Print

Print


<html>
Alvin - and others:<br>
<br>
I've been lurking behind this long discussion,<br>
hoping to jump in when the time was right<br>
(i.e. when I had a real question to ask)<br>
and wouldn't you know...fate lends a hand.<br>
It's fitting that the HREF example Alvin used<br>
is from a Yale instance<br>
(<a href="http://webpac.library.yale.edu/webpac-bin/wgbroker?new+-access+top.Yale_Lib+" eudora="autourl">http://webpac.library.yale.edu/webpac-bin/wgbroker?new+-access+top.Yale_Lib+</a><br>
search+open+ke+FHF5564)<br>
<br>
Over the past two days, I've been trying to conquer the<br>
demon named <i>external linking</i>.<br>
Last February, after a series of problems indexing EAD 1.0
instances,<br>
I decided that the path of least resistance would be to switch<br>
to the HREF method - and, sure enough, LiveLink liked <br>
this a lot better. The residual problem is that I can't get<br>
my browser (MultiDoc Pro - Am I the only one using this?<br>
Anyone?...Anyone?) to parse correctly an HTML link embedded in an<br>
SGML instance.<br>
<br>
Now the concern about confusing a union system via shorthand<br>
notation of entities is certainly worth noting.<br>
All the info. I've seen about XML seems to indicate that<br>
external linking is not a big concern, however.<br>
Is the only answer the one suggested by Alvin -<br>
to declare these links in every conceivable way,<br>
hoping that your indexing and browsing software will<br>
be able to interpret at least one of these methods correctly?<br>
<br>
Just wondering...<br>
<br>
Tim Young<br>
Beinecke Library<br>
Yale University<br>
<br>
but seriously- DO contact me if you use MDPro!<br>
<br>
At 11:40 AM 4/27/99 -0700, you wrote:<br>
&gt;&gt;In the example above, we would encode the EAD link as<br>
&gt;&gt;&lt;extref href=&quot;myentity.sgm&quot;/&gt;<br>
&gt;<br>
&gt;Michael, this has a profound impacy on how we interchange<br>
&gt;our finding aids and use them in union databases. If you<br>
&gt;are using &lt;archref&gt;, say, to link to another finding aid<br>
&gt;you cannot assume your &quot;myentity.sgm&quot; is a unique
filename<br>
&gt;in a union environment simply because it's unique within<br>
&gt;your own repository. In our Museums in the Online Archive<br>
&gt;of California some of our members are using delivery software<br>
&gt;which does not understand the redirection employed by the<br>
&gt;entity mechanism. We've solved this problem by using both<br>
&gt;an href attribute and an entityref attribute. Our union<br>
&gt;server, Dynaweb, ignores the href attribute which it cannot<br>
&gt;use in a union environment, and uses the entity mechanism<br>
&gt;instead. The individual institution will use the href as<br>
&gt;that's all its software can understand:<br>
&gt;<br>
&gt;&lt;extref href=&quot;myentity.sgm&quot;
entityref=&quot;myentity&quot;/&gt;<br>
&gt;<br>
&gt;It's redundant information to be sure, but the fault is<br>
&gt;not with SGML or XML, rather it's in that particular<br>
&gt;software package. Interchange is the most fundamentally<br>
&gt;important aspect of EAD and it's important not to compromise<br>
&gt;this. I don't think everybody should use this duplicative<br>
&gt;method, just those few repositories which are using delivery<br>
&gt;software which is underpar.<br>
&gt;<br>
&gt;Href is not a good attribute. There is no indication of the<br>
&gt;*kind* of resource it's pointing to. If you are converting<br>
&gt;to html for example, in one case you would be translating<br>
&gt;to an &lt;A HREF=&quot;&quot;&gt; tag for hyperlinks, and &lt;IMG
HREF=&quot;&quot;&gt; for<br>
&gt;images, &lt;APPLET&gt; for still more complex resources. You
face<br>
&gt;the same problem in Panorama stylesheets. The conversion or<br>
&gt;rendering software can only guess at what it should do. If<br>
&gt;the software is sufficiently sophisticated, it could parse the<br>
&gt;filename or url and make a decision based on the file<br>
&gt;extension (Panorama cannot do this):<br>
&gt;<br>
&gt;&lt;extptr href=&quot;somefile.gif&quot;/&gt;&nbsp; we can guess is
an image<br>
&gt;&lt;extptr href=&quot;somefile.html&quot;/&gt; we can guess is a
hyperlink<br>
&gt;<br>
&gt;But how about more complex urls, such as this one taken from<br>
&gt;an actual finding aid:<br>
&gt;<br>
&gt;<a href="http://webpac.library.yale.edu/webpac-bin/wgbroker?new+-access+top.Yale_Lib+" eudora="autourl">http://webpac.library.yale.edu/webpac-bin/wgbroker?new+-access+top.Yale_Lib+</a><br>
&gt;search+open+ke+FHF5564<br>
&gt;<br>
&gt;Href has a powerful attraction in that it is so familiar<br>
&gt;to people used to dealing with HTML. The redirection employed<br>
&gt;by entities is something people are not familiar with but<br>
&gt;it's the *right* way to do this kind of thing in SGML and XML<br>
&gt;and not for pedantic reasons only.<br>
&gt;<br>
&gt;Alvin Pollock<br>
&gt;Lead Programmer<br>
&gt;Online Archive of California<br>
&gt;<a href="http://sunsite2.berkeley.edu/oac" eudora="autourl">http://sunsite2.berkeley.edu/oac</a><br>
&gt; </html>