Re: ISSUE-53: mediatypereg - suggest closing on 2009-09-03

On Aug 31, 2009, at 15:08, Julian Reschke wrote:

> Henri Sivonen wrote:
>> ...
>>> It might trigger observable processing in *some* recipients.
>> People or software?
> I meant software, but of course both is true.
>>> Are you trying to say that we should remove all "semantic"  
>>> elements if they do not have precise UA processing requirements?
>> I'm trying to say that "semantics" that don't trigger any  
>> observable effects in any class of UA are mere styling/scripting  
>> hooks, and mere styling/scripting hooks are an authoring-side  
>> convention--not a something that implementors of receiving software  
>> need to be concerned with beyond supporting generic styling/ 
>> scripting mechanisms.
> Just because they don't *need* to be concerned with it doesn't mean  
> that there is no semantics. And just because HTML5 doesn't define  
> processing requirements doesn't mean that some recipients will do  
> something specific with it.

So if the purpose of a MIME type registration is for recipient  
implementors to look up what they need to do to write receiving  
software, are you suggesting that the MIME type registration should  
support the case where an implementor looks up the meaning of text/ 
html and implements novel behaviors (i.e. ones not mentioned in the  
HTML5 spec) for obsolete language features by improvising from their  
previously-defined semantics? Is this a case that IETF/IANA rules  
require media type registrations to address?

I have a hard time understanding what part of this is wanting HTML5 to  
include the definitions of obsolete features from previous specs as a  
matter of independent principle and what part is strictly required for  
media type registration per IETF/IANA rules. That is, without chapter  
and verse from rules somewhere, I don't see why wanting semantics for  
obsolete features to be defined needs to be coupled with the media  
type registration issue blocking LC.

If the definitions aren't required by IETF/IANA rules for type  
registrations, the issue of definitions should be considered on its  
own right--not as a rider of the type registration.

Henri Sivonen

Received on Monday, 31 August 2009 12:37:57 UTC