Re: Google Booksearch Data API: Another blow to library metadata

From: Jonathan Rochkind <rochkind_at_nyob>
Date: Mon, 29 Sep 2008 16:51:22 -0400
To: NGC4LIB_at_LISTSERV.ND.EDU
Hmm, I wonder if in internationalizing the display of labels for the 
data-elements, a "language" could be provided for "legacy MARC".  If 
instead of seeing "title proper" or it's name in German, you'd rather 
see 245$a, hey, that's a language that can be provided by 
"internationalization".

Of course, as these things continue to evolve, data elements will be 
created and modified so they no longer neccesarily correspond to how 
they are/were defined in MARC. I suppose new numbers could be made up 
for those, if some think it would be more friendly to non-English 
catalogers.  I dunno.

Jonathan

Diane I. Hillmann wrote:
> Bernhard Eversberg wrote:
>> From this element set to a MARC replacement will still be some way to
>> go. Or can you show us a sample record as it will look when it comes
>> over the wire? Which is what matters, at the end of the day.
>>
> The DCMI/RDA TG page referred to in my post has a number of cataloger 
> scenarios that do show what the resulting records will look like.  We 
> will be adding more XML and RDF encoded records in the next few 
> months, and hope that others will also experiment with outputs so that 
> we'll be able to show more samples.
>>
>> Definitions ok, but labels? Who needs German labels? If we cannot stick
>> with numeric labels (the only language-independent solution), then we
>> should beware of creating different label sets in so many languages.
>> For that purpose, English will have to do. And the briefer the better,
>> which is more difficult in German. We certainly do not need something
>> like
>>  <RDA:erscheinungsjahr>2008</RDA:erscheinungsjahr>
>> which gobbles up 10 times more bytes for wrapping than for content.
>> But don't establish new language barriers where we had none. Of course,
>> record content matters much more than punctuation, but if you don't get
>> the latter nailed down, machine communication of the former must fail.
>>
> Labels, Bernard, not names.  The strategy we're looking at would share 
> the same URI and unique name token (in English, the original language 
> of the vocabulary), with standardized labels for the use of consuming 
> applications (e.g., catalogs, browsers).  So catalogers would deal 
> with the English, users would see their language of choice 
> (hopefully), but machines wouldn't care one way or another.
>
> Diane
>

-- 
Jonathan Rochkind
Digital Services Software Engineer
The Sheridan Libraries
Johns Hopkins University
410.516.8886 
rochkind (at) jhu.edu
Received on Mon Sep 29 2008 - 15:13:24 EDT