On Fri, Apr 16, 2010 at 1:31 PM, MJ Suhonos <mj_at_suhonos.ca> wrote:
> >> get a sense of what MARCXML is, and why it's not very good
> >
> > I have to disagree that MARC-XML is "not very good."
> >
> > It's actually quite useful as an *intermediary* format between (ISO 2709)
> MARC and other schemas, like MODS, Dublin Core, and so on.
>
Boy, I have to disagree again.
Let's start by saying there's no difference semantic between binary MARC and
MARC-XML. They're identical (i.e., round-trip-able) serializations of the
exact same data structure. Serializations are not "good" or "bad", they just
range from "less painful to work with" to "more painful to work with."
Having said that -- MARC-XML is demonstrably *better* than ISO 2709 in that
while every binary (ISO 2709) MARC record (at least those with "normal"
indicators, tags, etc) is easily represented in MARC-XML, the reverse is not
true because of the length issue. I've got plenty of records longer than
10,000 bytes in length, and for them binary MARC is worthless.
The MARC data structure (tags, indicators, subfields) is crappy, but that's
independent of its serialization. I've written up something about this at
http://robotlibrarian.billdueber.com/data-structures-and-serializations/
>
> > Trying to go from a binary MARC file *directly* to MODS, for example,
> would be a pain. So much easier to get the MARC into XML first, and then
> use XSLT to go from there to the schema of your choice.
>
> I agree with you 100% Dave, and that's precisely what I use it for myself.
>
> But as a standalone format, it's *not* very good. I didn't say "it's not
> very useful". ;-)
>
> MJ
>
> >
> > --Dave
> >
> > ==================
> > David Walker
> > Library Web Services Manager
> > California State University
> > http://xerxes.calstate.edu
> > ________________________________________
> > From: Next generation catalogs for libraries [NGC4LIB_at_LISTSERV.ND.EDU]
> On Behalf Of MJ Suhonos [mj_at_SUHONOS.CA]
> > Sent: Friday, April 16, 2010 8:16 AM
> > To: NGC4LIB_at_LISTSERV.ND.EDU
> > Subject: Re: [NGC4LIB] What do I need to know?!
> >
> > I'd suggest trying to get a grasp on XML as a starting point. Some will
> undoubtedly say that's not far enough, or that it's irrelevant next to the
> larger concepts like linked data, but really, XML represents the kind of
> web-world thinking that's going to power most of the implementations we see
> in the near future (as well as today, obviously).
> >
> > If you're a "traditional" cataloguer and familiar with Library of
> Congress stuff, then I might suggest starting there (
> http://www.loc.gov/standards/) — get a sense of what MARCXML is, and why
> it's not very good, despite being XML. Have a look at MODS and get an
> understanding of what a record in MODS looks like compared to what it would
> in MARC or MARCXML. Then, have a look at EAD and/or the Dublin Core
> Metadata Element Set (search on DCMES) — try to understand how
> non-librarians would use these schemas to describe a (maybe digital or
> archival) resource.
> >
> > That's just a starting point, but understanding the above should give you
> a better foundation to move on to other concepts like FRBR, RDA, RDF, Linked
> Data, etc.
> >
> > MJ
> >
> > On 2010-04-16, at 11:02 AM, Simpson, Sarah wrote:
> >
> >> I have to admit, I always hesitate to post a question of this sort to a
> list, for a variety of reasons. However, as I try to figure out how I'm
> going to contribute to the cataloging world, honestly, you just seem like
> the best people to ask.
> >>
> >> I'm a very good cataloger (and modest, too, as you can tell). However,
> I am becoming very aware that in order to remain a good cataloger and make
> any forward progress, I am going to need to know other things than
> cataloging. Specifically, technology. I will need to learn more about the
> workings of our catalogs, what we know and need to tell our customers about
> the digital resources we have, what is necessary to share our data with the
> Web, and so on.
> >>
> >> So, where do I start? Are there particular things to learn about
> technology that would be the most useful, particular classes, particular
> topics? What software do I need to know? I want to learn what I need to
> know, but just have no idea where to start! I have 3 universities and a
> community college in my immediate area, but do I need a computer science
> degree, or can I be more targeted than that?
> >>
> >> I want to know why RDA is a good or bad idea, I want to be able to tell
> my library how big of a project it will be to implement, and to tell the
> truth, I want the background to be part of the discussion as RDA continues
> to be modified and improved. ANYTHING you can tell me about how to learn
> what I need to know and what it is I need to know will be very, very much
> appreciated.
> >>
> >> Sarah Simpson
> >> Tulsa City-County Library
>
--
Bill Dueber
Library Systems Programmer
University of Michigan Library
Received on Tue Apr 20 2010 - 19:47:28 EDT