Print

Print


Let's not forget that the impetus to move to a new data format is not the data format itself but the need to encode a new kind of data. BIBFRAME arose out of tests that showed that it was not possible to produce true RDA data using MARC as its carrier. We should stop focusing on the data format and look more at what we want to do with our data. The data format is not holding us back more than our inability to imagine data that isn't based on left-anchored headings and transcribed title pages.

kc



On 3/6/15 6:32 AM, [log in to unmask] wrote:
[log in to unmask]" type="cite">

all. What I am saying is that for the library community, that is, the people who already know and understand--and even control--MARC format, changing the format they already control to Bibframe will not give them any new capabilities over what they have been able to do with MARCXML.

That's especially true as long as librarians continue to conflate "format" and "model", as you just did.
 
*Librarians* understand the MARC codes and that means they can work with MARCXML to fold in their records with what else exists on the Internet; they can do that now, and they've been able to do it for awhile.

Only by consuming. Library data is difficult to contribute to larger efforts. An option with Linked Data will be to contribute as well as consume.
 
So, if we can already do anything and haven't, the obvious question is: why will anything change with Bibframe/RDF?

Because you _can't_ do _anything_. You've repeatedly conflated what is possible with what is practical. Sure, MARC can be used, in theory, for any digital processing task. So what? One can make the same remark about any computerized representation of your data. The relevant question is what can you _practically_ do, with real technology and real budgets, and as Ross Singer hinted, the answer is: not much more than is now being done.
 
I wrote that message on Autocat to combat the popular idea that the reason libraries haven't done anything new or interesting is because of the limitations of the format. That was true until MARCXML arrived and then it became possible to do all sorts of new things. MARCXML may be nasty and difficult to work with, but no matter: if somebody wants to, it *can* be worked with *within the library community*.

If someone wants to, and has unlimited time, money, and technical support. The library community has none of those three things.
 
And people have worked with it, such as we see in catalogs that utilize Lucene indexing (which is based on MARCXML) to create the facets we see in different library catalogs. (That is one thing that has been done in the last 20 years, and it is due to XML)

This is flatly false. Lucene faceting (faceting in general) has no relationship with XML, and especially not with MARCXML. I can't imagine where you got this misinformation.

I gave the example of printing day-glo colors merely to emphasize that we can currently do anything we want right now, but of course, I was not suggesting we should waste our time on that. I want to try to open people's minds to what *can* be possible. *Anything* is a tremendous concept that is difficult to grasp.
 
That's because it has almost no meaning in this context. I can do anything that can be done with structured data with library data as soon as it obeys cataloging rules. That means nothing. A more interesting question is: what can I _afford_ to do? The answer to that question has everything to do with the marginal costs of processing library data. In its current forms, that cost is extraordinarily high compared to what it would be in modern forms.
 
Once we accept and begin to comprehend the idea that "anything can be done" the question of what would be better, or worse, uses of our labor and resources becomes far more complex and takes on different subtleties. Those who believe that the problems we have faced are because of the *format* so therefore, the solution is to get a "better format" and things will then be solved, will be sadly disillusioned.

No one has made such a simple-minded suggestion. Many people have suggested that removing some of the empty overhead costs of obsolete technology would free up resources to do more interesting things. Some people (I think of Karen Coyle here) have repeatedly requested that the library community begin this process by considering the end-user requirements for its products. That seems to have gone nowhere.

I'm sorry if this message seems a bit brusque, but some of the misconceptions in this conversation ("Linked Data is a format.", "There are no empty costs in the current system.", "Technologists think RDF is a magic bullet.") should have gone away a long time ago.
 
---
A. Soroka
the University of Virginia Library

-- 
Karen Coyle
[log in to unmask] http://kcoyle.net
m: +1-510-435-8234
skype: kcoylenet/+1-510-984-3600