W3C home > Mailing lists > Public > uri@w3.org > October 2006

Template URIs vs. URIs

From: Mark Baker <distobj@acm.org>
Date: Thu, 5 Oct 2006 14:27:18 -0400
Message-ID: <c70bc85d0610051127l271eacffjd77d302ce58ab5c@mail.gmail.com>
To: uri@w3.org

Regarding http://www.ietf.org/internet-drafts/draft-gregorio-uritemplate-00.txt

My concern with this draft is that template URIs are not easily
distinguishable from URIs.  mnot's blog entry[1] claims that the use
of "{}" is sufficient to distinguish the two, and while that may be
technically true, I'm not aware of any URI-detection algorithm that
checks for the presence of minor, unknown characters.  A quick test of
GMail shows that it presents template URIs as clickable URIs, for
example.

I'd recommend that the template URI syntax use a more significant
change to avoid this confusion.  Starting with a non-scheme character
("@"?) might be one way.  Or, given how detection algorithms usually
work, replacing the colon after the scheme with another non-scheme
character (or otherwise morphing it) would probably work.  Any
solution that helps make it more obvious to humans and (existing)
automata that template URIs are not URIs would be fine by me.

Thanks.

 [1] http://www.mnot.net/blog/2006/10/04/uri_templating

Mark.
Received on Thursday, 5 October 2006 18:27:23 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 13 January 2011 12:15:36 GMT