IRC log for #koha, 2005-07-04

All times shown according to UTC.

Time S Nick Message
17:44 thd kados: are you there?  Did I miss the announcement?
20:20 kados thd: no ... haven't done it yet because I'm waiting to hear from someone
20:20 kados thd: it'll all become clear soon enough ;-)
09:07 paul hello world
09:07 kados hi paul
09:07 paul hi joshua
09:08 kados have you seen bug 997?
09:09 kados and 998?
09:11 paul yes.
09:11 paul 997 is already partially fixed on my desktop.
09:11 paul 998 is strange.
09:12 paul but i have an idea
09:12 kados paul: cool ... thanks
10:00 thd paul: How did you partially solve 997?
10:01 paul hi thd.
10:01 thd paul: hello
10:01 paul for instance it's not completly done. but here is what i did :
10:02 paul - fix a bug in (a sql was ordered on subfieldcode and not subfieldorder)
10:02 paul - add the possibility to modify the subfieldcode.
10:02 paul so, we can store :
10:02 paul 700$x$x$y$a
10:02 paul as expected.
10:03 paul what is still to be done :
10:03 paul - adding a button to "up" or "down" a subfield.
10:04 paul (are you the bugreporter ?)
10:04 thd paul: what about a selection list for the value of the field?
10:04 paul (as me)
10:04 paul would make a too big html page.
10:04 paul some of my libraries already have a 75kB page.
10:04 thd thd: yes, blame the messenger :)
10:04 paul we reach the limits of HTML here..
10:05 paul (you're the bug reporter ?)
10:05 thd thd: yes I am the bug reporter for both 997 and 998, you can blame the messenger now :)
10:06 paul tu es francais ?
10:06 thd paul: no
10:06 paul (i ask because alinto is a french company, funded by a french of mine)
10:07 thd paul: alinto has the best freemail service for my uses right now, I like it very much.
10:07 paul (funded by a friend of mine)
10:08 paul (not french of mine :-( )
10:08 thd paul: few freemail services have POP support anymore
10:09 paul right. that's why I have my own, ;-)
10:09 thd paul: I use alinto for anonymity
10:10 paul I don't need anonymity, I need to be well known in the library world ;-)
10:10 paul (to get customers)
10:12 thd paul: Unserstood, in some contexts I avoid people from other contexts knowing what I am doing where the other context may hold something against me until I think it is safe.
10:13 paul of course I understand !
10:16 thd paul: if a subfield entry line presented a list of possible subfields adjacent to the entry line in a drop down selection box for that field only why would that be too much HTML?
10:16 paul because on a page with 40 subfields, that would mean 40 select lists.
10:16 paul (or there's something i didn't understood)
10:17 thd paul: how often do you have 40 subfields?
10:18 paul that's usual when creating a biblio
10:18 paul (with repeated authors, subject, that means around 15 fields, with 1-5 subfields in each)
10:19 thd paul: I guess I was thinking about the problem one field at a time
10:19 paul but we can't afford the price of a html request for each field/subfield isn't it ?
10:22 thd paul: even still with the fields you have commonly on one page, do you actually have 40 or too many in the 10 paged view?
10:22 paul ???
10:22 paul (not clear)
10:23 thd thd: you don't have all the fields in one page just five or so and then you fo to the next five
10:23 paul you mean in the page with the 10 tabs ?
10:24 thd paul: yes the whole record is not there on one page
10:25 paul i think the user don't want to have to wait between pages. The actual page loads the 10 tabs in 1 HTML page
10:25 paul I think that in the future, the best would be to have something not in pure html.
10:25 paul something like XUL
10:26 thd paul: I did not realise, I have not been thorough in my examination of koha yet.
10:26 thd paul: wouldn't 10 separte pages be better?
10:27 paul http protocol is too slow for this I think.
10:29 thd paul: how much slower is it to fetch 1/10th of the information 10 times as opposed to 10 times the information once?
10:30 thd paul: all 10 pages are not usually even used are they?
10:31 paul it's not a question of measurable speed, but user feeling. waiting for 2 seconds before beginning is better than waiting 4 times 0,4 seconds.
10:31 paul (+ there is html rendering time)
10:32 thd paul: everyone should be using elinks :)
10:34 thd paul: Is there a way to simplify the design so that any page renders faster on any browser?
10:35 paul maybe a little, but only a little I think.
10:35 paul in the future, I think the best solution would be to let the user having the MARC editor he want.
10:36 paul the "koha internal" or "XXX", or "YYY"
10:36 thd paul: I like the editor in Koha, most are free form subfield entry which allows mistakes to be made more easily.
10:37 paul right, that was my 1st goal : having something that does not require a professionnal MARC cataloguer.
10:37 thd paul: you did an excellent job with that
10:38 thd paul: even professional cataloguers make mistakes, just look at any large data set, especially OCLC.
10:39 paul (already enough to see with...)
10:41 thd paul: it would seem that having the whole record in html at one time is a great burden if full cataloguing on all fields is enabled.
10:42 paul right. But that's rarely needed. So you can define frameworks to limit the number of fields you need (1 for serials, 1 for monography...)
10:43 thd paul: I had envisioned a model where the cataloguer could press a link or button to activate less commonly needed fields for the type of item being catalogued.
10:44 thd paul: I had imagined the tabs worked in a similar way to fetch susequent pages in the record
10:48 thd paul: If the model does not require the whole record on one page, then very complete cataloguing can be supported where all relevent fields for the item type are a click or two away.
10:53 thd paul: under the current model, if you have too many fields enabled for the occasion that they might be needed then the MARC editor loads more slowly than it would need to for most records of the item type that would not need the obscure fields.
10:59 thd paul: am I being unclear, apart from my spelling errors?
10:59 paul no, you're clear.
11:00 paul & i agree with you on a theoric point of view. but for instance, the MARC editor has some weaknesses that makes them correct but not powerful.
11:00 paul s/them/it/
11:02 thd paul: I think you could change the model a 'little' and have something to support common MARC cataloguing and complete deluxue cataloguing in the same model.
11:03 paul explain how to do it...
11:03 thd paul: It would be a tool to die for :]
11:04 thd paul: specoify the item type first as you already do.
11:06 thd paul: then each page for an appropriate minimal standard is loaded separtely as I had thought the tabs did.
11:07 thd paul: at any page or point in the record a link or button fetches a more complete set of subpages for all the marc fields appropriate to the record at that point when needed.
11:08 thd paul: you still have separte pages when you expand the level of cataloguing detail so the page loading burden for each individual page is low.
11:10 thd paul: if you only need full marc in one section you can go back to the standard view in another section
11:11 thd paul: koha needs 'killer' tools to say that it not only does x but it does x better than any other program
11:12 thd paul: I will help for whatever I can to make it work, obviously I have a big gap in knowing just how koha does some things already
11:14 thd paul: the selection lists for the subfields would be less of a burden under that design
11:14 paul right. Your suggestions are good.
11:16 thd paul: if the user changed the order form a good default then the page might have to be reloaded to support the value lists or authorities thesaurus in the correct adjacent location.
11:18 thd paul: Of course it would help if there was a group institutions to support the work for a cataloguing tool that worked as well for the non-M
11:19 thd ARC expert as for cataloguers who know every field backwards.
11:20 thd paul: I think your basic user friendly forms design is better for everyone, even if expert cataloguers are faster with free form subfield entry into a text box.
11:21 thd paul: what you loose on speed for the experts you gain on a degree of automatic validation for everyone
11:24 thd paul: do you know of commercial tools that use your guided forms approach to subfield entry?  My experience seeing commercial cataloguing tools is a little to old and incomplete.
11:25 paul no thd. i'm not a cataloguer, i'm a developper. So i only have a minor knowledge of commercial ILS.
11:26 thd paul: the koha project needs a closer liason with the most expert users.
11:28 paul right. When I was Release Manager (2.0 and 2.2), I spoke a lot with "my" libraries, but they are mostly "small" libraries. very happy with actual MARC editor.
11:28 paul joshua/kados, has views/ideas/will for larger libraries.
11:29 paul where my code is a little bit under-featured.
11:29 paul i'm happy with joshua goals & objectives. He will make Koha a great product (& i'll help him where possible)
11:30 thd paul: Almost every developer sees programming solutions in terms of how to most efficiently solve problems rather than tackling the difficulties head on for what the users imagine they should when they are dreaming clearly about an ideal world.
11:32 thd paul: Programming becomes harder when you tackle the greatest difficulties head on but even the frustarations are more fun and the result is more rewarding.
11:36 thd paul: As for commercial ILS, it is good to know what the competition is doing so you can take the best ideas for your own and have a superior product.
11:38 thd paul: Of course, I do not believe in software idea patents.  You have to be careful about British Telecom claiming HTML even in Europe soon I fear.
11:40 thd paul: BT is over that foolishness over HTML but there will always be someone in the US at least causing trouble with patents.  So a modest amount of care is warranted in the current IP world.
11:43 thd kados: are you back from lunch?
11:44 kados thd: yep ;-)
11:44 kados thd:  just finished reading your discussion
11:46 thd kados: even though my programming experience is less deep than it should be I have generally employed it toward difficult problems and been pleased with how quickly they can sometimes fall once you consider them seriously.
11:48 thd kados: my lack of deep programming knowledge is often the greatest burden in knowing how to do some simple things the common way.
11:48 thd :)
11:49 kados :-)
11:50 kados I think the best approach to pattents is to simply ignore them
11:50 thd kados: Do you know of commercial tools that use Paul's guided forms approach to subfield entry?  My experience seeing commercial cataloguing tools is a little to old and incomplete.
11:50 paul joshua/kados : i've a ta_MY translation files (.po) in my mailbox
11:50 kados the best cataloging tool I've seen is ITS.MARC from TLC
11:51 kados paul: nice!
11:51 thd kados: Ignore patents and remain ignorant about them.  Knowlegde entails triple damages in the US.
11:51 kados :-)
11:51 kados well the best way to prove you had an idea is to code it
11:54 thd kados: I had not realised the tabs were not separate pages there is more work there than I had first supposed.
11:54 kados not sure why that means there is more work
11:55 kados what's wrong with having everything on one page?
11:56 kados I think we may need to employ some javascript on the cataloging tool to provide the ability to generate a new subfield for repeatable fields
11:56 thd kados: I thought the MARC record had separate pages already for cataloguing.
11:56 kados and also to handle some hotkeys for authorized value and for common tasks
11:56 kados macros, etc.
11:56 kados why does it need seperate pages?
11:57 kados separate even ;-)
11:58 kados paul: in the record you showed me earlier wherever there are little icons (magnifying glass) means there is an authorized value that you can search?
11:58 thd kados: Read the discussion with Paul about separate pages again.  One record is a  burden for supporting full MARC cataloguing for all the obscure fields when needed by a particular cataloguer.
11:59 paul kados : no. it's a way to search for a field value to "jump" from biblio to biblio
11:59 thd kados: , for an appropriate item that deserves full cataloguing.
11:59 paul i spoke of this with owen last friday
11:59 kados paul: I'll read the logs
11:59 paul ok, thx

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