> And now, Alex, I must ask you how much
> experience you have (and not in a snarky, or evil way ;-)
Ahah, you're on to me! :) Do you mean in or outside the library? I've
got over 20 years experience, and I've done 4 years on the inside.
> The fact that data is in a valid XML format in no way guarantees its usefulness.
That it is not in a valid XML format, however, is a guarantee that it
isn't useful to most people and most tools.
> The problems of technology in libraries have almost nothing to do MARC as a format in and of itself. Bad metadata can and does come out of any system where bad data is entered.
Sure, but wouldn't it be great that when you design your big exchange
meta data format that you think about mechanisms for, eh, exchange?
> And, by the way, MARC XML is just as useful as any XML schema you can dream up to contain bib information.
I can only assume ... uh, no, scratch that. Not sure how to respond to
this, but, um, let's just say that I dream in XML, and I can dream up
stuff that is a magnitude more useful than MARCXML. People, you need
to understand just how evil MARCXML is. Maybe I need to write
something about this.
> MARC has been around since the 1960's. XML since the 1990's. For all practical purposes, libraries around the country had the great bulk of their collections cataloged as MARC *before* XML became a viable alternative.
Cars have been around since the turn of the last century, and still
they evolved. It's this evolution (or lack thereof) which is the
problem, not the history of it.
> The problem with simply converting from MARC to some as-yet-undetermined XML Schema is the need to create a real, agreed-upon standard.
As much as I used to think that, I don't now think this is the
problem. I'd much sooner point my finger at the incredibly poor meta
data you find inside the culture of MARC (where the complete lack of
typed data is glaring at me from all corners), the focus of that meta
data (why is it that I always find very good information about the
physical size of books, but never good information about what the book
is all about?) and plans to move forward without knowing how (how do
you, for example, put FRBR in MARC in a simple sane way? I've seen at
least one proposal that cause my spline to do hiccups).
> So, should all libraries simply convert to the Amazon standard? Perhaps, but, then the standard would have to be extended to contain all of the possible information that you can store in a MARC record - so, perhaps it's not that easy after all. Don't extend the Amazon format and discard the detailed information that MARC can contain? Why would we want to do that?
Well, if you already did XML and did it good, extending the XML is a
piece of cake. Sorry to say, but there's more to XML than simple
markup.
> In any case, Amazon is not more popular than library catalogs because it uses some format other than MARC. It is clear that the problem for libraries lies in the fact that they share no openly accessible central database of bib records, a la Amazon. OCLC ain't giving away its records, as we all know...
No argument there; open library meta data to the world, NOW! Or perish.
Smiles,
Alex
--
---------------------------------------------------------------------------
Project Wrangler, SOA, Information Alchemist, UX, RESTafarian, Topic Maps
------------------------------------------ http://shelter.nu/blog/ --------
Received on Wed Sep 24 2008 - 15:18:26 EDT