Re: ALA Session on MODS and MADS: Current implementations and future directions

From: Ashley Sanders <a.sanders_at_nyob>
Date: Tue, 22 Jun 2010 15:06:28 +0100
To: NGC4LIB_at_LISTSERV.ND.EDU
Hi,

>> What can COPAC do, thanks to MODS, that a MARC-based catalog couldn't?
>>  From visiting it, I'm at a loss to answer that question, but I may have
>> failed to look hard enough. (The granularity of MODS, to mention one
>> point, is by design inferior to MARC's.)

First of all let me say that, in my opinion, what Copac can do is
independent of the format we store the data in...

We weren't allowed to store MARC records because of licensing
issue. We needed to store the record in some format and rather than
invent our format own we used MODS (with our extension for local
holdings.)

But, we don't index the MODS. The MODS is only used to drive the
displays (through XSLT) and provide a download format. Other
download formats are created from the MODS with XSLT.

So, MODS is easy to work with through XSLT and it saved us a load of
hassle coming up with our own format.

 >> (The granularity of MODS, to mention one point, is by design 
inferior to MARC's.)

I'm currently working an a new iteration of Copac in which we do
index a bunch of XML. We decided that the granularity of MODS was too 
fine for indexing and so we've come up with our own "simplified MODS" 
with less granularity. (Basically we've removed some tags, simplified
the hierarchy and dropped a load of attributes.) There'll still be a
MODS record in there as well, but again it'll only be used with XSLT
for driving the displays and export formats.

We also run a "MARC" database for cataloguing purposes. It does
store records in a format with a resemblance to MARC exchange, but
again it is only used to drive the displays and download formats.
The indexing is done separately.

(I've just spent a happy day writing code to translate the
coded fields in the Leader, 008 and 007 into human readable
genre keywords. It is far too easy to get that code wrong and
produce the wrong keyword. It is way more complicated that it
needs to be -- for the creator of the data and the poor sop
who has to process it. I guess all these coded fields were
invented when disk space cost real money. There's really no
need nowadays.)

Regards,

Ashley.
-- 
Ashley Sanders               a.sanders_at_manchester.ac.uk
Copac http://copac.ac.uk A Mimas service funded by JISC
Received on Tue Jun 22 2010 - 10:07:30 EDT