<?xml version='1.0' encoding='utf-8' standalone='yes'?>
<?xml-stylesheet type='text/xsl' href='/style/atom-comments.xsl'?>
<feed xmlns='http://www.w3.org/2005/Atom'>
<title>norman.walsh.name: Comments on /2004/03/03/266NorthPleasant</title>
<link rel='alternate' type='text/html' href='http://norman.walsh.name/2004/03/03/266NorthPleasant'/>
<id>http://norman.walsh.name/2004/03/03/266NorthPleasant/comments.atom</id>
<updated>2004-09-09T21:48:37Z</updated>

<entry>
<title>Comment 0001 on /2004/03/03/266NorthPleasant</title>
<link rel='alternate' type='text/html' href='http://norman.walsh.name/2004/03/03/266NorthPleasant#comment0001'/>
<id>http://norman.walsh.name/2004/03/03/266NorthPleasant#comment0001</id>
<published>2004-03-05T16:15:45Z</published>
<updated>2004-03-05T16:15:45Z</updated>
<author>
  <name>Tobi </name>
  <uri>http://www.pinkjuice.com/</uri>
</author>
<content type='xhtml'><div xmlns="http://www.w3.org/1999/xhtml"><p>If the "URI becomes both its name and its address" then
http://example.org/path/to/book.xsd
should always be the same (otherwise it's a different name thus stands for a different thing).
No matter if you call it "canonical URL" or "URI", your catalog resolver will always fetch the same local copy of the resource identified via the URI [1]. If it doesn't find a local copy, it can try to load it from the URL (the same string as the URI). One string can be a URI when used as identifier/name, or as a URL when used as locator. (but TBL disagrees ...)</p>
<p>All this is not simple, but the fact that one string can serve as URI (eg a namespace name, or resource name used for fetching a local copy) and also serves as URL (eg to download the resource) is actually not a problem AFAICS, but can be quite convenient and useful, if everyone participating is aware of that fact.</p>
<p>The problems arise when people make up identifiers by filling in some local path which isn't the name of the resource; you can't know what their made up name refers to. It's as if everyone would make up new names for everything; they shouldn't be surprised if noone understands them anymore.</p>
<p>[1]
Just like SGML "implementations made use of the fact that they could map from the name, the public identifier, to the appropriate local representation."</p></div></content>
</entry>

<entry xmlns:foaf='http://xmlns.com/foaf/0.1/'>
<title>Comment 0002 on /2004/03/03/266NorthPleasant</title>
<link rel='alternate' type='text/html' href='http://norman.walsh.name/2004/03/03/266NorthPleasant#comment0002'/>
<id>http://norman.walsh.name/2004/03/03/266NorthPleasant#comment0002</id>
<published>2004-03-08T21:14:00Z</published>
<updated>2004-03-08T21:14:00Z</updated>
<author>
  <name>Dan Connolly</name>
  <foaf:mbox_sha1sum>94b6eb0c835f928c5ed565dc3ed1a355ac1b41e5</foaf:mbox_sha1sum>
  <uri>http://www.w3.org/People/Connolly/</uri>
</author>
<content type='xhtml'><div xmlns="http://www.w3.org/1999/xhtml"><p>'266 North Pleasant Street' isn't a name for you; it's a name for your house. You're related to, but not the same as, your house.</p>
<p>On the web, http://norman.walsh.name/ is both a name and a location for your web site. Again, it's related to you, but it's not you.</p>
<p>You seem to use http://norman.walsh.name/knows/who#norman-walsh to identify yourself. On the web,
that's your name. Or... one of your
names, anyway.</p></div></content>
</entry>

<entry xmlns:foaf='http://xmlns.com/foaf/0.1/'>
<title>Comment 0003 on /2004/03/03/266NorthPleasant</title>
<link rel='alternate' type='text/html' href='http://norman.walsh.name/2004/03/03/266NorthPleasant#comment0003'/>
<id>http://norman.walsh.name/2004/03/03/266NorthPleasant#comment0003</id>
<published>2004-03-09T08:52:43Z</published>
<updated>2004-03-09T08:52:43Z</updated>
<author>
  <name>Norman Walsh</name>
  <foaf:mbox_sha1sum>9f5c771a25733700b2f96af4f8e6f35c9b0ad327</foaf:mbox_sha1sum>
  <uri>http://norman.walsh.name/</uri>
</author>
<content type='xhtml'><div xmlns="http://www.w3.org/1999/xhtml"><p>The alternatives you suggest, Dan, are all addresses. They're also names. I've argued in the past that my name is not my address, and the point of the title of this essay is that I'm abandoning that line of argument. Names and addresses are the same thing on the web.</p>
<p>But you're right, in the semantic web context, I use http://norman.walsh.name/knows/who#norman-walsh  as my canonical name.</p>
<p>The fact that, as a practical matter, I sometimes wish I'd used http://norman.walsh.name/knows/who/norman-walsh is a topic for another essay.</p></div></content>
</entry>

</feed>
