IRC log for #koha, 2006-03-15

All times shown according to UTC.

Time S Nick Message
12:41 osmoze hello
12:51 hdl hello
12:57 pierrick Hi
12:57 pierrick do we need to confirm the meeting is tonight ?
12:58 pierrick (I don't know if it occurs every monday)
12:58 paul yep.
12:59 paul (we don't need to confirm as yes, it occurs every monday)
12:59 paul although usually joshua publish sometihng on koha-devel.
12:59 pierrick OK, as I didn't anything from Joshua today...
13:07 kados paul: I'm here now
13:07 kados yes, we should have a meeting tonight though I neglected to post a notice to koha-devel -- sorry about htat
13:07 kados that even
13:08 kados paul: have we confirmed the arrangements for the Koha Conference?
13:08 paul which one ?
13:08 kados paul: well, at least the dates?
13:08 paul (also : hi)
13:08 paul yep, of course.
13:08 kados (heh, hi to you also :-))
13:08 kados ok ... so I will purchase tickets and schedule my time
13:08 paul 2-3 may in Paris, then 8-14 in Marseille
13:09 kados yep
13:09 kados I will probably fly into paris on April 24th or so
13:09 paul i'll take care of a cheap hotel & inform you.
13:09 kados then take the train to Geneva with Chris
13:09 paul do you need to rent a car ?
13:09 kados (he will join me there)
13:09 kados maybe, but IIRC public transportation is quite nice in EU :-0
13:10 kados last time I rented a car it was quite a pain
13:10 paul you need to do US =>Paris => Geneva ? can't do US => Geneva ?
13:10 paul you're right.
13:10 paul i'll try to find an hotel in marseille center. and hdl + me will both have a car to reach the CMI (where the dev meeting will take place)
13:10 kados US=>Geneva = $1,800 ; US=>Paris=>Geneva = $800 (with train included) :-)
13:11 paul gloups !
13:11 kados heh
13:11 paul Paris => Marseille is easy to do by TGV (3 hours for 850km)
13:11 kados ($800 is round-trip also)
13:11 kados cool
13:11 paul but not that cheap.
13:11 kados 's ok ... that's why we have a business right? :-)
13:12 kados so I will spend the last week of April between Geneva and Paris
13:13 paul you should do Geneva => Lyon => Paris.
13:13 kados meeting up with Regula as well in Switzerland (she is quite close to WIPO)
13:13 paul and take 2 days to visit Lyon.
13:13 kados is it in France?
13:13 kados yes I see from google :-)
13:13 paul yep. 150km west from Geneva
13:14 kados looks nice
13:14 paul (it's the 2nd/3rd city in France. The question being : is Marseille 2nd or Lyon ? everybody in Marseille, think Lyon is 3rd :-D )
13:14 kados owen: :-)
13:14 paul hehe...
13:14 kados so I can see the three best cities in one quick trip :-)
13:16 paul right ! but it could be nice to make a small trip to Italy too.
13:17 pierrick I made my studies in Lyon (the second city in France ;-) and it's quite nice
13:18 kados paul: I've spent quite a lot of time in Italy
13:18 kados paul: several months in fact
13:18 kados paul: so less of a priority :-)
13:19 kados paul: in fact, I spent about a week in Paris several years ago
13:19 kados paul: but I was travelling with a bunch of Poles, not local French people who know where everything is :-)
13:20 paul Poles ???
13:20 pierrick see you later, I must go now
13:24 kados paul: folks from Poland
13:24 paul ah, ok
13:24 paul (do you speak polish ? because the librarian from CMI, that will host us in MArseille, is polish)
13:25 kados only about 10 words :-)
13:33 Flash Barcode question...
13:34 kados sure
13:35 Flash I'm getting the following error in the log when I try print barcodes...
13:35 Flash [Mon Mar 13 09:35:02 2006] [error] [client 192.168.31.65] Can't locate
13:35 Flash PDF/API2.pm in @INC (@INC contains: /srv/www/htdocs/apps/koha/intranet/modules
13:36 Flash /usr/lib/perl5/5.8.3/i586-linux-thread-multi /usr/lib/perl5/5.8.3
13:36 Flash /usr/lib/perl5/site_perl/5.​8.3/i586-linux-thread-multi
13:36 Flash /usr/lib/perl5/site_perl/5.8.3 /usr/lib/perl5/site_perl
13:36 Flash /usr/lib/perl5/vendor_perl/5​.8.3/i586-linux-thread-multi
13:36 Flash /usr/lib/perl5/vendor_perl/5.8.3 /usr/lib/perl5/vendor_perl .) at
13:36 Flash /srv/www/htdocs/apps/koha/intranet/m​odules/C4/Barcodes/PrinterConfig.pm line
13:42 kados Flash: yep, you need the older version of PDF::API2 to use the current barcode printing utility
14:12 Flash Where do I get it?
14:20 kados ahhh ... I'm not sure
14:20 kados I think if you do some googling you'll find some notes on exactly which version you'll need
14:20 kados as well as where to get it
14:20 kados however, all of this is mute
14:20 Flash Why?
14:21 kados because in about a week there will be a brand new barcode / spine label printing utility
14:21 Flash I'll wait :)
14:21 kados :-)
14:21 Flash How will I know when it's available?
14:37 kados Flash: we'll announce it on koha-devel
14:39 owen kados: where's that coming from? I didn't hear about it.
14:40 kados owen: one of LibLime's clients is sponsoring it
14:40 kados owen: SMFPL actually
14:40 owen Cool.  Long overdue.
14:40 kados yep
14:41 paul another thing to add to 2.2.6 release notes ?
14:41 paul ;-)
14:41 kados if it makes it in time :-)
14:54 kados thd: you around?
14:54 thd yes kados
14:56 kados thd: in your opinion, are the MARC frameworks you're working on ready?
14:56 kados (ie, could they be used in a production system?)
14:58 thd kados: except for correction of the too many under-used fields appearing in the record editor.
15:00 kados thd: could you make it available to me?
15:00 thd kados: I have only been preparing the uber bibliographic framework, although, others are minor derivatives
15:00 kados right
15:01 kados I'm wondering what the best way to proceed with it is
15:01 thd kados: I could after I get the SQL to parse, which I am doing now
15:01 kados cool
15:01 kados thd: have you tried exporting / importing the bib framework before?
15:02 thd kados: Unless you want it complete with SQL errors from the latest revisions :)
15:02 kados heh
15:02 kados naw, go ahead and test it first :-)
15:04 thd kados: For only the default bibliographic framework you need to dump only 2 tables.  marc_tag_structure and marc_subfield structure.
15:05 kados I think our goal is to have default + trimmed-down itemtype/materialtype frameworks, right?
15:07 thd kados: yes and uncommon elements hidden by JavaScript.
15:08 thd kados: I have not yet added extra columns to support more sophisticated hiding.
15:09 kados paul: are you around?
15:09 kados paul: thd and I have discussed adding a new column to the biblio frameworks to support 'hiding' fields in the editor
15:09 kados paul: what is your opinion?
15:10 thd kados: multiple columns
15:10 kados thd: what do we need asside from a 'hidden' flag?
15:11 thd kados: I have not yet tested hidden in the patron view by leaving the label blank.
15:12 thd kados: just now it is rather all or nothing.
15:12 kados thd: I'm not sure it will even exist if you leave both intranet and opac labels blank
15:12 kados thd: could you try that real quick?
15:13 kados thd: that might be a 'quick and dirty' solution
15:13 thd kados: intranet has to be there to create and edit the data.
15:13 kados right ... too bad
15:14 kados wonder if we could label it 'hidden'
15:14 kados then set a flag with regex
15:14 kados (it's a lousy solution though :/)
15:15 thd kados: there is a hidden but that hides it completely from editor and and patron
15:15 kados hmmm
15:16 thd kados: the existing hidden option is for unseen automatic management of the subfield.
15:16 kados yep
15:16 kados it's poorly named :/
15:16 thd kados: I should be more awake but I have been working all night :)
15:16 kados heh
15:17 thd kados: we need a column for degree of hiding in the editor.
15:18 thd kados: a column for degree of use in OCLC.
15:18 thd kados: a column for degree of use in RLIN.
15:19 kados OCLC calls it 'degree of use'?
15:19 thd kados: I just made up that term to describe ...
15:19 kados what are you trying to represent?
15:19 kados I don't understand why we would need three different columns
15:21 thd the variations between must have a field open and ready for filling to hide as far away as you can unless it has data already.
15:21 thd kados: non-RLIN users would care little about some types of RLIN fields
15:22 thd kados: similarly with OCLC but those are fewer.
15:23 kados hmmm
15:23 kados that seems like it's a bit too advanced for this editor
15:24 thd kados: and then there is the standard need irrespective of union catalogue membership.
15:24 thd kados: extra columns do not break anything but I have not added them yet.
15:24 kados I have no idea what that last sentence means :-)
15:25 thd kados: which last sentence
15:25 kados what is the 'standard need'?
15:26 thd kados: hiding or collapsing uncommonly used fields from the editor yet making them accessible.
15:26 kados ahh
15:28 thd kados: that should not be difficult, however, some ARLIN, and OCLC fields need not even be accessible in the to non-users of those systems unless populated with data.
15:30 thd kados: I was supposing a corollary system preference indicating membership of OCLC, RLIN, WLIN (now part of OCLC), etc.
15:31 thd kados: record editor reads system preference and field importance to the respective system and treats it appropriately
15:31 kados that's a nice idea
15:32 kados but I think a simple 'visibility' flag will suffice for this release
15:32 kados I still haven't had a chance to build in subfield repeatability and reordering
15:33 kados and i think we still need to work more on the plugins
15:33 kados as they are incomplete
15:34 thd kados: we have already visibility and editability or invisibility and you could not edit it if you wanted to do so.
15:35 kados however, I assume that if a field is 'hidden' it's value is discarded
15:35 kados thd: have you tested this theory?
15:35 thd kados: there are also extra fields for Canadian users creating bilingual records.
15:36 thd kados: the actual existing 'hidden' column is meant for plugin management and I have not tested it.
15:37 thd kados: hidden is meant for automated filling with no human intervention.
15:38 kados test even
15:39 thd kados: there is also -1 tab setting which I do not like because it requires greater work to set the tab correctly but that does not loose the data to the extent of my experience.
15:40 kados I just tried 'hidden' on 008
15:40 kados as far as i can tell it did nothing
15:40 kados it still shows up and I can still edit it manually or with the plugin
15:40 thd kados: did nothing meaning that you can see the data even when it is hidden?
15:41 kados yep, at least from the addbiblio utility
15:42 thd kados: maybe hidden shows up if the plugin is there try deselecting the plugin.
15:42 thd kados: set it to hidden with only manual editing.
15:43 kados still shows up
15:43 kados so it seems to do nothing :-)
15:43 kados thd: try on your machine?
15:44 kados thd: here's another idea
15:44 thd kados: fixed fields in the '@'subfield or subfields in fields below 010 may behave differently than others.
15:44 kados thd: what about putting 'rare' subfields in a '11' tab
15:45 thd kados: that breaks orderliness, which I was trying to maintain.
15:45 kados thd: lets go that route
15:46 kados thd: put the rare ones in tab 11
15:46 kados thd: it's ok that it breaks orderliness for now
15:46 thd kados: curiously tab is TINYINT(1)
15:46 kados thd: I can't think of another way to do it in the current scheme
15:46 kados huh
15:47 kados well that gives you 256 possible values :-)
15:47 kados -128 to 128
15:47 kados (is -1 currently used for anything?)
15:47 thd kados: really, oh :)
15:49 thd kados: -1 is hidden in the former the preexisting setting for much of MARC that had been there formerly before I changed it all and setting it to the correct tab :)
15:50 thd my spell check ate the usage for that sentence :)
15:51 kados heh
15:52 kados if a subfield is in tab -1, is the field value saved?
15:53 thd yes but I changed all those to the correct tab :)
15:53 thd kados: there are abut 3,500 subfields now :0
15:53 kados yowsa
15:54 thd kados: MARC thought of everything and places to put it all more than once :)
15:54 kados yea
15:55 kados we need to ask paul at the meeting what the purpose of the 'hidden' option is
15:55 thd it is a beauty to behold except for the ARLIN mnemonics where I lacked enough of the documentation to record the proper field name
15:56 kados right
15:56 thd kados: hidden is plainly described as hidden to be managed automatically by a plugin.
15:57 kados yea, but as far as i can tell, it does nothing
15:57 kados if that's true, perhaps we can use it to do what we want :-)
15:58 thd kados: The code for the patron view at least clearly checks for hidden.
16:03 kados I think we've still got an hour before the meeting right?
16:03 kados 20:00 GMT is in one hour, right?
16:03 pierrick yes, I'm just testing my mother internet connection :-)
16:04 kados pierrick: ok :-)
16:04 kados I'm gonna head out for a bit of lunch ... be back in a bit
16:04 pierrick having diner right now... see you in one hour
16:13 thd kados: hidden seems to do nothing for the record editor even in the default templates.
16:40 kados T-MINUS 20 MINUTES TO KOHA MEETING
16:46 kados thd: I'm back now if you need to chat before the mtg
16:47 thd kados: I was disconnected just now
16:47 kados thd: nothing happened :-)
16:48 thd :)
16:50 thd kados: hidden seems to do nothing for the record editor even in the default templates.
16:50 kados yea, saw that
16:50 kados we'll have to ask paul if we can appropriate it :-)
16:52 thd kados: It certainly is used in the OPAC.  Maybe you broke it for the record editor when you fixed other things :)
16:54 thd kados: I have not been able to test tab 11 because I still have some minor SQL warnings to correct before trying that value not accessible from the templates but I suspect 11 may be trouble without changing the code which would more easily be changed with an additional column.
16:56 kados right ... we'll ask paul about that too
16:56 kados it's possible I broke it :-)
16:56 kados T-MINUS 5 MINUTES TO KOHA MEETING
16:56 kados AGENDA:
16:56 kados http://www.saas.nsw.edu.au/koh[…]ndAndNotes06mar13
16:57 kados (quite barebones)
16:57 kados this should be a quick meeting in fact
16:59 thd kados: have you put off working on authorities for now?
16:59 kados thd: not necessarily
17:00 kados thd: but I'd like to finalize our plans for the MARC editor first
17:00 kados thd: I'd rather have full MARC support and partial authorities support
17:00 thd kados: of course
17:00 kados thd: than partial both :-)
17:00 thd kados: )
17:01 kados thd: authorities work will probably pick back up again as soon as I hear back from SMFPL who will be the first library i manage to use them
17:01 kados I'm still waiting to get their data...
17:01 kados AGENDA:
17:01 kados http://www.saas.nsw.edu.au/koh[…]ndAndNotes06mar13
17:01 kados if anyone has anything to add, please do
17:01 thd the editor is much more important for most of your potential customers at this point
17:01 kados thd: yep
17:02 kados ok ... it's 20:00 GMT ... time for our weekly Meeting. The Agenda is at http://www.saas.nsw.edu.au/koh[…]ndAndNotes06mar13
17:03 kados lets start with a roll call
17:03 kados who's here?
17:03 thd here
17:03 pierrick here
17:03 kados hopefully paul will join us
17:04 thd all that javaScript really slowing my machine :)
17:04 shedges here
17:04 kados chris may not make it, or will be late
17:04 kados we can start with News and Questions
17:05 kados anyone have any? :-)
17:05 kados hey russ
17:05 russ hi everyone
17:05 pierrick hi russ
17:06 kados well I don't have any news
17:06 pierrick I have one general question
17:06 kados i do have some questions for paul but he's not around :-)
17:06 kados pierrick: sure
17:07 pierrick How do you manage new features ? I mean, do you write some specifications (functional and technical) somewhere like the wiki ?
17:07 russ that is a fantastic question
17:07 kados yep
17:07 pierrick russ: thanks :-)
17:08 pierrick I do on another project and it's a good practice IMO
17:08 kados so typically this has been handled by each koha support company
17:08 kados ie, katipo has a client that needs X, they handle the specs with the client, etc.
17:08 kados then the code is committed to the main project
17:08 kados after it's been approved by the client, etc.
17:09 pierrick When is it approved or rejected by other members of the dev team ?
17:09 russ chris sends his apologies
17:09 kados pierrick: so far it's been a small enough community that we haven't really needed formal approval
17:10 kados but I suspect we're going to want a more formal process as we grow
17:10 kados ie, code should meet our minimum standards for quality, style, etc.
17:10 thd pierrick: that is the answer for actual implementation.
17:11 kados pierrick: if you take 2.2 as an example
17:11 kados pierrick: all the decisions were basically up to the release manager
17:11 kados pierrick: paul approved / discarded the code for the release
17:12 kados pierrick: (though in practice, he didn't really discard anything)
17:12 kados pierrick: you've had quite a bit of experience working with similarly structured projects, do you bring any advice with you on different ways to handle this issue?
17:13 thd pierrick: this past weekend I was looking at my work just short of a detailed publicity and technical feature list from October that I had intended also to classify potential features..
17:13 kados russ: feel free to pitch in as well :-)
17:13 russ you were doing so well :-)
17:13 kados heh
17:13 pierrick My experience is not on business project but on public project (bigger than Koha in number of users)
17:14 kados right
17:14 pierrick I'm leading a project with a team of 8 developpers, and twice more really implicated users
17:15 pierrick In the wiki, we have a section to write specification for features that could be implemented
17:15 pierrick Once the spec written, it is discussed on the forum
17:15 kados pierrick: are those 8 developers working full-time developing the project?
17:15 thd pierrick: I imagine that some large projects must have layers of code approval to manage contributions that are absent from small projects like Koha.
17:15 pierrick no no, not full time
17:16 kados pierrick: it seems about the same size as Koha then, eh?
17:16 pierrick less code lines, but much more users
17:16 pierrick (around 200 downloads per day)
17:16 kados gotcha, so approx same number of developers but much more users?
17:17 pierrick yes
17:17 kados many more users I should say :-)
17:17 kados ok
17:17 pierrick as I said, it's not a business project
17:17 kados right
17:18 thd everyone should have a personal ILS, hat is a feature for feature development :)
17:18 kados well for us, I guess the assumption is that almost all the features are being sponsored by a real library
17:18 pierrick I think the system of wiki specification + discussion on forum lead to a good reflexion on feature before starting to code
17:18 kados and therefore, we assume that the specification and implementation will be up to par
17:18 kados ie, the library needs to be able to use the feature, otherwise all of us would be out of business :-)
17:19 pierrick yes, I understand ;-)
17:19 kados the trouble with getting approval from the whole community before coding is
17:19 kados that library contracts are so hard to get in the first place
17:19 kados and libraries are such penny-pinchers
17:20 pierrick the approval from the whole community might be too restrictive, but opening a reflexion with the team before implementing can save time sometimes
17:20 kados then, when you finally get the contract, they want the feature _yesterday_
17:20 kados :-)
17:20 kados pierrick: yes, we've done this on some projects, like the new serials stuff that katipo's working on
17:20 thd approval of anything that does not completely break the software for other users is understood
17:21 pierrick when you implement a feature for a customer, do you open a Koha demo with the new feature before installing on their server ?
17:21 russ pierrick : i agree with on the think before you do concept, the challenge is how to adopt that within the community, without making it too restrictive
17:21 kados pierrick: yes, I do
17:22 russ pierrick: yes, we have development versions of koha for all our clients on our own boxes
17:22 kados also, one problem we've faced thusfar is that several times we have features developed that break previous features
17:22 kados I think budget-based acquisitions has been broken no less than 3 times :-)
17:22 russ oh yes
17:22 kados and it's far from the only example
17:23 pierrick I think it's normal when a  software get bigger
17:23 thd kados: the problem there is that features not widely used or sufficiently well commented tend to be trodden upon by other coders.
17:24 kados so one way to approach this issue is to have a non-profit org
17:24 chris back
17:24 kados hey chris
17:24 kados just in time
17:25 thd examples are fairly steady if one does not think so much about major modules such as acquisitions
17:25 kados the non-profit's 'governing body' would be responsible for 'protecting' the feature development in the core ILS
17:26 russ thd : not sure i understand what you said there - part of the problem has been that major and minor features get trodden on
17:26 pierrick thank you for answering my question :-) When my company will ask me first new features, I'll prepare a paper and propose it on koha-devel, just to have a feedback (like "Oh no, don't do that this way" or "great, go ahead")
17:26 kados pierrick: sounds great!
17:26 chris that would be fantastic
17:26 russ excellent
17:27 chris the feedback might be
17:27 thd the best I have seen is someone erasing the former Amazon web services code not realising what it it did.
17:27 kados thd++ :-)
17:27 chris "what a great idea, but remember not to break such and such :-)"
17:27 kados chris: exactly
17:27 kados ok ... so any other news or questions?
17:27 owen What about: "If you do that, don't forget to do this too..."
17:27 kados owen++
17:28 kados "while you're in there ..."
17:28 chris that too :-)
17:28 thd more comments in the code would definitely help prevent others from breaking your work
17:28 chris certainly
17:28 chris more comments ++
17:28 pierrick chris: I understand very well, I often make this kind of feedback to new developers on my project :-)
17:28 chris any comments at all ++++++
17:28 chris :-)
17:28 kados hehe
17:28 kados ok ... I'm gonna move on here
17:29 kados perl-ZOOM Integration
17:29 kados chris: can you give us a status update?
17:29 thd I have very few commits but they are completely commented
17:29 chris i havent done much more on this since last meeting, i extended the searches in SearchMarc a bit more, such that you can search on all that you could in 2.2
17:29 kados cool
17:30 chris by default .. i havent figured out how to handle the query builder at the bottom yet
17:30 kados I remember something about deletions too ... where are we at with that?
17:30 chris that lets you search any marc tag
17:30 kados right
17:30 kados in fact, I'm not 100% sure we _can_ emulate that
17:30 kados as IIRC it does searches across subfields
17:30 chris yeah
17:31 chris deletions
17:31 chris i havent managed to get this to work yet
17:31 thd kados: is there a problem searching across subfields in Zebra?
17:32 chris deletions is the last blocker
17:32 kados right
17:32 kados modify works IIRC
17:32 chris yep
17:32 chris i can modify and blank a record out
17:32 chris but i dont really want an index full of blank records :-)
17:33 kados pierrick: related question: have you attempted a full utf-8 Koha install using the HEAD database?
17:33 chris add/modify both work in koha using zebra .. and searches work, but deletions im stuck on
17:33 kados right, I can try to help with that later today
17:33 kados the Biblio.pm routines probably need some tweaking
17:33 chris yes
17:34 chris there will be tweaking/tidying
17:34 kados yep
17:34 pierrick kados: no, I'm working on 2.2 with Stephen documentation to understand all what Koha does
17:34 chris more tweaking in SearchMarc.pm too
17:34 chris good idea pierrick
17:34 pierrick kados: I've worked a bit on UTF-8 to check if it was working or not
17:35 pierrick kados: on some pure MySQL examples, it works
17:35 kados pierrick: yea, I saw your email, that's really good news
17:35 kados pierrick: (even with table definition as utf-8 IIRC, right?)
17:36 pierrick kados: my screenshot shows it works in Koha
17:36 kados so our tentitive conclusion is that paul's difficulties were cause by non-UTF-8 data in his database
17:36 kados I'll try to get a test going with the new db definitions this week
17:36 thd pierrick: did you see the font funnies?
17:36 kados so we can confirm that theory
17:36 kados our final item is the Koha Con
17:37 pierrick http://le-gall.net/pierrick/im[…]koha-2.2-utf8.png
17:37 kados but paul's not here to tell us what we want to hear :-)
17:37 pierrick thd: what font funnies ?
17:37 kados I'll bug him tomorrow morning about making a formal announcement on the koha and koha-devel lists
17:38 kados we really can't wait any longer before we go public -- otherwise, noone will be able to attend :-)
17:38 kados anything else anyone needs to discuss?
17:38 thd pierrrick:  some fonts do not display the characters correctly in UTF-8 and need to be corrected in css.
17:39 russ do we need to consider a change to the meeting time?
17:39 pierrick thd: normal I think, font can represent a charset, not always unicode (very few fonts can represent unicode)
17:39 russ we haven't discussed some important matters lately becuase paul has been unable to attend
17:40 pierrick 21h40
17:40 pierrick GMT+1
17:40 russ cheers
17:41 pierrick paul's presence must depend a lot on kids sleep
17:41 chris what time is it in the us joshua?
17:41 kados it's 3:41
17:41 chris ahh that makes sense pierrick
17:41 kados well ... in Ohio :-)
17:42 kados I'd be happy to change the time
17:42 kados the other option where we could get both NZ and France while they are both still awake is going to be 12 hours from now
17:43 chris and thats 3am
17:43 kados for me yes :-)
17:43 chris for you
17:43 pierrick kados: don't you ever sleep ?
17:43 russ ack - that's no good for the us
17:43 kados which I'm ok with
17:44 chris i think this is probably the best we can do
17:44 owen Can you tell kados is a bachelor?
17:44 chris what we might need to do, is write up the minutes and get paul to comment
17:44 kados heh
17:44 kados good idea
17:44 kados I'll do that now in fact
17:44 chris thanks
17:46 pierrick can I ask a question ?
17:46 chris you just did :-)
17:46 chris sure pierrick
17:46 kados pierrick: sure
17:46 thd :)
17:47 pierrick updating Joshua's document "Updating_koha", it talks about Arch... shouldn't it be deleted ?
17:47 chris ah yes
17:47 kados yep, definitely
17:47 pierrick is there any a plan to migrate from CVS to ... ?
17:47 chris we did use arch for a little while as an experiment
17:47 thd only updated
17:47 pierrick was it a good experiment ?
17:47 chris when/if savannah support subversion
17:47 kados if savannah supports svn in the future
17:48 chris id like to move to that
17:48 kados yep
17:48 kados me too
17:48 pierrick (me too)
17:48 chris pierrick: i couldnt get my head around arch
17:48 thd savannah does have arch
17:48 pierrick I've heard Arch was very hard to use
17:49 chris i found it so
17:49 thd give people power and they complain about the speed :)
17:49 chris :-)
17:49 pierrick I don't know when Savannah will support Subversion... doesn't seem to be a priority
17:50 chris yeah, arch is a gnu tool, so it makes sense they would concentrate on that
17:50 thd they have actually been working heavily to support subversion
17:50 chris ohh thats good news thd
17:51 pierrick Gna, the french Savannah (using the same software) supports Subversion and it works great
17:51 thd they hired some people a few months ago but subversion seems to have been giving some problems
17:51 pierrick I'm sure Savannah will make Subversion work soon
17:51 chris cool
17:52 pierrick I just hope they will ask for help to Gna admins :-)
17:53 thd I think it is working but the was some difficulty a month ago and they have gradations of reliability where savannah is not equal to cvs on Savannah yet.
17:54 pierrick Can I ask another question ?
17:54 thd s/savannah/subversion/
17:54 chris always pierrick
17:55 pierrick How do you plan the roadmap ? do you define a set a features you want or do you define a date ?
17:55 pierrick I mean, for the new branches
17:55 pierrick 3.0 for instance
17:55 kados that's a question for me :-)
17:56 russ to date, the release manager has coordinated that
17:56 kados so far I have defined a set of features
17:57 kados pierrick: if you would like to work on a new system for branches that'd be fine with me
17:58 thd features have tended to be driven by user funding which too often is merely more and better of the same.
17:58 kados the release date for 3.0 has gotten pushed back a few times ...
17:58 kados I originally thought we'd be done before the end of last year
17:58 kados but by the time we got funding for the perl-zoom stuff
17:59 kados (which took forever)
17:59 pierrick kados: if you defined a set of features, you can't define a data... or the quality will be poor
17:59 kados yep
17:59 kados that's my approach currently
17:59 kados I want 3.0 to be a very stable release
17:59 kados and I'm prepared to push the date back until it's ready
17:59 pierrick is 3.0 features frozen ?
17:59 thd Users do not always have the foresight to commit to features kados might like to direct if he choose independently of market considerations.
18:00 kados pierrick: no
18:01 thd as long as HEAD is open for commits there is space
18:01 kados pierrick: really, when we talk about new features, we're talking about you, paul, hdl, chris, and me :-)
18:01 kados pierrick: it's a small group and we pretty much know what each other is working on and what we want
18:01 pierrick kados: don't you fear 3.0 will never be ready ? :-/
18:01 kados pierrick: I'm sure paul fears it :-)
18:01 chris :-)
18:02 pierrick (I think of project like Debian)
18:02 chris im sure at some point in the nearish future
18:02 chris we will branch a rel_3_0
18:02 kados yep
18:02 chris and call a feature freeze
18:02 kados but we dont' want to be premature with that
18:02 kados because then we have to deal with merging bugfixes
18:02 chris and then its just finishing whtas been started and fixing
18:03 chris id love to do it in subversion
18:03 chris where merging is so much easier
18:03 kados yea, that would be nice
18:03 chris lets branch when we get subversion support :-)
18:03 kados heh
18:04 kados does anyone have any specific questions for paul?
18:04 pierrick dangerous idea ;-)
18:04 kados asside from details about the Koha Con?
18:04 kados russ: ?
18:04 thd pierrick: when Koha release cycle  has the administrative complexity and overhead of the Debian project the Boxer will look at the farmers and look at the pigs and not be able to tell the difference.
18:04 pierrick kados: about Koha Con... oups, no question on this
18:04 russ kados: kohacon details are what i am after
18:05 kados righto
18:06 pierrick thd: a release cycle is "date based" or "feature based". Anywhat the size of the project. If you always wait for a feature while another buggy features are added, you never release :-/
18:06 thd does not pierrick know at least half of the details?
18:06 pierrick I know things, but not all
18:06 pierrick what questions do you have ?
18:07 kados pierrick: there are some features I'd _like_ to have in 3.0
18:07 kados pierrick: should I tell you what they are?
18:07 pierrick kados: I understand very well, "feature base" or "date based" is a matter of choice. There is no bad model
18:07 kados pierrick: maybe send an email to koha-devel explaining briefly some of my ideas?
18:07 thd pierrick: I am still working on fixing bugs from 3 years ago before I was involved :)
18:08 pierrick kados: a mail on koha-devel with the list of features for 3.0 would be very interesting to me :-)
18:08 kados well I already sent off the meeting minutes so I guess our meeting is over :-)
18:08 kados pierrick: ok, I'll put it on my list for this week
18:09 russ nice way to constrain a meeting kados :-)
18:09 chris :)
18:09 chris i think the best thing we can do
18:09 chris is communicate lots
18:09 kados russ: :-)
18:09 thd russ: I was waiting for you to ask your insightful questions about Koha con
18:09 russ yep - the devel list is grat for that
18:10 chris so we know what people are working on, and why, etc
18:10 russ thd: no point asking if paul isnt here
18:10 chris i think we have been getting good at using each others strengths, eg thd, kados and I worked well together on the marc editor
18:11 chris russ: pierrick might be able to answer some he said
18:11 kados yep, that's been very successful
18:11 russ sorry i missed that
18:12 russ pierrick: do you know if anyone has been confirmed to talk
18:12 russ and when, on what?
18:12 thd rruss: please ask pierrick now and you can ask paul later also.
18:13 kados heh, some bastard changed the password for the 'circ' user on LibLime's demo
18:13 pierrick russ: the only thing we were talking about with Paul was Joshua talk about 3.0 new features
18:14 pierrick on the second day with Koha french users
18:14 russ right
18:15 thd pierrick: is someone speaking about the virtues of FOSS or are you expecting only the converted to attend?
18:15 pierrick On the first day, Paul wanted an important french talker to come, but I don't know if he found the good person
18:15 russ right
18:16 pierrick thd: I don't know who could speak, Tristan Nitot (Mozilla Europe president) would be a good speaker, but he's ver busy
18:17 pierrick INEO and Paul are expecting many librarians, not especially FOSS addicted librarians :-)
18:17 thd pierrick: I would imagine that FSF could find someone if you did not have someone already.
18:17 chris russ and I would be willing to do a talk
18:18 chris i could do the talk (with russ's) help that I gave at linux conf australia earlier this year
18:18 pierrick chris: great :-)
18:18 chris which is mostly the history of Koha, why/how it exists etc
18:19 russ that talk gives some good background as to why koha is open source - why it made sense for a little library in nz
18:19 chris :)
18:19 pierrick On the first day, a demo of Koha is planned. IMO it is the hardest talk, it needs to be interesting and not asleeping
18:20 chris it covers the benefits of FOSS, etc
18:20 chris yes, thats the trick pierrick
18:20 russ pierrick : agreed it is a challenge
18:20 thd it is a comparative for some senses not an absolute :)
18:21 russ ok i think that answers my question - what i really wanted to know is how much has been decided upon and confirmed
18:21 russ and i guess the answer to that is "we have ideas and an outline but not much detail yet"
18:21 pierrick maybe more than I know has been confirmed
18:22 russ pierrick: sure, i think that is why we should ask paul
18:22 pierrick russ: yep, Paul is obviously the only who knows
18:22 thd pierrick, I had thought that you had an attractive brochure already to distribute to Inneo customers about Koha Con.
18:23 pierrick thd: hum... not yet
18:24 pierrick the contacted libraries are not INEO customers yet... INEO contacted libraries with the mailing-lists that all libraries are subscribed to
18:25 thd pierrick: I presumed that one of your principal goals was to develop more interest in Koha form busy people who would need to schedule well in advance :)
18:25 pierrick I don't know if the invitation was already sent and if some libraries have reserved a seat
18:26 thd pierrick: So the invitations were at least prepared for sending.
18:26 pierrick thd: my principal goals is not so well defined, I've been hired to work on Koha, not because I already knew Koha
18:27 pierrick thd: yes, invitation is prepared, I'm sure of it
18:27 thd pierrick: I did not man your personal goals but  rather the goals of the conference.
18:28 pierrick thd: sorry, I misunderstood your sentence...
18:29 pierrick thd: so your right, Kohacon goal is to promote Koha, from busy people but not only
18:30 thd pierrick: for your prospective future customers to commit to attending they need to be able to plan well enough in advance.
18:30 thd yes capture all the people, not just the busy ones
18:31 pierrick thd: ILS upgrades are often big projects running on several months or years
18:32 thd pierick: is there a URL for Kohacon?
18:33 pierrick thd: I don't know
18:38 thd I think hdl mentioned something about a link on koha.fr,org but I do not see anything.  Part of the home page is not appearing for me.
18:39 thd the page appears now but I do not see any prominent reference.
18:40 pierrick URL ?
18:41 thd pierrick: A website promoting Kohacon and referenced in the invitation for more information.
18:42 thd not that koha.fr.org could not serve that purpose but I do not see anything.
18:43 chris we will put information up at koha.org as well
18:43 chris when we have talked to paul
18:43 chris i think we just need to get pauls repsonse thd
18:43 russ /me is in the process of making all those pages now :-)
18:44 russ hence why i am after the details
18:44 thd chris: yes of course
18:44 pierrick I hope Koha Con will be planned by the end of the week, I'll ask Paul if I need to see my boss
18:44 russ cool thank you pierrick
18:45 russ lol
18:45 russ i get that all the time too :-)
18:46 chris cya later pierrick
18:46 thd pierrick: I had assumed that Kohacon had already been planned and I was merely uninformed :)
18:48 thd pierrick_away: see you at another play date
20:11 kados chris: For update, the record option should be set to the full text of the XML record to added, deleted or replaced. Depending on how the server is configured, it may extract the record's unique ID from the text (i.e. from a known element such as the 001 field of a MARCXML record), or it may require the unique ID to passed in explicitly using the recordIdOpaque option.
20:11 kados it may be that we need to pass along the full record when we want to delete
20:12 chris ahh
20:12 chris could be
20:13 chris heres what i have
20:13 chris $rs->sort("yaz", "1=4 <i");
20:13 chris which should sort by title, ignoring case
20:13 chris but i get this error
20:14 chris [request] Sort  ERROR 207 (1)->1
20:14 chris in the zebra log
20:14 kados huh
20:14 chris which leads me to believe there something in the config
20:14 chris wrong
20:14 chris in collection.abs
20:15 chris http://72.14.203.104/search?q=[…]l=en&ct=clnk&cd=4
20:16 kados could you commit your collection.abs so i can have a look?
20:16 chris will do, ill commit SearchMarc.pm too
20:17 chris so you can have a play
20:17 chris there we go
20:17 kados http://indexdata.dk/zebra/doc/[…]ort.tkl#id2528289
20:18 kados Sort
20:18 kados Z39.50 specifies three different types of sort criteria. Of these Zebra supports the attribute specification type in which case the use attribute specifies the "Sort register". Sort registers are created for those fields that are of type "sort" in the default.idx file. The corresponding character mapping file in default.idx specifies the ordinal of each character used in the actual sort.
20:18 kados Z39.50 allows the client to specify sorting on one or more input result sets and one output result set. Zebra supports sorting on one result set only which may or may not be the same as the output result set.
20:18 chris hmm
20:18 kados not sure if that's relevant, but it might be
20:18 chris from the ZOOM man
20:19 chris At present, the only supported sort-specification type is "yaz".  Such a specification con-
20:19 chris       sists of a space-separated sequence of keys, each of which itself consists of two space-sepa-
20:19 chris       rated words (so that the total number of words in the sort-specification is even)
20:19 kados what the heck?
20:20 chris if ($rs->sort("yaz", "1=4 >i 1=21 >s") < 0) {
20:20 chris            die "sort failed";
20:20 chris        }
20:20 chris For example, the sort-specification in the code-fragment above will sort the records in $rs
20:20 chris       case-insensitively in descending order of title, with records having equivalent titles sorted
20:20 chris       case-sensitively in ascending order of subject.  (The BIB-1 access points 4 and 21 represent
20:20 chris       title and subject respectively.)
20:20 chris so it makes sense .. if i could understand how to get it to work :-)
20:20 kados right
20:22 chris googling leads me to believe its erroring because it hasnt indexed the field to be sorted on
20:22 chris but im not sure
20:22 kados I didn't get the sortby code in SearchMarc.pm
20:22 kados or is that not where it is
20:22 chris yep thats where it is
20:22 chris (HEAD)
20:22 kados if ($rs->sort("yaz", "1=4 <i") < 0) {
20:22 kados should'nt that be:
20:23 kados if ($rs->sortby("yaz", "1=4 <i") < 0) {
20:23 kados ?
20:23 chris umm it says sort in the man page
20:23 kados http://search.cpan.org/~mirk/N[…]d#ZOOM%3A%3AQuery
20:23 kados ZOOM::Query says sortby :-)
20:23 chris yeah this is resultset
20:24 kados ahh
20:24 chris ZOOM::ResultSet
20:24 kados well shouldn't you do the sort in the query?
20:24 chris either or
20:24 chris resultset is the cache of all the results
20:25 chris allows you to resort without requerying
20:25 chris ie without researching, you can reorder the results
20:25 kados ahh right
20:25 kados that'd be nice
20:25 chris yeah
20:26 chris and its working
20:26 kados it is?
20:26 chris in the way that its not erroring/dieing
20:26 chris its passing the query to zebra
20:26 chris and zebra is saying error 207
20:27 kados ahh
20:27 kados hmmm
20:27 kados so prolly the .abs file then
20:27 kados sigh
20:28 kados melm 245/?/a            title           !:w,!:p,!:s
20:28 chris yeah i think its to do with !:s
20:28 chris yeah that was me playing around
20:28 kados heh:
20:28 kados http://lists.indexdata.dk/pipe[…]nuary/000971.html
20:29 kados Tumer :-)
20:29 chris ohh so we need elm not melm?
20:30 kados that would be weird
20:30 kados melm:
20:30 kados This directive is specifically for MARC-formatted records, ingested either in the form of MARCXML documents, or in the ISO2709/Z39.2 format using the grs.marcxml input filter. You can specify indexing rules for any subfield, or you can leave off the  $subfield   part and specify default rules for all subfields of the given field (note: default rules should come after any subfield-specific rules in the configuration file). The  attributes   have the same syntax and me
20:31 chris yeah
20:31 kados I wonder if you can mix and match
20:37 chris good question
20:38 chris can you make zebra reindex itself?
20:38 chris i wonder if thats the problem all the old records made before we told it :s ?
20:39 kados hmmm, yea could be
20:39 kados zebraidx -s rings a bell, but I'm not 100% sure
20:39 kados (for reindexing)
20:40 kados btw: I just tried a delete
20:40 kados and I get a 404
20:40 chris oh maybe it worked?
20:40 chris hmm
20:40 chris where did you delete from?
20:40 kados no, the item still exists
20:40 kados http://kohatest.liblime.com/cg[…]?biblionumber=138
20:40 kados could be a template problem
20:40 kados where are you deleting from?
20:41 chris havent tried today but from there
20:41 chris http://koha.koha2.katipo.co.nz[…]Cdetail.pl?bib=11
20:42 chris i dont get a 404, but it doesnt kill the record either :)
20:43 chris it is trying to do it
20:43 chris 12:42:47-14/03 zebrasrv(6) [request] EsRequest  ERROR 224 only XML update supported
20:43 chris but its not happy
20:43 chris (thats the delete)
20:45 kados huh, mine doesn't even try
20:45 kados I just get a 500
20:45 kados (found the bug in the template causing the prob and fixed it in cvs)
20:46 kados heh
20:46 kados [Mon Mar 13 16:46:11 2006] [error] [client 70.106.173.65] HTML::Template->new() : Cannot open included file doc-head-close-addbiblio.inc : file not found. at /usr/share/perl5/HTML/Template.pm line 2178., referer: http://kohatest.liblime.com/cg[…]?biblionumber=138
20:46 kados damn it
20:46 chris that would do it :-)
20:50 kados ok ...
20:50 kados so now it doesn't 500
20:50 kados I delete an item and I get:
20:50 kados 16:51:17-13/03 zebrasrv(9) [request] EsRequest  ERROR 224 only XML update supported
20:51 chris yep
20:51 chris me too
20:52 kados I've got to get something to eat
20:52 kados bbiab
20:52 chris cya
20:53 kados one thing to try
20:53 kados is passing it the whole record
20:53 kados (though that just seems silly)
20:54 kados it might be the only way
20:54 chris yeah
20:54 chris ok so you cant mix melm and elm
20:55 chris 12:54:43-14/03 zebrasrv(2) [warn] collection.abs:35: Bad level increase
22:24 rach hi ya
22:28 kados heya rach
22:28 kados ok, so I'm starting to see how this works
22:28 kados there's a default.idx file in the 'tab' dir
22:29 chris yep
22:29 chris it looks ok
22:29 kados that's where the index types are specified
22:29 kados and yea, it look ok
22:29 chris Idx: [s]bib1:ISBN [7] data XData:"0525247688"
22:29 kados as does the string.chr it refers too
22:29 chris i ran update with -s
22:29 chris it says its creating an sorting index
22:29 chris on isbn
22:30 kados same error for me
22:30 chris yeah
22:30 chris im just playing in yaz-client
22:30 chris to see if i can get it to work
22:30 kados cool
22:32 chris grr
22:32 chris Diagnostic message(s) from database:
22:32 chris    [207] Cannot sort according to sequence -- v2 addinfo ''
22:32 chris same error in yaz-client
22:32 chris there must be something im missing
22:36 kados huh
22:36 kados what's the advantage to indexing with 'melm' rather than just 'elm'?
22:37 chris pass
22:37 chris i dont understand anything in the .abs file to any real extent :-)
22:39 chris melm is specifically for MARC records
22:39 kados I'm gonna reindex with 'elm'
22:39 chris sounds like a plan
22:39 chris in yaz-client
22:39 chris if you do a find
22:39 chris then you can do a
22:39 chris sort 1=4 ai
22:40 chris which should sort on title
22:40 chris to test
22:42 kados 18:43:58-13/03 zebrasrv(1) [warn] collection.abs:34: Bad # of args to elm
22:42 kados 18:43:58-13/03 zebrasrv(1) [warn] collection.abs:35: Bad level increase
22:43 kados changed all the 'melm' to 'elm'
22:43 chris i have a plan
22:45 chris http://prdownloads.sourceforge[…]3.tar.gz?download
22:45 chris lets look at how emilda have theres set up
22:47 kados k
22:47 chris the usmarc.abs one
22:47 chris is all done with elm
22:47 chris not melm
22:50 kados huh
22:51 chris maybe we can copy that, and call that collection.abs :)
22:51 kados heh
23:05 kados btw: switching gears a bit
23:05 kados http://www.suite101.com/article.cfm/paris/109815
23:05 kados tina found that article which recommends renting an apartment in paris :-)
23:11 rach renting an appartment in vienna was a lot cheeper than a hotel
23:17 kados + you get a kitchen :-)
23:18 kados it might be worth it to find an appartment where chris, rach, tina and I can all stay ... that way we could split the cost
23:18 kados s/rach/russ/
23:18 kados unless you're planning to go to France rach :-)
23:22 russ kados: ben has summoned rach :-)
23:43 kados russ: ben?
23:43 kados ahh ... ben :-)
23:43 kados hehe
01:03 thd kados: are you still up?
01:12 kados thd: yep, kinda
01:12 kados thd: what's up?
01:15 thd kados: I could send you an SQL file that would be useful for a new installation as it could be subsequently updated after I am awake enough to finish a few things that I was too tired to focus on today.
01:16 kados thd: I can wait until tomorrow
01:16 kados thd: get some sleep :-)
01:17 thd kados: It would also work on an existing installation but you would need to add whatever MARC fiend/subfields you had mapped to the items table outside the original default.
01:18 thd kados: I was too tired too sleep before.
01:18 kados heh
01:18 kados yea I know how that is :-)
01:19 thd kados: I was productive until early this afternoon and then I became a zombie :)
01:20 kados yea, I've been kinda burned out for the last couple days
01:21 kados tomorrow's another day
01:22 kados and with that, I'm gonna head out
01:22 kados ciao all
01:23 thd kados: the SQL errors were only careless copying of the form from the previous line without changing the field or subfield for the next line.  Just a few minutes to fix those earlier just as I have usually had between manually editing the SQL dump and and then executing it.
01:24 thd kados: good night.
04:59 russ paul you around?
04:59 paul yes
04:59 russ hi there
04:59 paul hello russ
05:00 paul today i'm well, as baby made a complete night
05:00 paul as well as us ;-)
05:00 russ ah that is good
05:01 russ baby is better now?
05:01 paul I've read the summary joshua made about the meeting.
05:01 paul (yes)
05:01 russ cool and cool
05:01 paul I plan to open a wikipage about KohaCon
05:01 paul hello pierrick
05:02 russ cool - i have started work on pages for koha.org
05:02 russ including a space on the front page to highlight it
05:02 paul this wikipage will be for organisational questions.
05:02 russ good idea
05:02 paul will mostly be for our internal use
05:03 russ cool - yep the web page is for promotional use
05:03 russ i think it will be good for those going and even those who wont be
05:04 russ ah i need to make you a login for the cms so you can see the site before it goes up
05:07 russ how are you getting on with the programme for the first day?
05:07 paul I have to work on it a little.
05:07 paul but i'm a little overbooked those days.
05:07 russ i know that feeling :-)
05:08 paul (for example : i've 214 unread commits on koha-cvs head-only. + 47 for rel_2_2. As 2.2 release manager I think it's poor :-( )
05:09 russ oh dear - that sounds like a lot of work :-(
05:10 paul + 2RFP + some accounting questions + +++ ...
05:10 pierrick hi russ and paul
05:10 russ yep to do lists only get longer
05:10 russ hi pierrick
05:26 chris evening
05:29 pierrick chris: :-)
05:29 paul hello chris
05:29 paul are you happy with France - GB last rugby result ?
05:30 chris yes very happy
05:30 paul it seems france scored the most rapid try in the "tournoi des 6 nations".
05:31 chris ireland or france now .... so im happy, as long as england dont win, im always happy in the 6 nations :-)
05:31 paul less than 1mn...
05:31 chris england play boring rugby
05:31 russ hi chris
05:31 chris and if they win, its a win for boredom :-) i much prefer the way france plays :-)
05:32 hdl hello world
05:39 russ hi hdl
05:39 hdl hi russ
05:39 hdl hi paul, chris and pierrick.
05:40 chris hi hdl
05:50 pierrick hi hdl
05:51 osmoze hi too
05:51 pierrick hi osmoze
05:51 chris hi osmoze
06:03 russ night all
06:04 hdl nite russ.
06:04 pierrick good night
06:04 paul good night kiwis
08:30 pierrick osmoze: are u there ?
09:40 osmoze pierrick>  now yes
09:40 paul les parisiens et les autres n'ont pas tout à fait le même horaire pour déjeuner ;-)
09:41 osmoze (les fonctionnaire et les autres ;))
09:41 osmoze +s
09:43 osmoze au passage bonjour Paul
09:43 osmoze hier soir est passé uyne conversation vis a vis de kohacon, des news ?
09:44 paul super !
09:44 paul pas de news particulière. En gros : on bosse, on rame sur l'utf8, ca marche pas trop mal avec zebra
09:52 osmoze hdl here ?
09:52 paul (il peut être pas loin, mais si tu l'attrapes pas cet aprèm, il est absent 2 jours ensuite)
09:53 osmoze je vais faire mes remarques par mails, ce sera plus simple ^^
09:53 pierrick omoze: ma question portait sur ta gestion des branches, mais la lecture de la bonne section dans la doc a répondu de manière générale
09:53 osmoze ok pierrick  :)
09:56 osmoze paul, peut etre que tu peux me renseigner, le nettoie_bdp ( que tu nous avez fais a l epoque) marqué "L |" pour les cds, es ce corriger ?
09:56 osmoze -avez + avais
09:57 paul mais effectivement, hdl serait plus apte à une réponse correcte.
09:57 osmoze ok :) je m en depatouille avec lui ^^
10:01 pierrick logbot, status?
10:23 paul pierrick : logbot est l'automate qui archive le canal sur www.koha.org/irc
10:44 pierrick yes paul ?
10:44 paul tu as quelques minutes ?
10:44 paul (à propos de l'utf8)
10:44 pierrick oui, bien sûr
10:44 paul j'essaye de refaire ton test uft8
10:44 paul ce n'est pas bienconcluant...
10:45 pierrick :-/
10:45 paul mais je ne pense pas être en utf8 sur mon terminal
10:45 paul comment je peux le vérifier ?
10:45 paul locale me donne :
10:45 paul [root@bureau ~]# cd /home/paul/tmp
10:45 paul [root@bureau tmp]# locale
10:45 paul LANG=fr_FR.UTF-8
10:45 paul LC_CTYPE=fr_FR.UTF-8
10:45 paul LC_NUMERIC=fr_FR.UTF-8
10:45 paul LC_TIME=fr_FR.UTF-8
10:45 paul LC_COLLATE=fr_FR.UTF-8
10:45 paul LC_MONETARY=fr_FR.UTF-8
10:45 paul LC_MESSAGES=fr_FR.UTF-8
10:45 paul LC_PAPER=fr_FR.UTF-8
10:45 paul LC_NAME=fr_FR.UTF-8
10:45 osmoze c est quoi l avantage de l utf8 ( depuis que ca en discute beaucoup, j arrive pas a faire le d"istingo)
10:45 paul LC_ADDRESS=fr_FR.UTF-8
10:45 paul LC_TELEPHONE=fr_FR.UTF-8
10:45 paul LC_MEASUREMENT=fr_FR.UTF-8
10:45 paul LC_IDENTIFICATION=fr_FR.UTF-8
10:45 paul LC_ALL=
10:45 paul mais ton test me renvoie :
10:45 paul {1} ������
10:45 pierrick osmoze: je te réponds après :-)
10:46 paul là ou j'avais mis un coeur, un coeur renversé et un symbole musical
10:46 pierrick 3 caractères ?
10:46 paul si je me mets en CTRL-ALT-F1 (donc console pure)
10:46 paul exact
10:46 paul j'ai trois ? blancs sur fond noir.
10:46 paul (accessoirement, je DOIS me logguer en root avec su - pour que ca marche, si je me loggue en "paul", j'ai un locale en fr_FR tout simple
10:46 paul )
10:47 pierrick alors il faut cour-circuiter le copier/coller
10:47 paul ???
10:47 paul j'oubliais :
10:47 paul [root@bureau tmp]# ./versions.pl
10:47 paul DBD::mysql : 3.0002
10:47 paul      Perl : 5.8.7
10:47 paul     MySQL : 4.1.12
10:47 paul       DBI : 1.48
10:47 pierrick il faut enregister la page html sur ton disque et la faire lire à Perl
10:47 paul qué page HTML ?
10:48 pierrick http://www.columbia.edu/kermit/utf8-t1.html
10:48 paul j'ai fait du copier coller dans kate, en me mettant en utf8, et je vois bien dans le fichier ce que j'ai copié
10:48 pierrick déjà, ta version de DBD::mysql est plus récente
10:49 pierrick OK, alors ça doit être plutôt bon... Combien pèse ton fichier ?
10:49 paul le fichier utf8.txt ?
10:49 pierrick oui
10:49 paul 9 octets
10:49 paul 3x3 quoi
10:50 paul j'ai mis ca dedans :
10:50 paul HEAVY BLACK HEART
10:50 paul ROTATED HEAVY BLACK HEART BULLET
10:50 hdl osmoze : Je n'avais pas vu... Dsl
10:50 paul CURVED STEM PARAGRAPH SIGN ORNAMENT
10:51 pierrick si tu fais $(cat utf8.txt), ça affiche quoi ?
10:52 paul -bash: ������,: command not found
10:52 pierrick ton terminal ne doit pas être en utf-8, définitivement...
10:53 paul et comment je le mets en utf8 ?
10:53 paul zat is the kestion...
10:53 pierrick tu utilises quoi comme terminal ?
10:53 paul konsole
10:53 pierrick mais de toute façon, ça n'est pas important, la finalité, c'est d'envoyer de l'UTF-8 dans une page web
10:54 pierrick moi j'utilise gnome-terminal, c'est en UTF-8 par défaut sous Ubutun
10:54 osmoze paul>  set-language-env en tant qu utilisateur, ca marche pas ?
10:54 paul [root@bureau paul]# set-language-env
10:54 paul bash: set-language-env: command not found
10:54 osmoze je ne sais pas sous Mandriva
10:54 pierrick donc je te propose de rediriger la sortie de readfile_insertdb.pl dans un fichier
10:55 paul mandriva powa, sauf pour l'utf8 ;-)
10:55 pierrick modifie un peu mon script
10:55 paul BINGO !
10:56 paul [root@bureau tmp]# ./readfile_insertdb.pl utf8.txt >utf8B.txt
10:56 pierrick dans le dernier foreach, ne garde que print $string->{value};
10:56 paul si j'ouvre le fichier utf8B.txt sous kate, que je demande l'affichage en utf8, ca marche bien.
10:56 pierrick :-)
10:56 paul ce qui m'avance en partie.
10:56 pierrick j'allais te propose de comparer les md5sums
10:56 paul reste maintenant à tout mettre en utf8...
10:57 pierrick mais si visuellement dans Kate c'est bon, alors ça le serait aussi dans un browser
10:57 hdl ce qui est assez difficile, c'est que nos utilisateurs ne seront certainement pas tous en clavier utf-8. Il faudra certainement toujours vérifer et convertir nos variables.
10:57 pierrick mouais... pas évident que mettre son terminal en UTF-8 soit si utile que ça
10:58 pierrick personnellement, je ne comprends pas ce que "clavier utf8" veut dire
10:58 hdl En faisant export, tu as quoi comme variable CHARSET et KBCHARSET.
10:58 paul pierrick : moi non plus ;-)
10:58 hdl pierrick: En fait, cela veut dire que les touches que tu tapes sont directement interprétées en utf-8 et génèrent des codes UTF-8.
10:59 paul declare -x LESSCHARSET="utf-8"
10:59 paul rien pour kbcharset je crois

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