W3C home > Mailing lists > Public > public-html@w3.org > June 2008

Re: heads-up about "new" URLs section in HTML5 editor's draft

From: Robert J Burns <rob@robburns.com>
Date: Mon, 30 Jun 2008 12:25:48 +0300
Cc: "Roy T. Fielding" <fielding@gbiv.com>, public-html@w3.org
Message-Id: <0FBA7EE4-DF78-4ADE-B29F-D530DCD0A375@robburns.com>
To: Ian Hickson <ian@hixie.ch>
Hi Ian,

On Jun 30, 2008, at 8:25 AM, Ian Hickson wrote:

>
> On Sun, 29 Jun 2008, Roy T. Fielding wrote:
>>
>> Please note that we spent 12 years reaching a global agreement on the
>> meaning of URI, URL, URN, Web addresses, or whatever else you might  
>> call
>> them, in order for all implementations to be interoperable and for  
>> all
>> protocols to obey the same restrictions on generation of those
>> identifiers.  The result is IETF STD 66, RFC 3986, and it defines the
>> most important standard of all the standards that make up what we  
>> call
>> the Web.
>
> With all due respect, if the goal was for all implementations to be
> interoperable, then it failed.

You can't have it both ways here Ian. You claim you're selecting the  
one true way that the web actually works and we can't tell  
implementations what to do. Now you claim here that there is no  
interoperability and thus there is not already ONE TRUE WAY. You're  
doing the same thing here that you're doing with parsing. Its hard to  
imagine this duplicity is not intentional since it comes up repeatedly

>> I suggest that the section be removed or replaced with the limited  
>> and
>> specific needs for parsing href and src attribute values such that  
>> the
>> attribute's value string is mapped to a URI-reference with a defined
>> base-URI.
>
> How is that different from what the spec does now?

The spec states explicitly:
> Note: The term "URL" in this specification is used in a manner  
> distinct from the precise technical meaning it is given in RFC 3986.  
> Readers familiar with that RFC will find it easier to read this  
> specification if they pretend the term "URL" as used herein is  
> really called something else altogether.

Perhaps error recovery and document conformance norms are not enough.  
I do think there may be a need to round out the matrix of definitions  
provided by RFC 3986 and RFC 3987.

                  Names     Locators   All Identifiers
ASCII       URN           URL               URI
l18n          IRN             IRL                 IRI


In which case the only new terms that need to be introduced is an IRL  
which is simply an identifier among the locator subsets of an RFC 3987  
IRI. This is really only an extension of the URN and URL definitions  
already in RFC 3986 and applying those to RFC 3987.

Take care,
Rob

issue-56
Received on Monday, 30 June 2008 09:26:30 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:16:18 GMT