields;$x_i++) {
14:18 paul (line 97)
14:19 kados ok
14:21 kados I get lots of numbers like this:  2 => 7
14:21 kados 2 => 9
14:21 kados etc ...
14:24 paul (what did you warn)
14:25 paul i bet $tabloop & $x_i
14:25 paul so the loop is done. more & more mysterious...
14:26 kados warn "$tabloop => $x_i";
14:26 kados yes
14:26 paul did you look at the html source ? do you "see" the loop (with empty values)
14:26 paul or not ?
14:26 owen No.
14:26 paul no to which question ?
14:26 paul look or see ?
14:26 paul ;-)
14:27 kados :-)
14:27 owen You can't see the loop in the source.  sorry, I just saw the second part :)
14:27 paul mmm... a stupid idea : what does hidden contain (in marc_subfield_structure)
14:27 kados I don't see it either: http://66.213.78.67/cgi-bin/ko[…]tail.pl?bib=16595
14:28 paul ok, nothing here I agree.
14:28 kados all nulls
14:28 kados 20 NULLs
14:29 paul just 20 ?
14:29 kados when I do select hidden from marc_subfield_structure limit 0,20;
14:29 paul ok
14:29 kados select hidden from marc_subfield_structure where hidden is not NULL limit 0,20;
14:29 paul select distinct hidden from marc_subfield_structure
14:29 kados yeilds nothing
14:29 paul ok, so it'not because you've asked to hide everything ;-)
14:29 kados NULL
14:30 kados :)
14:30 paul really really really really strange...
14:30 paul should be easier to convince your president that there is no massive weapons in Irak than understand this bug...
14:30 kados :-)
14:31 kados I think he knew all along ;-)
14:31 kados but yes
14:31 kados this bug is very strange
14:32 paul can you edit the marc biblio ?
14:32 kados from addbiblio.pl?
14:32 paul yep
14:32 kados no ... it shows up as blank too
14:33 paul cgi-bin/koha/acqui.simple/addb​iblio.pl?oldbiblionumber=16595
14:33 kados right
14:33 owen and it shows up blank too if you try to start with an empty record
14:34 owen so it seems that it isn't related to retrieval, it's related to building the marc display?
14:34 paul yes
14:34 paul seems
14:34 paul what is your marc_tag_structure table ?
14:34 paul (structure & content size)
14:35 kados | tagfield         | char(3)    |      |     |         |       |
14:35 kados | liblibrarian     | char(255)  |      |     |         |       |
14:35 kados | libopac          | char(255)  |      |     |         |       |
14:35 kados | repeatable       | tinyint(4) |      |     | 0       |       |
14:35 kados | mandatory        | tinyint(4) |      |     | 0       |       |
14:35 kados | authorised_value | char(10)   | YES  |     | NULL    |       |
14:35 kados | frameworkcode    | char(4)    | YES  |     | NULL    |       |
14:35 kados | hidden           | tinyint(1) | YES  |     | NULL    |       |
14:35 kados select count(*) from marc_tag_structure;
14:35 kados |      108 |
14:36 paul (i've no "hidden" field)
14:36 paul you could delete it
14:36 kados ok ... wonder how it got there
14:36 paul (but i don't think it would change anything, but you can give it a try)
14:37 paul ! another idea !
14:37 paul sub get_authorised_value_desc ($$$$$) {
14:38 paul add a return $value just after the
14:38 paul   my($tag, $subfield, $value, $framework, $dbh) = @_;
14:38 kados ok
14:38 paul (that will not change anything : addbiblio should work at least)
14:39 kados hmmm, didn't seem to do anything
14:39 kados should I "warn $value"?
14:40 paul no, just return the value (=> do nothing)
14:40 paul (it's a new sub, maybe it's the origin of your pb)
14:40 paul so my idea is to de-activate it
14:40 kados hmmm, warn $value doesn't show up in the log
14:40 paul mmm... interesting
14:41 paul should have (is called at line 124)
14:41 kados meaning never called eh?
14:41 paul so, the line 124 is never reached
14:42 paul i'll have to leave for dinner
14:42 kados but line 97 is
14:42 paul but that may be a good point
14:42 kados ok ... thanks for the help paul
14:42 kados could we make an appointment to further troubleshoot?
14:42 kados maybe 2:00am for me ;-)
14:42 paul look if the tag()<10 is reached or not
14:42 paul not tommorrow (teaching Koha at CMI)
14:43 paul (so, monday)
14:43 kados darn ... we were hoping to deploy 2.2 over weekend
14:43 kados the tag()<10 is not reached btw
14:44 paul (test the } else { line 108
14:45 paul the line 112 (while)
14:45 paul then line 115, after the 2 next if
14:45 kados yes ... 108 is reached
14:45 paul you shoud find where Koha exit
14:45 paul (ok, I must leave now, sorry)
14:49 kados thanks for the help paul!
14:49 kados adding warn after the next ifs doesn't work ... so that's probably the problem
15:12 kados yep ... should be set to go
15:12 kados paul if you pop back in here's something interesting
15:13 kados I put a warn $tagslib->{$fields[$x_i]->ta​g()}->{$subf[$i][0]}->{tab}; just before the first next if
15:13 kados and it gives me "Warning: something's wrong" in the log
15:40 kados here's an interesting thing ... on 101 (the server that's working) marc_subfield_structure looks like this:
15:40 kados | tagfield           | char(3)      |      | PRI |         |       |
15:40 kados | tagsubfield        | char(1)      |      | PRI |         |       |
15:41 kados on 69 there is no PRI in Key column
15:41 kados I wonder if that's our problem
15:41 kados it could also just be because updatedatabase hasn't finished
16:06 paul_away kados ?
16:06 paul_away (here after dinner. I had an idea)
16:06 paul_away I've FOUND YOUR PROBLEM
16:06 paul_away owen, kados, come, come
16:06 paul_away owen, kados, come, come
16:07 paul_away marc_tag_structure
16:07 paul_away frameworkcode  is NULL default NULL
16:07 paul_away THAT'S WRONG
16:07 paul_away should be
16:07 paul_away frameworkcode NOT NULL default ''
16:07 paul_away with null default null, I bet USD100000000 that you have null everywhere
16:08 paul_away so the tag list is always empty
16:08 paul_away and the MARCdetail... is empty too...
16:08 paul_away so the next unles... is always used
16:08 paul_away & the template stay empty !
16:08 paul_away you know what ? I'm happy
16:10 nausicaa ... juste en passant... j'ai réinstallé koha et j'importe des notices bientôt
16:10 nausicaa ;-)
16:12 nausicaa bon appétit et bonne nuit...
16:12 nausicaa @ bientôt
16:16 kados ok ... here
16:16 kados wow ... paul .. thanks!!!
16:17 kados so how do I fix it?
16:17 kados ;-)
16:18 kados alter table marc_tag_structure set frameworkcode NOT NUll default ''; ??
16:18 kados paul_away: still there?
16:18 paul_away yep
16:18 kados great!
16:19 paul_away i was with my baby.
16:19 paul_away and suddenly.
16:19 kados hehe
16:19 paul_away he told me where was your problem !
16:19 kados hehe
16:19 paul_away nice baby ;-)
16:19 paul_away ok, i leave now. Have a good day
16:20 kados wait !! how do I fix it?
16:20 kados :-)
16:20 paul_away still here
16:20 paul just update marc_tag_structure set frameworkcode=''
16:20 paul then your alter table
16:21 paul (the alter table being for future marc add, the update being for what exists)
16:21 kados ?
16:21 paul the update solves the problem NOW
16:21 kados ok
16:21 paul the alter will prevent any future problem if you edit the marc structure & add a new marc tag
16:22 paul (that would have frameworkcode empty again !)
16:22 kados ahh ... so I run 'update marc_tag_structure set frameworkcode='''
16:23 kados then 'alter table marc_tag_structure set frameworkcode='''?
16:23 paul yep
16:23 kados great ... thanks paul!
16:23 paul (the alter can't be run before there is no more empty values)
16:23 paul note you have to do this in marc_tag_structure too
16:24 kados ?
16:24 kados I thought we were talking about marc_tag_structure
16:24 paul oups, sorry... you have to do this in marc_subfield_structure too
16:24 kados ahh
16:24 kados ok ...
16:24 kados I am also thinking I will drop the tagsubfield from marc_word before the next time I run updatedatabase
16:25 paul mmm... you will also have problem with primary key.
16:25 kados yes we have that problem ^^ see above
16:25 paul (you can't have the same as me, as frameworkcode is in my PK, and can't be null)
16:26 paul right. I've missed the real origin of the problem
16:26 kados I'm a bit lost :-)
16:27 kados I'd really like to get our database "standard koha" if possible but I'm not sure how
16:27 kados somehow we've got extra fields (like the hidden)
16:27 kados and I've a feeling our indexes are different, etc.
16:29 kados sigh
16:54 paul kados : between 2 official versions of Koha, updatedatabase is as intelligent as possible.
16:54 paul but for ppl that have non-standard DB it can't
16:54 paul i've updated successfully from 2.0 to 2.2.0 3 libraries.
16:54 paul updatedatabase works like a charm
16:54 paul ok, I go to bed now
16:55 paul happy to have be able to help you
17:03 kados thanks paul!
18:25 owen rosa, I don't remember--are you a library person?
18:28 owen Koha's renewal behavior has changed recently and I'm trying to figure out if it's the expected behavior for anyone.
18:58 rosa hi owen. can't help sorry. Yes I am a library person, but we haven't implemented 2.2 yet
18:59 chris rosa: how 2.2 is working is different to 1.2 and 2.0
18:59 rosa that's why I can't help, I assumed
18:59 chris the way it works is when u renew, instead of extending the due date to today + issuing period
18:59 chris which is how it used to work
19:00 chris its makes the new due date be old due date + issuing period
19:00 chris i think owen is wondering if this is desirable ... it doesnt sound right to me
19:00 rosa oh, that's tricky
19:00 rosa and unpopular, methinks
19:00 chris i think the way it worked before was less suprising
19:00 kados I haven't looked at the code yet but we need it to act as before
19:00 kados yes
19:01 kados I'm thinking it can't be that hard of a fix
19:01 chris yeah
19:01 rosa Old birds who renew their books today wanta nother three weeks, not two weeks and 3 days if they're already overdue
19:01 chris yeah
19:01 chris thats what id expect
19:02 kados btw: chris (since your about) the updatedatabase command has been running for about 7 hours ... when i do mysqladmin proc it shows a process in SLEEP mode ... does that mean it's died?
19:02 chris sounds like it
19:02 kados there are actually two processes listed ...
19:02 kados 148 | kohaadmin | localhost | Koha | Query   | 4380  | copy to tmp table | create index Search_Marc on marc_word (tagsubfield,word)
19:03 kados 293 | kohaadmin | localhost | Koha | Sleep   | 10347 |                   |
19:03 kados is there some way for me to tell if 293 is a child of 148?
19:03 chris probably .. but im not sure how
19:03 kados k
19:04 kados well I'll wait a couple more hours
19:04 chris id expect it to take 10-15 mins
19:04 kados hmmm, our database is quite large now ... 1.5 gig
19:04 chris even so
19:04 kados the import takes about 5 hours
19:04 kados ahh
19:04 chris yeah the import does a lot more
19:04 chris update database should be just running some alter tables
19:05 chris maybe if its building new indices it might take a while ..
19:05 chris not sure
19:05 kados it's creating indexes too
19:05 kados I'll wait just to be sure
19:09 owen rosa and chris, thanks for the input on the renewals thing.
19:09 owen I'm glad I'm not crazy thinking it's not working as expected.
19:09 chris i think it would just have been an oversight when paul was rewriting the circ code
19:10 chris i doubt it was done on purpose
19:10 owen I'm surprised it took me so long to notice.  I assume it's been that way for a little while now.
19:11 chris since 2.2.0 anyway .. i dont think many ppl have that in production yet tho
19:12 kados renewscript.pl still how renewals are done?
19:12 chris couple of places
19:12 chris there and in issues
19:12 kados right
19:13 kados looks like both use renewbook in Circ2.pm
19:13 chris id hope so :)
19:14 kados this line:
19:14 kados $datedue = UnixDate(DateCalc($iteminforma​tion->{date_due},"$loanlength days"),"%Y-%m-%d");
19:14 chris there ya go
19:14 kados so instead of $iteminformation->{date_due
19:14 kados it should be today ... right?
19:14 chris "today"
19:14 chris yep
19:14 chris that oughta fix it
19:17 kados hmmm, didn't seem to fix it
19:33 kados nevermind ... it is working
19:33 kados thanks chris!

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