IRC log for #koha, 2006-04-12

All times shown according to UTC.

Time S Nick Message
12:01 tim yeah.  That's fine.
12:01 tim Thanks!
12:02 kados np
12:05 kados T-MINUS 2 HOURS TO KOHA MEETING
13:06 owen tim: I just committed a fix for your deletions problem
13:15 tim That was quick owen
13:15 owen It was a silly template error on my part, easily fixed.
13:16 tim Some day I hope to be able to make (and fix) silly template errors.
13:35 kados tim: I've updated your install with the new changes
13:41 kados T-MINUS 25 MINUTES TO KOHA MEETING
13:54 tim Thanks kados and owen
14:01 kados well by my clock it's 19:00 GMT ...
14:01 kados anyone around for the meeting?
14:02 kados hey pierrick
14:02 kados looks like it's just you and me :-)
14:02 pierrick rally?
14:02 pierrick really?
14:03 kados so far yes :-)
14:03 pierrick what time is it in Ohio?
14:03 kados 3:00 pm
14:03 pierrick 9:04 pm in France
14:03 pierrick (strange offset ;-)
14:05 kados :-)
14:06 pierrick I've worked with hdl today, interesting working session
14:06 kados what did you work on?
14:06 pierrick we worked separately, but we had discussion about how Koha works (the application, the project)
14:07 kados ahh, cool
14:07 kados any conclusions? :-)
14:08 pierrick I was wondering whether each Koha supplier (Paul, Liblime, slef, Katipo) had a local version manager to work on customer specific branches
14:08 pierrick Paul and hdl don't
14:08 pierrick I think I'll do it for INEO
14:08 kados imo all customer-specific code should be managed in sysprefs
14:08 pierrick we spoke about localization, template status for 3.0 and so on
14:09 kados otherwise the koha project looses local customization
14:09 pierrick kados: you're right, but IMO adding a feature on a stable branch is dangerous
14:09 kados why?
14:09 pierrick you never know how many bugs your new feature is bringing
14:10 kados and that's where QA comes in :-)
14:10 pierrick anyway, this is the release manager role to give rules about that :-)
14:10 pierrick :-)
14:10 kados I suspect we are approaching Koha development quite differently than other software projects
14:10 kados of course, RM makes the rules about that
14:11 pierrick kados: how does Koha differs from other opensource projects?
14:11 kados I've found that it's a nightmare to try to maintain multiple client installations using a non-stock CVS
14:11 kados well, we don't have very many users/developers compared to the size/complexity of the code bases for one
14:12 pierrick (maybe we should think about the tool we use)
14:12 pierrick CVS was not designed to manage branches
14:12 kados right
14:12 pierrick so evidently, working on many branches is not a good idea, we should only have a few branches (dev/stable)
14:13 kados I think two branches is best
14:13 kados three at the most
14:13 pierrick maybe we'll think about that when we move to Subversion :-)
14:13 kados moving to subversion would require retraining all koha developers and some koha users
14:14 kados so while I agree in theory
14:14 kados i think it might be tough to pull off successfully
14:14 pierrick kados: moving to Subversion is a really easy task I can assure you
14:14 kados cool
14:14 pierrick Subversion was designed to replace CVS and to be compliant in the way it works
14:14 kados the other worry I have is that if we use subversion folks will have less incentive for contribbuting their changes back
14:15 pierrick why???
14:15 kados 1. because they will not be familiar with sv
14:15 kados 2. because they will not know now to create their own sv 'branch'
14:15 kados 3. because even if they do, it will be hard for the RM to keep track of all changes and have an 'authoritative' branch
14:16 kados but maybe I don't understand well how it works
14:16 pierrick kados, I think you should not worry so much :-)
14:17 kados hehe
14:17 pierrick kados: do we say the meeting is adjourned?
14:17 kados pierrick: almost
14:17 kados pierrick: in your opinion, is the new wiki ready?
14:18 kados pierrick: if so I will announce it on koha-devel
14:18 kados pierrick: also, do you think we have time for a bug squash mtg before paul releases 2.4?
14:18 pierrick IMO remaining migration issues need manual work
14:18 kados pierrick: IMO we should have one because surely 2.4 has bugs
14:19 pierrick bug squashing next week for example?
14:19 kados that's good for me
14:19 kados pierrick: final question
14:19 pierrick Paul have to say, I'll organize the party, but I need Paul to say
14:20 kados pierrick: I need instructions for UNICODE
14:20 kados no, you can organize the party without paul's approval
14:20 kados you are the QA Manager now :-)
14:20 pierrick kados, OK, I'll give a date for the BSP as soon as possible
14:21 kados great!
14:21 pierrick so Unicode?
14:21 kados yes
14:21 kados I would like to confirm that head is 100% unicode safe
14:22 pierrick I can tell you about MySQL communication, but not about zebra
14:22 kados sure, I already know zebra :-)
14:22 pierrick IMO, UTF8 is working fine on HEAD
14:23 kados did you alter your db schema or any code to get it working?
14:23 pierrick My tests are successful, I didn't work anymore on this subject
14:23 pierrick I did
14:23 kados ok
14:23 pierrick my code modification was commited
14:23 kados can you commit your changes?
14:23 kados ahh, great
14:24 kados ok, so meeting adjourned then :-)
14:24 kados have a good evening
14:24 kados I'll read you tomorrow :-)
14:24 pierrick my database modification was not commited, I made the conversion manually, but today I saw hdl had commited the conversion to UTF8 in updatedatabase
14:24 kados unless you have anything more to add
14:24 kados ahh ... is hdl's commit the same as your db modif?
14:25 pierrick yep, if you have time to tell me about the new stopwords management page I wrote, it would be great :-)
14:25 kados ok, I'll try to look at it today
14:25 pierrick kados: I need a check about hdl commit
14:26 owen pierrick: how is the number of items per page controlled with the new pagination function?
14:26 pierrick each page has its own $pagesize
14:27 owen Ah, I see it.
14:27 pierrick having a common $pagesize wouldn't be a good idea
14:27 pierrick IMO
14:27 owen I agree
14:27 pierrick listing stopwords and budgets is not the same thing
14:28 pierrick owen: your opinion about the pagination bar?
14:29 owen Is in use somewhere else now too?
14:29 owen besides stopwords and budgets?
14:29 pierrick yep, itemtypes, stopwords, authorities
14:29 owen that's right, itemtypes
14:29 pierrick (budgets not commited yet)
14:32 owen itemtypes is only displaying 5 items at a time
14:32 pierrick owen, when will the "beautiful" template work start for 3.0? is this action planned?
14:32 pierrick owen: yes, I commited an extreme test value :-(
14:33 owen I'm not really sure. The last time we had an in-depth discussion of new templates for 3.0, I thought Katipo had agreed to work on a new design
14:34 owen I was to help code it. But I haven't heard anything about that since then.
14:35 pierrick OK, because in the current state, we can't show the prog template to any customer
14:35 pierrick (IMO)
14:35 owen about the pagination bar: I worry each time I see HTML generated by a script, because as a template designer I see that as a limitation.
14:35 owen However, the HTML generated by the pagination script is very clean
14:35 pierrick owen: I understand, but what a simplification !
14:35 pierrick owen: thank you
14:35 pierrick I've tried to have semanticaly correct CSS classes
14:36 pierrick I'm sure we'll produce beautiful and useful pagination bar
14:36 pierrick always looking the same on every page
14:37 pierrick owen, if you have nothing else to ask, I would like to go to bed (feeling very tired today)
14:38 owen I have nothing to add other than "Good night!"
14:38 pierrick thank you, read you tomorrow :-)
15:12 kados owen: noticed your recent commit to opac-authorities...
15:12 owen Everything look okay?
15:13 kados yea, it's just that I had been working on a version too :-)
15:13 kados and cvs can't resolve the conflicts
15:13 kados only difference that I can see
15:13 owen All I did was copy over some updates from the css templates
15:13 kados is that mine includes links if there are existing bibs
15:13 kados I'll just manually add that change
15:13 kados since it's kinda nice
15:14 kados also, the dictionary searches don't seem to be working for me
15:14 kados they return results, but i can't close and insert the values into the search box
15:14 kados have you had similar probs?
15:26 kados yea, me too
15:31 kados owen: so remember you were asking how authorities would change things?
15:31 kados owen: http://smfplopac.liblime.com
15:31 kados owen: should give you a basic idea
15:31 kados owen: for instance, when you pull up the detail page
15:31 kados owen: the subjects are links to authorities searches if they exist in the auth database
15:32 kados owen: otherwise they are regular subject searches as per usual
15:32 kados owen: the same could be done with authors and uniform titles
15:32 kados owen: really, with any link
15:33 owen So they pay for authority records?
15:33 kados well, in the case of sfmpl, they have _very_ minimal records
15:33 kados _just_ the headings
15:33 kados minimal authority records I mean
15:33 kados because their old dynix system could only save the headings
15:33 kados and not the full records
15:34 kados but yes, they did pay for the auth records at some point
15:34 owen Do you have an example of a record which links to an authority subject search?
15:34 kados most of them should
15:34 kados http://smfplopac.liblime.com/c[…]tail.pl?bib=38273
15:35 kados is one that does for sure
15:35 owen Does it indicate an authority recrd search if the link shows 'marclist=6509'
15:35 kados yes
15:36 kados that means 'search tag 650 subfield $9 (the linking field) for the authid=<whatever that id is>
15:36 owen It's too bad the results list says "Results for Search: 6509 = 2576582"
15:36 kados in otherwords, it does a search for every other bib with the same authid
15:36 kados yea, we could easily fix that
15:36 kados I'm sure SMFPL will want that changed
15:37 kados the whole searchdesc var needs to be changed in fact
15:37 kados it's a poor design IMO
15:38 owen It has never worked well out of the box for us
15:39 kados there's so much we could do
15:39 kados but so little time
15:40 kados yea, I'll look at that just as soon as I've finished merging our changes to opac auth display
15:40 owen It works fine in the css templates... the javascript is identical...
15:42 kados huh ... the markup must have changed
15:42 kados are we still valid xhtml?
15:43 kados w00t, Tumer just made his first commits
15:43 kados owen: thanks for walking him through that!
15:44 owen No problem. Is the Wiki open for editing?
15:45 kados the new wiki?
15:45 kados ahhh ... better wait
15:45 kados I need to confirm with pierrick that we've got the latest snapshot
15:46 kados but you could email your stuff to Stephen, that'd be a useful kohadocs addition
15:48 chris morning
15:48 owen Okay, I'll do that too after I get a chance to clean it up
15:48 owen Hi chris
15:49 kados morning chris
15:49 kados chris: Tumer did his first commit today :-)
15:51 chris woo hoo
15:51 kados owen: so the error I'm getting is 'opener.document.f.value has no properties'
15:53 owen But everything I look at in the two versions of the templates matches
15:53 kados in the js we have:
15:53 kados opener.document.f.value[myindex].value!=""
15:53 kados what is value[myindex] in the markup?
15:55 kados <!-- TMPL_VAR NAME="value" ESCAPE="URL" --> maybe?
15:55 kados arrg, not that
15:57 kados huh
16:00 kados hmmm
16:00 kados owen: 'opener' referes to the previous window
16:00 kados owen: so maybe our problem is in the opac-searchresults.tmpl markup, not the opac-dictionary.tmpl markup
16:01 kados and notice it says 'opener.document.f.value' ... it's not getting the id for some reason
16:02 kados owen: shoudl this be a href?<a  href="javascript:SelectEntry('Lewis%2C%20C.%20S​.%20%28Clive%20Staples%29%2C%201898-1963',2,0)" class="button catalogue">Select</a> |
16:03 kados shouldn't it be something like <a href="" onclick=javasript...
16:03 kados ?
16:03 owen It's not relevant to this issue, but yes.
16:04 owen The guidelines I've seen say that you should always include a relevant href value to allow for non-javascript browsers. That doesn't really apply here, of course, because it's a pop-up window.
16:04 owen The alternative would be to use href="#" onclick="SelectEntry..."
16:06 kados yea, that's not what's causing this prob
16:06 kados well I'm burned out on it
16:07 kados let me know if you find anything
16:07 kados otherwise I'll work on it some more tomorrow
16:07 owen I'll keep poking at it.
16:08 owen I've encountered this problem before... not that it helps.
22:19 kados anyone have ideas for a poll related to Koha or open source in libraries for LibLime's first Newsletter?
22:19 kados I was thinking maybe a question about the Koha foundation ...
22:21 kados russ: any ideas?
22:22 russ hmm
22:22 russ now i was thinking about something along these lines
22:22 russ as i had some q's i wanted to ask
22:22 russ our clients
22:50 kados chris: stephen agreed to do an op-ed column
22:50 kados chris: we're calling it 'the hedge'
22:50 chris heh
22:50 kados chris: and tina just made a nice graphic
22:50 chris hedging your bets
22:50 kados http://newlabs.liblime.com/new[…]pete-with-web-20/
22:50 kados hehe
05:20 pierrick_ hi there
05:21 pierrick does anybody know how to insert a line in a table with a null value for one column, using the binding method?
05:21 pierrick (I've tried with undef value in Perl, but it doesn't seem to work)
05:22 chris how do you mean?
05:22 chris is it a different column every time? or always the same column?
05:22 pierrick insert into aqbookfund (bookfundid,bookfundname,branchcode) values (?,?,?)
05:22 pierrick the branchcode can be null
05:23 chris and you want null, not '' ?
05:23 pierrick if I sth->execute($bookfundid, $bookfundname, undef), it doesn't work
05:23 chris $branchcode='';
05:23 pierrick I want SQL NULL value, not an empty string
05:23 chris right
05:23 chris i dont think that you can
05:23 pierrick :-(
05:24 chris the way id do it
05:24 chris is with an if, and 2 $query
05:24 chris one
05:24 chris insert into aqbookfund (bookfundid,bookfundname) values (?,?)
05:25 chris and one with all three
05:25 pierrick I understand what you mean, but this is not a scalable solution :-/
05:25 chris what about 'NULL'
05:25 chris does that work?
05:25 pierrick imagine ou have 23 columns to fill, and 20 can be NULL, will you write 40 queries?
05:26 chris yep
05:26 pierrick I can insert the string 'NULL', but it is absolutely not what I want. I'd prefer inserting an empty string than 'NULL' string
05:26 chris that definitely inserst the string null
05:26 chris ?
05:27 pierrick yes, it does
05:27 chris right
05:27 chris hmm
05:27 pierrick I've tested sth->execute($bookfundid, $bookfundname, $branchcode || 'NULL')
05:28 chris Undefined values, or "undef", can be used to indicate null values.  However, care must be taken in
05:28 chris       the particular case of trying to use null values to qualify a "SELECT" statemeny
05:28 chris the DBI man page reckons we can somehow
05:29 pierrick OK, that's what I supposed at the beginning (using a Perl undef to insert an SQL NULL value) but it doesn't work :-/
05:29 chris right
05:30 pierrick wait wait wait
05:30 pierrick sth->execute($bookfundid, $bookfundname, $branchcode || undef)
05:30 pierrick works
05:30 pierrick ???
05:31 chris weird
05:31 pierrick I thought $input->param('branchcode') was returning undef if the parameter was not defined!
05:31 chris i think it returns blank
05:32 pierrick no because I would have had the error:  DBD::mysql::st execute failed: called with 2 bind variables when 3 are needed
05:32 chris so my $branchcode=input->param('branchcode') || undef;
05:32 pierrick an empty string would have been a correct value
05:34 pierrick anyway, thank you chris :-)
08:19 kados morning folks
08:19 pierrick morning Johsua
08:20 pierrick s{Johsua}{Joshua}
08:57 pierrick kados, do you know how to have two distinct actions (destinations) in a form, depending on the submit button you click?
08:59 kados pierrick: not off the top of my head, sorry
09:01 pierrick kados, OK, I continue to search a solution
09:32 kados morning owen
09:32 owen Hi
09:34 owen No luck yesterday on debugging the opac-dictionary popup.
09:35 pierrick hi owen, have you ever used a multiselect with Perl as scripting language?
09:35 owen My Perl skills are very limited.
09:36 pierrick OK, :-/

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