RE: don't create new URI schemes just for namespaces

["Hollenbeck, Scott" <shollenbeck@xxxxxxxxxxxx>]

(Trimming the cc list...)

Since I've already sent the guidelines document to the IESG I'll capture
this as a comment for their last call.

-Scott-

> -----Original Message-----
> From: Martin Duerst [mailto:duerst@xxxxxx]
> Sent: Sunday, July 21, 2002 12:42 PM
> To: Julian Reschke; uri@xxxxxx; ietf-xml-use@xxxxxxx
> Subject: don't create new URI schemes just for namespaces
> 
> 
> 
> I fully agree with Julian. I just have seen some other such examples
> in the jabber protocol at the IETF BOF, and have pointed this out
> as a mistake.
> 
> I propose that we add something like this to the XML guidelines
> (http://www.ietf.org/internet-drafts/draft-hollenbeck-ietf-xml
> -guidelines-05. 
> txt), just before 4.8.1:
> 
> "Although there are some examples where this has been done in the
> past, specifications MUST NOT create new URI schemes to 
> define namespaces."
> 
> Regards,     Martin.
> 
> 
> At 12:26 02/07/21 +0200, Julian Reschke wrote:
> 
> >Sadly, many projects are following WebDAV's example of 
> defining a special
> >URI scheme just for the purpose of identiying their 
> namespaces. In the case
> >of WebDAV (RFC2518), the scheme was at least registered, and 
> the official
> >errata document recognizes this as a valid issue with the protocol.
> >
> >Subversion (http://subversion.tigris.org) sadly is doing the 
> same thing --
> >it defines a new URI scheme ("svn:") for internal 
> identifiers which now
> >shows up as XML namespace both in internal files (entries files for
> >instance) and in external marshalling (WebDAV 
> request/response bodies). It
> >would be nice if this would be fixed before they reach 1.0. 
> The issue has
> >certainly been raised but was rejected by their respective 
> developers.
> >
> >Julian
> >
>