W3C home > Mailing lists > Public > public-html@w3.org > September 2009

Re: ISSUE-81 (resource vs representation)

From: Nikunj R. Mehta <nikunj.mehta@oracle.com>
Date: Mon, 28 Sep 2009 12:03:46 -0700
Cc: "public-html@w3.org" <public-html@w3.org>
Message-Id: <CA3645F7-1BC2-417F-BB2E-73A77FC35919@oracle.com>
To: Larry Masinter <masinter@adobe.com>

On Sep 27, 2009, at 7:16 PM, Larry Masinter wrote:

> Folks, some of this discussion isn't too helpful.
> The serious technical issue is that what HTML5 calls a "URL"
> allows (MAY) recommends (SHOULD) or mandates (MUST) doesn't
> match syntactically or semantically what other Internet
> applications call variously URLs, URIs, IRIs, LEIRIs, etc.,
> either syntactically or semantically, and that the mismatch
> in semantics will cause interoperability problems when
> copying or moving or translating between web applications
> and the other applications.

I was only aware of one issue being discussed here - whether a  
resource is itself a bag of bits or a thing that sends a bag of bits  
when requested. The issue you cite here is separate from the one I  
reported and the one Julian started a discussion about.

> Fixing this is hard; I think the only direction that
> makes sense to me is to mandate that compliant producers
> of URLs (or whatever they're called) be more conservative
> than what browsers currently accept.
> In some cases, the mismatch can be accomplished by saying
> that browsers do "pre-processing", e.g., strip off leading
> and trailing spaces.
> I'm still working on producing a new version of a
> document that captures that sense (on
> http://larry.masinter.net/iribis-hack.html), but
> the next version I was hoping to have ready by tomorrow
> (Monday) might take a little longer, now estimated
> Tuesday.

Just to clarify, neither mine nor Julian's issue is not about the  
syntax or meaning of URIs.

> Once we've really resolved the technical interoperability
> issue of how to get the systems to work together, then
> deciding what these things are named or called or described
> can be much more easily resolved.  I took a cut at that
> in my current IRIBIS draft, but I think I can do better.
> Of course, the provocative style of the HTML5 document
> and its irreverence toward existing other technical standards
> is annoying, but we should all be used to that by now and
> just get over it, at least until resolve the technical
> incompatibility issues.

Received on Monday, 28 September 2009 19:06:34 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:15:51 UTC