<html>
<body>
<br>
Yes, the specification in MRAC 21 will be open-ended about the wort of
identifier that may be included. This will parallel the
specification in other fields where subfield $u has been defined.
MARC doesn't try to restrict usage.<br><br>
However, as I pointed out in my last message, I think that records that
we contribute to shared cataloging databases should use links that are as
widely applicable as possible -- to publicly available resources through
persistent and public identifiers. However, that is an application
guideline, not something that would be in the MARC
specifications.<br><br>
One thing that occurred to be after I sent my last message was the
sequence of subfields. In Elizabeth's example, the URL is for the
Evans bibliography as a whole and so it comes between subfield $a and
$c. However, I could imagine a case in which the bibliography is
available online and the URL could point to the particular item; in that
case, the subfield $u would follow $c. Can anyone think of an
example of this that doesn't involve a proprietary database? I'd
like to include an example in the MARC change proposal, and I don't want
to include something that requires a subscription to access.<br><br>
<x-tab> </x-tab>John<br>
<br>
At 04:52 PM 9/17/2008, Elizabeth Robinson wrote:<br>
<blockquote type=cite class=cite cite="">Part of this afternoon, I've
been reading the FAQ on the LCCN Permalink (see
<a href="http://lccn.loc.gov/lccnperm-faq.html">
http://lccn.loc.gov/lccnperm-faq.html</a>). While the 510$u's URL doesn't
have to be the LCCN Permalink, I think it might be simpler to reference
it since it should be stable. Notice that the permalink program searches
both $a and $z of the 010.<br>
<br>
You are right though, Nina, that there will be resources that even LC
won't have (or won't have an online bib record for). The latter is very
real. (No, we are not done with recon!) So in those cases, it would have
to be some other institution's record.<br>
<br>
I suppose the instruction for the 510$u could be left open ended, and
those who chose to use the LCCN Permalink could, and those who chose to
use their institution's (or another institution's) bib record of the
resource could do that too. <br>
<br>
Can anyone think of any unforeseen problem with this?<br><br>
--- On <b>Wed, 9/17/08, Schneider, Nina
<i><nschneider@humnet.ucla.edu></i></b> wrote:<br>
<dl>
<dd>From: Schneider, Nina <nschneider@humnet.ucla.edu><br>
<dd>Subject: Re: [DCRM-L] ESTC and the revision of SCF<br>
<dd>To: ElizRob@alum.emory.edu, "DCRM Revision Group List"
<dcrm-l@lib.byu.edu><br>
<dd>Date: Wednesday, September 17, 2008, 4:19 PM<br><br>
<dd><font face="Arial, Helvetica" size=2 color="#0000FF">I like it.
Here's a question, though: Does it matter what URL we provide in the 510
$u? In other words, it might be preferable to include the link to my
institution's copy of Evans rather than LC's copy. But then what happens
if my institution doesn't have a copy of the reference? <br>
</font>
<dd> <br>
<dd><font face="Arial, Helvetica" size=2 color="#0000FF">I realize this
is in its preliminary stage, but it's something to consider.</font>
</dl></blockquote></body>
</html>