IRC log for #koha, 2005-10-07

All times shown according to UTC.

Time S Nick Message
12:19 Sylvain paul je viens de mettre un fix pour le bug #1013 dont je t'avais parler l'autre jour. Si tu peux vérifier que ça ne casse rien et que ça fonctionne aussi chez toi
12:31 Sylvain bon, j'y vais, si y a un problème, dis le moi
12:31 Sylvain ciao
18:13 russ kados: you about?
18:28 russ hi thd
18:28 russ thd: have you made any more progress with the features section for the website?
19:40 thd russ: yes I have about three weeks work into it so far and practically put myself on  NZ time with sleep loss.
19:40 russ crikey
19:41 thd russ: As I said it is more than just a features list.
19:42 russ have you got anything we can take a look at?
19:43 thd russ: Yes, but I would be afraid to give the wrong impression from just one whole document.
19:49 russ i'd really like to have a look if at all possible
19:52 thd russ: It is currently one large document that needs division by module and development level if not more than that to be readable.
19:55 thd russ: I would have posted something and kept adding to it, however, the design as a large document to be divided has meant that posting it by degrees would be double work for each dividing process.
19:55 russ right
19:56 thd russ: even suitably divided it might need to be trimmed back to serve as a feature list.
19:57 russ right
19:57 russ any idea how much more you have to do?
19:57 thd russ: I was grepping for some descriptive text I had written in a paragraph from yesterday but I may not have save that document.
19:59 thd russ: That is the consequence of sleep loss.  Forgetting to save ones work :)  That was not of great importance.
20:00 thd russ: I have a tendency to work until I cannot stay awake.  That only works for a few days without good rest then productivity falls to almost nothing until the sleep deficit is restored.
20:00 russ oh well take care, i don't think it is so important that you need to jepordise your sleep
20:02 thd russ: As a feature list it is not overly important however I am trying to do other things with it.
20:03 thd russ: Koha Analytical Feature List and Documentation Index with Demonstration and Example Implementation Links is a title that I have attached to the comprehensive draft file.
20:05 russ that is a mouthful :-)
20:05 thd russ: So there is the feature list purpose to show not just the obvious and superficial but also the features that cannot be seen without close examination.
20:05 thd russ: Feature list is good for
20:06 russ yep
20:07 thd russ: It is easy to come away with a poor impression of Koha by only seeing what is obvious in the demonstration and a cursory examination of documents.
20:07 thd russ: This is especially true with MARC Koha.
20:07 russ well i dont think poor is the right word, maybe incomplete is a better description
20:09 russ if you could keep us posted on your progress that would be great
20:09 thd russ: The default configuration gives the impression that MARC support is extremely incomplete instead of modestly incomplete in some ways.
20:10 thd russ-lunch: I will explain some other utility after lunch
20:10 rach :-) you really are on nz time
20:11 thd :-)
20:11 thd rach: actually I have not had breakfast yet today
20:12 rach maybe you should do that so you're in sync with each other :-)
20:15 thd kados: are you around
20:15 thd kados: I have some information and a question about adding MARC serials support to Koha
20:30 thd russ, rach: I found the missing paragraph, for when you get back from lunch.  It is good to know that I can save documents while sleepwalking.
20:43 thd russ: From my lost and now found paragraph...
20:44 thd This is meant to show prospective adopters of Koha the full range of hidden features that are not obvious from superficial examination.  It is hoped to guide prospective feature funders and supporters otherwise to know what features may need support to initiate or accelerate development.  Developers are also reminded of what is needed for the roadmap and possible future directions for Koha.
20:46 russ ok cool - that gives me a better idea of what you are trying to do - thanks
20:46 thd russ: The point where it becomes complex is that is an analytical features list.
20:47 thd russ: Features are not merely listed in a flat sequence.
20:48 thd russ: Many important features are broken down into there constituent elements.
20:48 thd russ: The organisation is a large indented outline.
20:50 thd russ: In HTML, it requires ordered list and dictionary list tags.
20:51 thd russ: Even suitably broken down, it needs appropriate stylesheet elements for the ordered list and dictionary list tags for readability.
20:53 thd russ: I have written a stylesheet for those tags and tested it.
20:54 russ ah right
20:54 thd russ: Without such a stylesheet, gazing upon it is uniewably painful :)
20:55 russ ok well we will just have to see how we get on with that when populating the content into
20:55 russ a seperate style sheet for that section may be a problem, may not
20:55 russ but I wouldnt worry about that for the moment
20:56 russ the most important thing is getting that content together.
20:56 russ we can sort out how it gets displayed and organised on later
20:56 russ gotta go - client is after a report i am writing
20:57 thd russ: It is only the list, ordered list, and dictionary tags that need special stylesheet treatment
21:01 thd russ: I have been translating some roadmap content from French.  The babelfish had reduced some posted 3.0 roadmap features to gibberish.
22:50 rach thd do you know about inventory reports?
01:42 thd rach: what about inventory reports?
04:23 thd good morning paul
04:23 paul (on phone)
04:23 thd paul: I have a question when you are off phone
05:01 thd paul: are you still on phone?
05:02 paul 58mn on phone...
05:02 hdl s/thiks/thinks/
05:02 paul i'm here now !
05:03 paul (on phone again...)
05:03 paul (unilim)
05:03 thd paul: I have seen your expert MARC editor diagram.  It looks very nice.
05:05 thd paul: It is designed as I expected it to be with an exception which I will explain when you are off phone again.
05:14 Sylvain ho
05:14 Sylvain hi
05:15 thd ho Sylvain
05:48 paul thd, i'm back
05:50 thd paul: I had not expected that the authorised value pop-up would have been part of you expert MARC editor implementation.
05:51 paul why ?
05:51 thd paul: However, I think a proper implementation of the full guided editor forms tool should be the best for every user including experts.
05:52 paul ???
05:53 thd paul: I think the guided forms with value lists and plugins for everything possible should be ideal for every user ultimately.
05:55 thd paul: However, I just thought the experts would be typing the authorised forms for authorities expertly without a tool to aid them.
05:56 thd paul: I prefer the pop-up for authorised values in the expert editor as you have represented it.
05:57 thd paul: However, there is a problem with the way the frameworks represent authorised values from authority records.
05:59 thd paul: Yes, I am saying that the help is good but there is a design defect in the help.
06:00 thd paul: The frameworks have no provision for designating field groupings for authorised values.
06:00 thd paul: consider the case of authorised personal names.
06:02 thd paul: I have noticed that implementations of UNIMARC Koha seem to use personal names in a manner that is a little closer to MARC 21 than UNIMARC.
06:03 thd paul: After lunch I will explain with an example.
06:09 thd s/field groupings/subfield groupings/
08:15 thd paul: consider the case of the personal names
08:17 thd paul: Examples I have seen in UNIMARC Koha look like MARC 21
08:17 thd paul: Sorry, do you mean that you are busy reading now?
08:17 paul no, I mean I'm reading you ;-)
08:19 thd paul: I find 700##$aSmith, John in UNIMARC Koha just as would expect to see it in MARC 21 Koha
08:19 paul except that you also should have a $3 with the authority number
08:21 thd paul: Yes there would be an authority number that is true but to an improper authority form.  The equivalent for MARC 21 Koha would be 100#1$aSmith, John.
08:22 thd paul: UNIMARC divides forenames from surnames by different subfields, not merely punctuation.
08:23 paul right.
08:23 paul smith => $a
08:23 paul john => $b
08:23 thd paul: exactly
08:23 paul + title (king, Sir, Lord...) in $dontrememberwhich
08:23 paul + birth/death dates in $f
08:23 thd paul: $c
08:23 paul + number in $d
08:24 paul (like King Louis XIV)
08:24 thd paul: exactly and the cumulation of all those parts is the authorised form of the name
08:27 thd paul: This allows the direct distinction of $700#1$aSmith$bJohn$f1840-1900 from $700#1$aSmith$bJohn$f1950-
08:28 thd paul: for the user visually as well as the system in $3
08:31 thd paul: The frameworks should support treating subfields as groups or whole fields collectively and filling the authorised form as a whole with the standard authorised form from BnF or wherever..
08:33 thd paul: frameworks should fill with 700 #0$aAlexandra,$cEmpress,$cConsort of Nicholas II, Emperor of Russia ; note the repeated $c and not merely 700 $aAlexandra
08:35 thd s/frameworks should fill/plugin pop-up should fill/
08:41 thd paul: do you have comment?
08:41 paul (on phone, sorry ;-) )
08:47 thd s/$aSmith$bJohn/$aSmith,$bJohn/g  # I had omitted the comma after Smith
08:58 hdl thd: comma automatically added after $cEmpress ?
09:01 hdl thd: Is this much troublesome to link with the appropriate authority but fill with only a summary of informations ?
09:04 hdl thd: I don't understand your change in Indicator 700#1 becomes 700#0
09:09 thd hdl: UNIMARC 700#1 for designates $a as a surname
09:09 thd hdl: UNIMARC 700#0 designates $a as a forename
09:10 hdl wow smileglasses
09:10 thd hdl: Alexandra is a forename
09:11 hdl thd: you really are a MARC professionnal are you not ?
09:13 thd hdl: The comma depends on your bibliographic cataloguing rules.  Remember I asked what is the name of the cataloguing rules used in France.
09:14 thd hdl: I am an amateur at everything :)
09:15 thd hdl: Although, I wish I had as much experience programming as I have cataloguing.
09:17 thd hdl: I have never worked as a cataloguer in a library only in books businesses where the important cataloguing issues are more usually descriptions of condition after copy cataloguing the rest of the information.
09:18 thd hdl: I learnt to care about the fine details of MARC while writing my own copy cataloguing system before Koha existed.
09:24 thd hdl: I do not understand this question.  " Is this much troublesome to link with the appropriate authority but fill with only a summary of informations ?"
09:26 hdl thd: Indeed, as you said, MARC21 uses a kind of summary field to display authority derived information, and UNIMARC does it differently.
09:27 hdl thd: if we follow MARC21 practise in UNIMARC, is this THAT horrible ?
09:27 thd hdl: Do you mean MARC 21 has no linkage to the authority record number as UNIMARC $3 provides?
09:29 hdl Errr...
09:29 thd hdl: Or do you mean putting the forename and the surname in $a as is the UNIMARC practise.
09:29 thd ?
09:29 hdl thd: yes, that latest assert
09:30 thd hdl: Then you do not have UNIMARC and your authorised names will not match the BnF authority file at the encoding level.
09:30 hdl thd: that's the spirit, except that you seemed to talk about MARC21 as  putting the forename and the surname in $a ... Or I mix up
09:31 hdl ok.
09:31 thd hdl: Yes MARC 21 puts the both the forename and surname in $a however UNIMARC does not.
09:32 thd hdl: This aspect may be minor since would be easy to add a $b with a script.
09:33 thd hdl: The issue that is important is that multiple subfields work together for the authorised form of the name.
09:35 thd hdl: Dates, etc. if present in the authorised form are important distinguishes so the user can use the same full name with dates to search another system for the exact same author.
09:35 thd s/distinguishes/distinguishers/
09:37 thd hdl: consider a MARC 21 example 100 1#$aChurchill, Winston,$cSir,$d1874-1965.
09:38 thd hdl: There are other Winston Churchill's who are less famous but wrote books.  Without the dates it is impossible to distinguish them from the bibliographic record.
09:40 thd hdl: MARC 21 has no $3 for authority record linking and even in UNIMARC the user searching another system has no meaningful access to $3.
09:40 hdl thd: are the other Winston also Sir ?
09:41 hdl ;)
09:41 hdl ok.
09:41 hdl I understand your point of view.
09:42 thd hdl: I do not remember.  Quite possibly.  Knighthoods are common but even still it is from another subfield.
09:44 thd hdl: There are actually quite a few Winston Churchills, more than one of which has published books.  Some are direct descendents or also ancestors of the best known Winston Churchill.
09:45 thd hdl: Another good example just for fun is 100 0#$aJohn Paul$bII,$cPope,$d1920-2004
09:49 thd hdl: There are certainly some John Pauls in the world where Paul is a surname.  And a previous Pope John Paul.  His given Polish name is not the authorised form in the NACO names database.
09:50 thd paul: My reference showed 1920- with no death date and I could not remember when he died.
09:50 paul 2005, april 2
09:50 paul (was the day my last son was baptised)
09:51 thd paul: I was feverish at the time with influenza.
09:51 thd paul: All my memories then are muddled.
09:54 thd paul: So what about the issue of treating groups of subfields with indicators for filling with authorised forms.
09:55 thd paul: UNIMARC 700 $4 could not be filled from an authority record but I think everything else could in 700.
09:56 thd paul: Should there not be some way to define such behaviour in an authority framework?
09:59 thd paul: If you copy catalogue from BnF you will at least have 700 $a and $b for most every case unless you concatenate them.
10:04 thd paul: The issues for subject subdivisions are a little different because subject headings usually have the subfields filled individually from the authorised heading in a subject authority record when cataloguing except when copying the subdivision usage from another bibliographic record as a whole.
10:05 thd paul: comments, or are you still on telephone?
10:06 hdl paul away at the doctor.
10:06 thd hdl: Have I made him ill?
10:07 hdl No. :D  had an appointment. and remembered late.;)
10:08 thd hdl: Koha will be very difficult to market in the US without addressing these issues.
10:09 hdl but sure, when we speak with you, it is both interesting but not very productive time for our own developpments.
10:10 thd hdl: I realise that but would not most important libraries in France have the same need?
10:11 hdl First, let's have Koha Safe and sound even with some lacks.
10:11 hdl Then let's fill in the gaps.
10:11 hdl Don't you agree ?
10:12 hdl But It IS interesting to speak with you.
10:13 thd hdl: Yes of course and these issues will certainly be easier for the OPAC in 3.0 but I had assumed they would also be addressed for the authority record at the same time.
10:15 thd hdl: I have a librarian in Illinois who had told me two weeks ago that she was willing to write a grant application for two full time programmers for a year to address these issues in Koha.
10:19 thd hdl: What do you do now about 700 $a and $b when copy cataloguing from BnF?
10:26 thd hdl: Why does ENSMP only have Koha for testing and not yet for production?

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