IRC log for #koha, 2006-03-22

All times shown according to UTC.

Time S Nick Message
12:08 hdl kados ?
12:08 hdl I need some information here.
12:09 hdl Which fields do you link to
12:09 hdl itemtype
12:09 hdl bibliosubject
12:09 hdl author (100$a guessed)
12:09 hdl in MARC-21
12:10 hdl itemcalnumber needed too.
12:18 kados thd: some of my clients use both DDC and LCC
12:19 kados thd: so use 'dewey' to may the DDC and 'classification' to map the LCC (I think)
12:21 kados hdl: I've got to run on a service call
12:21 thd kados: I see that would be the simplest way of doing it
12:22 kados hdl: thd can answer your question (as we're trying to decide how to handle this currently)
12:22 kados thd: do you think it's a good way?
12:23 thd kados: it does not conform with what chris had for range search
12:23 kados thd: I'll be back in ten minutes
12:24 thd kados: range search might work with LCC but maybe the letters would need conversion to numbers
12:29 hdl thd : can you answer my question ?
12:30 hdl MARC-21 is quite confusing to me.
12:31 thd hdl: yes I can probably answer
12:31 thd hdl: UNIMARC is more logical
12:32 thd hdl: There is no proper standard MARC field to link to biblioitems.itemtype in either MARC 21 or UNIMARC
12:33 hdl 200$b seems to be the one we generally use
12:33 hdl I saw 942$c for itemtype
12:34 thd hdl: This is because the field that you would want to map for general material designation with is 245 $h in MARC 21 is not used for books.
12:34 thd s/with/which/
12:35 hdl 942$k for itemcallnumber ?
12:36 thd kados: for books Anglo American cataloguing Rules specify that books are assumed when nothing is specified and I am certain that the rule is the same for French bibliographic rules
12:36 thd hdl sorry above
12:37 hdl 700$q for author ? (seems always empty would 100$q be more accurate ?).... .
12:38 thd hdl: Therefore, a separate field is needed which could be filled by a script in MARC 21 from 245 $h and if 245 $h were not present then BOOK would fill the special field.
12:38 hdl 650$a bibliosubject.
12:39 hdl thd: Is such a script designed already ?
12:39 thd hdl: items are currently in a nonstandard 952 by default
12:40 thd hdl: so currently 952 $k by default
12:41 hdl is 942 a standard for items ?
12:41 thd hdl: No there is no such script yet but since paul has told kados the secret of writing cross field scripts and he is currently working on the record editor I expect there will be a script soon
12:42 paul ;-)
12:42 kados I can confirm that :-)
12:43 kados thd: however, currently what I need is just a basic mapping that will work in the time being
12:43 thd hdl: 942 is used for biblioitems not items and is an invention of paul but fortunately does not conflict with any other uses
12:44 thd hdl: 952 is used for items and conflicts with the world's second largest library network RLIN so I intend to move that.
12:45 paul it's a proposal from NPL
12:46 thd paul: Is there anything in Koha that would cause a problem for 95k with a letter instead of 952?
12:46 kados thd: we won't move 952 in rel_2_2 I don't think
12:46 kados thd: as it's completely unnecessary
12:47 kados thd: if we find a client that uses RLIN records we can just re-map the values using our import scraipt
12:47 kados thd: script I mean
12:47 thd kados: It would be very simple with a little thought to write the SQL
13:00 pierrick read you all tonight at the meeting :-)
13:01 pierrick (I have some topics to discuss on)
13:02 paul I leave too.
13:02 paul not sure to be here tonight, but i'll try
13:02 pierrick paul: will you be there tonight ?
13:02 pierrick cross post...
13:02 paul_away you've got the answer ;-)
13:13 thd kados: If you look at 952 as my email had suggested, it can almost be used as if it were 852, 876-8 and I am much more interested in rearranging the subfields according to my recommendation so that the match to 852, 876-8 is more perfect but I left the ones you already had where they were.
13:14 kados thd: checking now
13:15 kados thd: dewey  942  k  Classification base (DDC to decimal or LCC letter class padded after single letter classes with trailing 0
13:15 kados ?
13:16 kados thd: 942K should not be LCC ... it should be DDC
13:16 thd kados: that is how chris intended to use it for range searching which does not really work with LCC
13:16 kados thd: 942J is still marked as obsolete
13:16 kados thd: forget searching by classification in rel_2_2
13:16 thd kados: you do not need 942 j
13:17 kados thd: what I'm woried about is display
13:17 thd kados: it never existed before so just do not fill it
13:17 thd :)
13:17 kados hmmm
13:17 kados NBBC uses that
13:17 kados classification is LCC
13:18 kados dewey is DCC
13:18 kados DDC I mean
13:18 thd kados: NBBC has it mapped to 050
13:18 kados that's fine as long as it will display!
13:18 kados which is the whole point
13:19 kados 050 for classification?
13:21 thd kados: the map that you gave me had 050 $a mapped to items.classification and 050 $b mapped to items.subclass.
13:21 kados thd: I believe I mapped it that way because you told me to some months back :-)
13:21 kados thd: why don't you like that method now?
13:21 thd kados: What I did is allow the framework to be agnostic as to DDC or LCC
13:22 kados thd: but many clients use both DDC and LCC
13:22 kados thd: so the framework has to support both simultanously
13:22 thd kados: you discovered plugins and 942 should be filled with a plugin so the cataloguer never has to touch it
13:23 kados thd: itemtype needs to be manually set
13:23 kados thd: also, you aren't taking into account local classifications
13:23 kados thd: or locations
13:23 kados thd: both of which some libraries use
13:24 thd kados: I may have been a little mistaken about the mapping months ago but you did not seem to have enough time then to see how the plugins worked in French else we may have had a very different conversation at the time
13:25 kados there are three things to keep in mind
13:26 kados wait ... nevermind
13:26 thd kados: oh yes but if the question is joining all the matches in a unified manner at the biblio level and not the item level the classification system does not even matter much
13:27 kados ok ... the problem with classifications and call numbers and locations is that every library does it differently
13:27 kados some use classification and call number interchangably
13:28 kados some have multiple classifications but only one call number
13:28 kados some have call numbers at the record level AND at the item level
13:28 kados etc etc.
13:29 kados there may be a standard way to represent classification / call number / location in MARC21 but I've yet to see a library that uses it
13:29 thd kados: call number is the classification system number plus the item number plus prefixes and suffixes
13:29 kados thd: not in actual practice
13:29 kados thd: that may be the 'standard' but I've never seen a library do it that way
13:30 thd kados: at standards based libraries that are afraid to try or are still to large for Koha it is done that way.
13:30 kados ok, but we're not going to conform to standards to the exclusion of libraries like NPL that don't use them
13:31 kados we need to be all inclusive
13:31 thd kados: but most libraries in the world are not as standard as they might be
13:32 thd yes, I agree but I do not see how anything broke except that I was trying to thick of what would allow the system to work for any library with the least cataloguer effort
13:34 kados thd: the latest file you sent me has the following in the record editor's 942 field:
13:34 kados Item type *
13:34 kados Classification base (DDC to decimal or LCC letter class padded after single letter classes with trailing 0
13:34 kados Classification subclass (DDC after decimal or LCC number after letters
13:34 thd kados: if the library is using a nonstandard classification that is fine there are MARC 21 fields for that and they can also be captured by a plugin if filled and inserted into 942
13:35 thd kados: I was trying to think of how the range search that chris devised could work for LCC
13:35 kados NBBC has:
13:35 kados Institution code
13:35 kados item type
13:35 kados dewey
13:35 kados (LCC is autofilled by 050)
13:35 kados (I suppose dewey could be autofilled as well)
13:36 thd kados: I think the range search is numeric so LCC would need some manipulation
13:36 kados thd: why don't we just map the 'dewey' and 'classification' koha fields directly to the marc fields 030 and 080?
13:36 kados thd: the range search is being fixed by Tumer in head
13:36 thd kados: like I said the cataloguer should never have to touch 942 even if the library has its own special system
13:37 kados thd: it's never worked with LCC classification or with the MARC version of Koha and I won't have time to fix it
13:37 kados thd: so just forget about it for rel_2_2 I think
13:37 thd kados: it does do something with DDC if it is mapped the way I described
13:38 thd kados: the reason it did not work is that it was not mapped correctly
13:38 kados thd: fine, but that's the last thing I'm thinking about
13:38 kados thd: none of my clients have every asked for such a search
13:38 thd kados: my parenthetical comments were really meant as partial instructions for writing a plugin
13:39 kados thd: what is highest priority is ensuring that the classification / call number / location are clearly visible on the results pages
13:39 kados thd: so that people can actually find the items they're looking for on the shelves :-)
13:39 kados thd: we can deal with the search issues later
13:40 kados thd: so should we map Dewey to 080 and Classification to 050?
13:40 thd kados: Koha works fine as long as itemtypes is filled and you could write a plugin for the classification parts of 942 in an hour and save cataloguers a lot of work
13:41 thd kados: do you mean 082 DDC or 080 UDC?
13:42 kados thd: 082 DCC
13:43 thd kados: The search issues were not really my greatest focus as much as releaving cataloguers of work
13:43 kados gotcha
13:43 kados ok, I can work on that plugin
13:43 kados I still dont' quite understand what it should do
13:43 kados but first, how should we map things?
13:45 thd kados: it should search for the local cataloguing system fields to see if they were filled and then fill from the standard field that is specified in a system preference or with the possibility of modification by a per record flag for those libraries that want several different classifications in addition to the default
13:46 thd kados: that was the broad overview
13:48 thd kados: inspect 084 for a value if a system preference allows 084 to be used for a less common classification system.
13:52 thd kados: oops first check if some field ,942 whatever has overridden the usual pattern by 942 $whatever so that any record can do something different
13:53 thd kados: resuming, If 084 is empty even if other classifications are allowed then go to the common standard classification specified by a classification preference
13:55 thd kados: if the preference specifies DDC then the 3 digits before the decimal from 082 $a fill 942 $k biblioitems.dewey
13:57 thd kados: and the digits form 082 $a after the decimal fill 942 $l biblioitems.subclass
13:58 kados thd: ok, hang on
13:59 kados thd: we need to determine how to map things first
13:59 kados thd: we'll discuss the specifics of the plugin later when I have time :-)
13:59 thd kados: how to map what?
13:59 kados koha tables to MARC tables for classification
14:01 kados should we map 'dewey' to 082?
14:01 kados should we may 'classification' to 050?
14:01 thd ok kados: but it is very simple and similarly you can fill values for 952 in the same way so the cataloguer presses a few buttons and everything is done as if by magic :)
14:01 kados thd: i don't have time to do that today
14:01 kados thd: that's a weekend project :-)
14:02 thd kados: weekends and the days between are all the same to me :)
14:02 kados hehe
14:03 thd kados: the mapping is stable as long as you are not moving the framework between DDC and LCC or whatever
14:04 kados thd: what mapping?
14:04 thd kados: If you let the plugin do the work according to a preference then you can have a single default framework
14:06 thd kados: biblioitems.classification was intended for the equivalent of 852 $c or 852 $k
14:08 thd kados: biblioitems.dewey does not map cleanly to MARC because it was meant for part of 082 $a but you can certainly fill it consistently from a plugin
14:10 thd kados: biblioitems.subclass does not map cleanly to MARC because it was meant for the latter part of 082 $a but you can certainly fill it consistently from a plugin
14:12 thd kados: splt the $a string by the decimal if DDC and by the space if LCC or local system with spaces
14:12 thd s/decimal/decimal point/
14:14 thd kados: that algorithm will work every time so the cataloguer should only have to ensure that there is something in 050, 082, or 084 as needed
14:18 thd kados: if you are migrating records without such fields then you map the values from where they are in 852, 090, 092, 096 or wherever; or else you upgrade the records with a script that will be easy once we have struggled a bit for a well tested method
14:22 thd oh yes 090 has to move absolutely because it can very easily complicate copy cataloguing
14:51 kados thd: back from lunch
14:55 thd kados did you see my posts about mapping above?
14:59 kados thd: yes
14:59 kados what would the syspref be?
14:59 kados what would the possible values be?
15:01 thd One preference for defaultClassificationSystem
15:01 thd values
15:01 thd DDC
15:01 thd UDC
15:01 thd LCC
15:02 thd NLM
15:03 kados what about systems like NPL where there is more than one classification/call number system active at the same time?
15:03 kados also, what about systems that distinguish between classification and call number and location?
15:03 thd National Library of Agriculture (I do not know th e abbreviation)
15:04 thd kados: other systems are coming and then concluding with local system
15:04 kados ??
15:05 thd
15:05 thd kados: lastly local custom system
15:08 thd kados: I left out government document classification
15:11 thd kados: another boolean system preference would be allowOtherClassSystem for using an alternate classification system if filled.
15:11 thd kados: classification systems are the fun part of library science
15:11 thd kados: that OCLC list is very very short
15:12 thd not that all of library science is not fun but the most fun of all can be had with classification systems
15:14 kados thd: I'm just going to manually set up my client's classification system
15:14 kados thd: can't wait any longer :-)
15:14 thd :)
15:15 thd did you get my email with the small changes?
15:32 kados barcode and 942$j?
15:38 thd kados: yes that one
15:54 hdl hi
15:54 kados hi hdl
16:02 pierrick is anybody here?
16:02 hdl yes
16:03 hdl and kados and thd-back-soon
16:03 hdl ooops thd-back-soon will bbs :)
16:03 kados we can postpone the mtg until next week unless either of you have anything to report
16:03 kados I don't have anything new ... been struggling to get zebra 'sort' to work ...
16:04 hdl But pierrick said he had.
16:04 kados ahh ... ok
16:04 kados pierrick: what's up?
16:04 pierrick well... yes I still have questions :-)
16:05 pierrick first topic : the bugtracker
16:05 pierrick there seems to remain very old items
16:06 kados yep, I'm sure there are :-)
16:06 pierrick and in 2 weeks, only one update
16:06 kados it's not frequently used I'm afraid
16:06 pierrick what is the current status of the bugtracker ?
16:06 kados I think 2.2 is quite stabliized (except for the many bugs that noone wants to work on)
16:07 kados for 3.0 it's too early to begin using it since we don't even have a running system
16:07 pierrick having a bugtracker is a very good point, but keeping very old items unresolved is not a good point at all :-/
16:07 kados once 3.0 stabilizes a bit more it will become more important
16:07 kados pierrick: I agree, but the problem is noone seems willing to fix the old items
16:07 russ morning
16:07 kados pierrick: morning russ
16:07 kados oops :-)
16:08 paul2 hello world !
16:08 kados so guess we'll have a meeting after all
16:08 kados paul's here even :-)
16:08 pierrick wouldn't it be useful to close too old items that nobody will fix ?
16:08 kados pierrick: they still exist :-)
16:08 kados pierrick: so I think it wouldn't be that useful
16:09 pierrick old bugs still present, I suppose there are really minor bugs
16:09 pierrick does the bugtracker also use as "request tracker" ?
16:09 kados pierrick: in fact, some are not very minor
16:09 kados pierrick: sometimes it's used as an RT
16:10 russ i hope that during the dev week in may we will be able to do a bit of a tidy up
16:10 kados yea, it's a very good point
16:11 kados for a while we were doing weekly bug squasing sessions
16:11 kados maybe we need to revisit that
16:11 pierrick btw kados, closing a bug does not mean it's fixed, we can justify the close as "won't be fixed" (in the general case, I don't have an example in mind)
16:11 pierrick kados: BSP would be great
16:11 russ you can resolve as "later" as well - but i am not sure if that is all that usefeul
16:12 kados I count 174 unresolved bugs
16:12 kados for Koha>2.2.2
16:12 pierrick oldest ?
16:12 thd pierrick: that means that you did not try hard enough to fix it or it is a feature :)
16:12 kados we can ignore Koha<2.0
16:12 kados as those branches aren't maintained any longer
16:12 pierrick no customer still in koha<2.0 ?
16:13 thd kados: I suspect some of Katipo's customers would hope you are wrong about 1.X not being maintained
16:14 russ well i dont thing we have anyone on the 1.x series anymore
16:14 kados hmmm ... well russ'll have to speak to that
16:14 paul at least they are not officially maintained !
16:14 thd except that Katipo has its own bug tracking system in addition to
16:14 kados (as do all Koha companies I suspect :))
16:14 russ i think it is fairly safe to assume that those people using the older versions either don't notice the bugs, have fixed them or developed work arounds
16:15 paul just one note : I wake up at 4AM tomorrow. and it's already 9:15PM in France. So I won't stay for more than 45mn.
16:15 paul going to bed at 10 for sure.
16:15 russ yikes 4am
16:15 kados ok, so I will post a message about reinstituting the weekly bug squashing meeting
16:15 kados anything else to discuss?
16:16 kados paul: yowser!
16:16 russ i have put the conference pages live on the koha site
16:16 paul pierrick : no probl you didn't know : I come to see flc, not you ;-)
16:16 russ paul - i sent a test message to the form last night
16:16 kados russ: cool
16:16 paul I recieved it russel.
16:16 thd russ does that have the complete schedule?
16:16 russ paul: great
16:16 russ no not at the moment
16:16 pierrick paul: hope you'll come to say hello at least :-)
16:17 paul for sure !
16:17 russ it has broad details at the moment whilst the complete schedule is developed on the wiki
16:17 paul what do you call "complete schedule" thd ?
16:17 pierrick russ: I read KohaCon pages on today
16:17 thd paul: please clarify what the schedule actually is for KohaCon and after
16:18 paul ???
16:18 thd as complete as you would like to specify
16:19 thd paul: what dates are clearly or tentatively defined for what places
16:19 thd ?
16:19 pierrick maybe thd wants to know the estimated length of each presentation...
16:19 russ they are all confirmed
16:19 paul russ : right.
16:19 paul ENSMP is OK, CMI too, so dates are definitive. Unless there is an earthquake in France.
16:19 paul or a revolution.
16:19 kados or a riot :-)
16:20 kados heh, paul beat me :-)
16:20 russ yep closer to the time it would be good to publish a programme of speakers, topics and the schedule
16:20 paul yep.
16:20 paul we should publish it in french this week.
16:20 thd paul: now I am trying to work out the possibility of attending and would need to purchase air fair before the ticket would become to expensive.  Not that my passport has arrived yet
16:20 paul I'll translate it for you.
16:20 kados paul: thanks
16:20 paul thd : you'll come too ?
16:21 paul good news, I didn't knew.
16:21 thd paul: not if I do not know the dates clearly
16:21 paul you'll come for both weeks ?
16:21 thd paul: yes
16:21 kados thd:
16:21 thd if I can
16:21 paul easy : may, 2nd : meeting for everybody that want to discover Koha. Not useful for you..
16:22 paul may, 3rd : meeting for all koha users in Europe. Mostly french I bet, but maybe some other, we will see.
16:22 thd I still want to discover the mysteries of Koha :)
16:22 paul may 8th -> May 12 = developper week in Marseille.
16:22 paul you're welcomed, of course.
16:22 paul there will be 2 other librarians.
16:22 paul just beware that the hotel should be reserved ASAP
16:23 paul all infos on
16:23 paul KohaCon page.
16:23 thd paul: I am actually not a librarian nor have I ever been employed as a librarian
16:23 paul pierrick : same note for you : reserve your hotel too.
16:23 paul ? really ? I thought, as you know everything from MARC21 & UNIMARC.
16:23 pierrick paul: all right, I'll do it tomorrow
16:24 thd paul: I was formerly a bookseller with a library science background and some understanding of computer systems
16:25 thd paul: I wrote a system using MARC records for my bookshop at the same time that Koha was being created
16:25 thd 1999
16:25 paul did everybody look at[…]63@N00/115366281/ ?
16:26 pierrick course I saw it, he send an email with it on koha-devel
16:27 thd very pretty
16:27 kados paul: yes, it will fit nicely on when I finally launch it :-)
16:27 kados (the plugin I mean)
16:28 pierrick this is the maybe moment for my second question of the meeting : templates
16:28 kados sure
16:29 pierrick in current head, there are many template/language, why don't we work only on prog/en during development
16:29 kados I think that's the plan
16:29 paul we aare supposed to do that
16:30 paul everybody ONLY work on english at least.
16:30 pierrick Wouldn't it be easier to have only one template/language in CVS and other downloadable on as "extensions" ?
16:30 paul pierrick ++
16:32 pierrick anybody would be able to add/update extensions
16:32 paul although i'm still thinking that it would be best to have only 1 template.
16:32 paul with stylesheet + maybe header/footer specific.
16:32 pierrick thus encouraging participation from librarians
16:34 kados right
16:34 kados very good ideas
16:34 paul "right" for who ? pierrick or me ?
16:34 kados both :-)
16:35 paul there is something very basic & needing manual setup on, with stylesheets for css templates.
16:35 kados it would be neat if we had a tool for translating templates online
16:35 pierrick something like launchpad???
16:35 thd I suspect the reason that has not worked to date is that people want to use the template to support different features and presentation to an extent that exceeds mere stylesheet differences
16:35 kados I don't know launchpad
16:35 pierrick (rosetta more precisely)
16:36 paul kados : we already have something :
16:36 pierrick launchpad is a kind of project manager for Ubuntu
16:36 paul it just need something to "install" and "update" the text strings with koha
16:36 kados ahh ...
16:36 paul I never found the 2-3 hours needed to do this.
16:36 kados we need to popularize the kartouche
16:37 paul if someone has a better software for this, he can suggest : kartouche is a dead project It seems
16:37 kados paul: will you have time before 3.0? :-)
16:37 pierrick kartouche is exactly the same as what launchpad/rosetta offers
16:37 paul no new release since 0.2
16:37 kados bummer
16:38 kados pierrick: does rosetta use 'PO' files for translation?
16:38 pierrick I suppose, but rosetta is not a free software :-/ we can't install it on
16:38 kados hmmm
16:38 paul so it's not a possible candidate :-(
16:38 pierrick it's a servicve
16:39 thd pierrick how unfree is it?
16:39 pierrick thd: it is not donwloadable
16:39 kados "No, Rosetta is not Open or Free Software at the moment. Rosetta will become open source sometime in the future but we don't have a date, although some parts of the Launchpad have already been released under the GPL by Canonical Ltd."
16:39 kados so it must still be in 'beta'
16:40 pierrick thank you kados for the fast search
16:40 kados and they want to wait until it's fully functional before making a public release
16:40 kados understandable if that's the motivation
16:41 kados so for now we all agree to only work on prog/en
16:41 pierrick would you be interested in an extension manager for ?
16:41 kados and we'll continue to look for a good template translation system
16:41 kados pierrick: could you explain what you mean?
16:42 paul the only/main difficulty being to port templates from rel_2_2 to head.
16:43 pierrick kados: imagine a new customer has decided to fully rewrite a new template, the extension manager would let this customer offer his work to the community
16:43 kados right, that will require a lot of grunt work :-)
16:43 kados pierrick: that would be great!
16:43 pierrick kados: without adding it to CVS
16:43 kados pierrick: in fact, that's how I envisioned
16:43 kados pierrick: (doesn't exist yet)
16:44 kados pierrick: kind of like Wordpress 'plugins'
16:44 pierrick kados: I'll work with you on this if you want... we developped such a tool on my other project
16:44 kados pierrick: that'd be great!
16:44 russ or oscommerce contributions
16:44 kados right
16:44 kados ok ... don't want to keep paul up
16:44 kados anything else ?
16:44 pierrick russ & kados: do WP and oscommerce give away their tool ?
16:45 paul a question, about pierrick : what kind of official position could we give to ineo/pierrick ?
16:45 kados paul: good question
16:45 kados pierrick: WP is open source
16:45 kados QA Manager is up for grabs IIRC
16:45 paul pierrick : no need to stop speaking, you can have an idea !
16:45 pierrick kados: in know WP, but the plugin manager ?
16:46 kados pierrick: what position would you like? :-)
16:46 kados pierrick: I think so ...
16:46 pierrick how do you define QA manager position on Koha ?
16:46 kados very loosely :-)
16:47 kados responsible for ensuring high quality of the project code
16:47 russ how do you define any position in Koha
16:47 paul the guy that we could kick if we have bugs on a stable reelase!
16:47 russ very loosely
16:47 kados hehe
16:47 paul lol
16:48 pierrick paul: you meant "the guy who would be allowed to kick if coders let bugs in their code"
16:48 kados pierrick: if you are QA manager I can assign you with the task of organizing the Bug Squashing sessions :-)
16:48 thd the guy responsible for ensuring that there is no stable release until all bugs are fixed
16:48 paul no pierrick : I mean the guy that libraries can kick if he accept a release that has bugs.
16:48 kados hehe
16:49 pierrick not sure I see this role as you :-/
16:49 thd This is how Debian goes 3 years between stable releases
16:49 pierrick thd: I agree
16:49 kados pierrick: how would you define a QA manager?
16:50 kados better yet, what official role do you want to fill? :-)
16:50 paul I think the QA manager is responsible for heavy testing. and organising bug squashing.
16:50 pierrick code reviewing, keeping an eye on the bugtracker so that no bug stays too long
16:50 kados excellent
16:51 thd pierrick: you can invent a role that has not been described yet
16:51 kados pierrick: is it something you are interested in holding? or would you prefer something else?
16:51 pierrick paul: I do not agree with you, QA manager do not have to heavy test the application, not in open source model
16:51 pierrick QA manager is an interesting position, as long as I do other things :-)
16:52 russ sounds like a good position to learn alot about koha
16:52 thd pierrick: what is the distinction between open and closed source QA?
16:52 kados pierrick: Koha may be different than other OSS systems because the software must be tested before placing into production as bugs would seriously comprimise a library's workflow
16:52 pierrick if in 2 months QA is stil on the same level, it would mean I'm not made for this
16:53 kados heh ... fair enough :-)
16:53 kados ok ... congratulations then pierrick, our new QA Manager
16:53 russ :-)
16:53 pierrick let's make a try
16:53 kados sounds good
16:53 kados anything else to discuss at the meeting?
16:53 paul maybe pierrick should ask for confirmation from ineo ?
16:54 pierrick thd: I'll answer you later, but the question is very interesting, I've worked on the two sides
16:54 paul as he's ineo employee.
16:54 russ pierrick - perhaps to the list
16:54 paul (although I already spoke of this many times to flc & jps)
16:54 russ the devel list - i'd be interested to read your observations
16:54 pierrick kados: yes I have another point
16:54 kados ok
16:55 pierrick koha project on Freshmeat has not been updated for a long time
16:56 kados I didn't even know we were listed
16:56 pierrick russ: OK, I'll make my comments on difference between open and closed source QA on koha-devel
16:56 russ
16:56 pierrick we are :-)
16:56 kados we need a PR manager :-)
16:56 russ pierrick : thanks
16:56 pierrick PR ?
16:56 kados pierrick: I'll see if I can update the entry
16:56 kados Public Relations
16:57 pierrick a very hard task I suppose
16:57 thd pierrick: the people who tell you that toxic sludge is good for you :)
16:57 paul and probably not one for a non native english
16:57 thd and do it with a smile
16:58 kados yep, thanks for that russ
16:58 paul and if he's really good, once you're ill, he sell you the medicine to help you...
16:58 pierrick an unusual position in an opensource initiative
16:58 pierrick only big projects have that (Mozilla or
16:59 kados right
16:59 russ it something i'd like to talk about at the dev week
16:59 kados yep, me too
16:59 paul russ :[…]discoverkoha.html
16:59 paul you write "May, 3rd" when speaking of may 2nd.
16:59 russ doh!!
17:00 russ 2 secs
17:00 paul[…]registration.html
17:00 paul the form contains 3rd & 4th of may, instead of 2nd and 3rd
17:00 kados it's 3rd and 4th in NZ :-)
17:01 pierrick kados: impressive
17:01 paul "we will contact you to arrange payment"  + the price : it will be free of charge
17:01 paul as we would have to pay ENSMP if the meeting is not free.
17:01 paul + it would be highly more complex for ENSMP as well as for us.
17:02 paul so, we have decided to have it for free.
17:02 thd paul: formerly you spoke of fees to ensure attendance but now that has changed?
17:02 paul yes.
17:02 russ ok that should all be fixed now
17:02 paul when we spoke to Pascale Nalon, ENSMP, she explained to us that
17:03 paul ENSMP can't (legally) host such a meeting for free if it's not free.
17:03 paul ENSMP is a school from french government.
17:03 paul so very heavy on administrative rules !
17:04 thd paul: free software should have free meetings but I understand your reasoning to ensure larger attendance
17:04 paul that was my opinion (+ we hope this meeting will provide us some customers !), but not ineo one.
17:05 paul finally, it's ENSMP that forces the choice !
17:05 thd as people would be disinclined to waste something for which they have paid
17:05 paul right
17:05 paul pierrick : in fact, I think that pate, our previous kaitiaki was quite a good PR guy.
17:06 paul it's 10:10. anything else ?
17:06 thd paul: is there some way to substitute something that is not money but would function in the same way for the purpose of larger attendance?
17:06 paul if you have an idea, let me know, we are open to any !
17:08 thd paul: you might ask for a confirmation of attendance by email, telephone etc.
17:08 paul that's what we plan to do.
17:08 pierrick another very small point: would it be possible to log your commits on the Wiki ?
17:09 thd paul: If you could extract a promise or intention of attendance then people may be reluctant to break there declaration of intention.
17:09 pierrick (add a small summary when you create a new version of a page) the RSS feed becomes far more readable
17:11 thd paul: we have in English the passive closing of an invitation with the abbreviation of the French phrase RSVP
17:12 paul ???
17:12 thd paul: But you could be more aggressive and attempt to personally contact anyone who was sent an invitation or announcement
17:13 pierrick thd: good idea
17:13 paul I think that Ineo plans to invite all important libraries in France personnaly.
17:13 paul with a mail & maybe a snailmail as well.
17:13 russ paul - are you going to do an email for the koha and koha-devel lists?
17:14 paul for sure.
17:14 pierrick on the last free (gratis) open source conference I went to, I had to register and I was called by phone to confirm
17:14 thd paul: r?pondez s'il vous pla?t is RSVP
17:14 russ ok i gotta go
17:14 russ back in an hour or two
17:15 pierrick bye russ
17:15 russ bye all
17:15 thd pierrick: I expect to see all the important librarians in France there
17:15 pierrick good night paul
17:16 thd good night paul
17:16 hdl good night paul
17:16 pierrick thd: INEO expects that too, but not biggest libraries
17:16 pierrick paul: and see you tomorrow :-)
17:16 paul bye all.
17:19 thd perrick: will no one from BNF be there?
17:20 pierrick BNF ??? I can't be 100% sure but I don't think so
17:21 thd pierrick: They should have some computer systems person who is curious
17:21 pierrick thd: it's not how I understood this meeting, IMO it was for potential customer, and BNF is not a potential customer
17:21 thd pierrick: yes you are correct
17:21 pierrick thd: but I may be wrong, we'll see on May, 2nd :-)
17:22 thd of course I have not seen the announcement notice that you would have sent
17:23 pierrick meeting status ?
17:23 pierrick (I've got nothing to add)
17:24 thd kados must have wandered already as he is too busy today
17:25 pierrick OK, it's 22h24 here, I go back home to help my wife and the baby, good night #koha :-)
17:26 thd good night pierrick
17:57 thd kados: are you around?
17:58 kados thd: barely
20:33 russ mason - what is the uid/pass for[…]/koha/
20:34 russ ack
20:34 russ wrong #
22:06 thd kados: what are you seeing that I did not or that I did?
22:06 kados thd: nevermind that, I've got it resolved
22:07 kados thd: what I'd like to discuss is itemcallnumber if you have a couple of minutes
22:07 thd kados: stared at the problem and it evaporated
22:07 thd ok
22:07 kados thd: paul has a syspref for itemcallnumber but I'm not clear on what it's supposed to do
22:08 kados thd: or how to relate his syspref to the 952 spec you created
22:09 thd kados: it did until quite recently fill items.itemcallnumber with the contents of whatever field/subfield was given in the preference
22:09 kados thd: it doesn't do that now?
22:09 thd kados: I had suspected your great bugfixes to the editor of breaking it
22:09 kados thd: hmmm ...
22:09 kados that could be
22:10 kados hmmm
22:10 kados well actually, I don't see how
22:10 kados because I've never edited or .tmpl except one minor change
22:10 kados that only affected encoding
22:11 thd kados: no it does nothing now but formerly you could set 050a in the preference and it would fill half of the LC call number in whatever field/subfield had been set to item.itemscallnumber
22:11 kados thd: as with MARC subjects, I'm very confused about how classifications are supposed to work
22:12 thd kados: It had the defect of being limited to one subfield when what you would want is 050 $a space $b
22:13 thd kados: $a is the classification number for the base classification
22:14 kados what is the purpose of splitting up the classification in the record, is there any advantage other than that's the standard way to represent it?
22:15 thd $b is a number to distinguish a particular title from all the other titles sharing the same place in the classification system and is usually derived form the author's last name according to rules in a table
22:16 kados but why split it up? why not just have everything in a single field?
22:16 thd kados: the splitting allows treating the separate elements separately and it is certainly an advantage for searching
22:17 kados in simple cases, like the client I'm working with now, there are only a few hundred items in the collection
22:17 kados I think a minimal holdings and classification system is needed to deal with this case
22:18 thd kados: everything in a single field gives false matches when you want to search only the classification hierarchy and not the silly number that distinguishes multiple biblios occurring in the same place
22:18 kados homebranch/holdingbranch, cost, classification, date acquired, public note
22:18 kados I think those 6 fields are all that are needed
22:20 thd kados: several of the subfields should be filled by a plugins so there is no work for the catalaloguer
22:21 thd kados: the cataloguer should not even need to fill any classification fields for copy catalogued records
22:22 kados thd: unless they want item-level classification
22:25 thd kados: classification subfields should cascade from the front of the record maybe a value list for JUV prefix or appending the copy number after the suffix or something might be needed but then there is no need do anything for items.itemcallnumber because that would auto-fill from concatenating the other subfields with space or slash separation.
22:26 thd so a copy number usually establishes the item
22:27 thd kados: 741.53 A54 copy 2
22:29 thd kados:852 $h741.53$iA54$t2
22:31 thd kados: the spine label would be 741.53\nA54\nc. 2
22:33 thd kados: certainly different copies would need different barcodes where preprinted barcodes are usually the most efficient with scanning them in rather than trying to print a known number.
22:34 thd although barcodes are not required by MARC
22:34 thd only by the Koha framework
22:35 thd after I changed it back earlier :)
22:37 thd however 003 is mandatory within the system.
22:38 kados ok
22:40 thd obviously LC is not exporting records with 03 but that may be Eneavor Voyager's fault.
22:47 thd kados: so if the appropriate subfields are minimised for items or not appearing then the cataloguer should not have to do much more than scan in a preprinted barcode as anything else relevant such as holding branch can be set from context unless all cataloguing is centralised.
22:48 thd kados: are you still there?
22:51 kados thd: kindasorta
22:51 thd kados: are you still awake?
22:51 kados thd: barely :-)
22:52 kados thd: I agree the cataloger should hardly have to add anything but the barcode
22:52 kados thd: but that plugin will be a weekend job unfortunately :-)
22:52 kados thd: I've got several clients that need some love this week :-)
22:52 thd kados: let us leave the other question until tomorrow if you will have some time then when you are fully awake
22:52 kados ok
22:56 kados thd: do you happen to know if there is an authoritative list of non-filing words/characters for MARC21 in English?
22:56 kados thd: The the a A obviously, any others?
22:56 kados (not stopwords but rather words that are not used for sorting)
22:57 thd kados: you use the indicator position if you are reading a preexisting record
22:57 kados thd: we can't do that with zebra
22:57 kados thd: we need to rely on a list of words that should be ommitted for sorting (but not for searching)
22:58 thd kados: so are you speaking about a plugin for setting that position in the editor when creating an original record?
22:58 kados thd: no, I'm talking about a zebra configuration file that specifies which words to omit for sorting
22:59 thd kados: there may be an AACR list for both possibilities as the answer would be the same.
23:01 thd kados: I think it is simple in English but French is tricky because they contract their prepositions
23:01 kados yep
23:03 thd kados:s/prepositions/articles/
23:04 kados prepositions ... you sure?
23:05 kados 1-9 Number of nonfiling characters present.  A field begins with a definite or indefinite article that is disregarded in sorting and filing processes. Enter the number of characters in the article, plus spaces, punctuation, diacritical marks and special characters that precede the first significant word.
23:05 kados Any diacritical mark or special character associated with the first filing character is not included in the count of nonfiling characters.
23:05 kados from:
23:05 kados thd: so just articles The and a then
23:05 kados w00t, got that working already
23:09 thd horay and I could not find the rule in my old printed copy of AACR2, suggesting that it was a MARC and systems question at least in 1978.
23:13 kados heh
23:34 thd kados: did you miss 'an'?
23:34 kados thd: yes, thanks :-)
23:46 rach good work on presuring perick into taking on qa :-)
23:46 rach well persuading
00:18 thd kados: INITIAL ARTICLES
00:18 thd For languages: Arabic, Dutch, English, French, German, Hebrew, Hungarian, Italian, Portuguese, Spanish, Yiddish
00:18 thd[…]ols/initiala.html
00:18 kados thd: nice, thanks!
00:20 thd kados there may be something in Cataloger's Desktop. Have you tried using it?
00:21 kados thd: not yet, but I'll give it a show asap
00:21 thd unfortunately frames retard sharing links
04:14 osmoze bonjour :)
04:27 hdl bonjour osmoze
09:34 pierrick thd: I've answered your question about proprietary Vs OSS regarding QA, on koha-devel
10:11 kados morning all
10:11 pierrick hello Johsua
10:11 pierrick Joshua even
10:12 kados pierrick: nice email sent to koha-devel
10:12 pierrick kados: thanks :-)
10:12 pierrick very personnal point a view, need to be discussed
10:12 kados yep ...
10:13 kados one problem I forsee with no QA testing is that many clients don't want to run buggy software to discover the bugs
10:13 pierrick I saw Paul his morning just to say hello, he won't be online until friday
10:13 kados ahh
10:14 kados Koha is not like many other OSS applications because it is actually quite difficult to convince a library to use it :-)
10:14 pierrick I didn't mean no QA testing, I mean QA testing with involved customers
10:14 pierrick oh OK :-/
10:15 kados ahh ... yes the customer who sponsored a feature should be heavily involved in QA testing, I agree
10:16 pierrick IMO, we can't garantee the 100% stability having such an open development model
10:16 pierrick did you test the commits I made ?
10:16 pierrick who did ?
10:16 hdl hi kados
10:16 pierrick If I didn't, nobody did it I suppose
10:17 pierrick (hi hdl)
10:19 hdl kados or thd : For authors, you can have up to Three major subfields : 100$a, 245$c and 700$a, what is the rule ?
10:19 kados hdl: there are more than three :-)
10:19 kados but 100$a is the most important one
10:20 kados 700$a is where additional authors go
10:20 kados 245$c puts author as it appears on title page of the book
10:29 pierrick kados: when you have time and if you're still interested, you can ask m about the extension manager...
10:32 hdl kados: For our customers, what is commonly used is 200$f which could be related to your 245$c
10:32 hdl That is why I am quit confused.
10:32 kados hdl: thd is the expert, I just dabble :-)
10:34 hdl But when linking to Koha NonMARC field, I know only ONE MARC-subfield should be linked.
10:34 hdl unless results could be quite awful.
10:48 kados I think 100$a is the 'one' to link
10:50 pierrick kados: if you answer to my mail about QA management, could you explain what *you* are expecting from a QA manager please?
10:51 kados sure
10:51 kados I will try to find the time today
10:51 kados and about extension manager, maybe we can discuss it on Thursday? (today I'm strapped for time and tomorrow I have a demo of perl-zoom for NPL)
10:51 pierrick :-) not a high priority, do it when you have time
10:52 kados BTW, hdl and pierrick, chris and I got perl-zoom to sort last night
10:52 pierrick thursday, I'm visiting a big customer to see what is the usual work of INEO
10:52 kados cool
10:53 hdl Great !
10:53 hdl (kados and chris)
10:54 pierrick sort on Dewey you mean ? I saw something like that in CVS
11:00 kados well, the index now supports sort by Dewey
11:00 kados but the code's not there I don't think
11:00 kados unless chris got to it while I was sleeping
11:00 kados now, for some breakfast
11:02 osmoze by all
11:02 hdl bon apettit.
11:02 osmoze e
11:02 hdl bye osmoze

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