Re: html4 vs html5 and "superseded RECs" when there isn't a new REC

Hi Wayne,

I think we are in closer agreement than we seem...

On Wed, 07 Mar 2012 13:03:03 +0100, Carr, Wayne <wayne.carr@intel.com>  
wrote:

> My original post was not issue 2 (what I was asking for was snipped out  
> in what's below).   HTML4 isn't a superseded REC because html5 isn't a  
> REC.   That was the point.  But, HTML5 is defacto the new html spec and  
> it will take a very long time to get to REC even though it is widely  
> implemented.  We should do something in the html4 REC to at least point  
> to html5.

Yeah, issue 2 is meant to cover that case, the case where there is a  
published working draft but the editor's draft is more useful, and the  
case where a REC has been superseded. We could split them if you think  
that's worthwhile (I think the last of those is pretty much a no brainer  
if we figure out the rest, and is reasonably well-handled anyway).

> The suggestion was to put a note in the HTML4 spec pointing to the  
> latest html5 TR WD.  (and keeping that up to date is a different issue -  
> but it's better than not indicating anything in the old and out of date  
> official REC).

Yep. And my suggestion is that while that is something that could be  
helpful, it is probably insufficient for people trying to understand what  
they are actually getting. Essentially for a family like the (X)HTML  
specifications I think there is a real need for an "intro to the HTML  
family" that is maintained, and describes the status of different things.  
Something like:

HTML4.01 is a REC but large parts of it are considered under-specced or  
even incorrect, HTML 3.2 and 4 are simply obsolete, XHTML 1 and 1.1 are  
RECs that are sort of complementary but based on HTML4, XHTML2 has been  
abandoned and HTML5 is the version where current work is going on and  
while unfinished it contains largely stabilised parts including important  
technology not described elsewhere, and unstable parts - see WD or  
Editor's Drafts according to your requirements

(with links, info about changelogs and so on, and a pointer to or note  
about the different kinds of document, and why you shouldn't link to a  
dated WD in most cases and often don't want a 2-year contract to depend on  
the "latest editor's draft" but do want to see that for the hot news, etc.)

cheers

Chaals

>> -----Original Message-----
>> From: Charles McCathieNevile [mailto:chaals@opera.com]
>> Sent: Wednesday, March 07, 2012 3:05 AM
>> To: Carr, Wayne; Marcos Caceres
>> Cc: public-w3process@w3.org
>> Subject: Re: html4 vs html5 and "superseded RECs" when there isn't a  
>> new REC
>>
>> Note - this is ISSUE-2 - if you add that string in mails about the  
>> topic they will be
>> collected at
>> http://www.w3.org/community/w3process/track/issues/2
>>
>> On Wed, 07 Mar 2012 11:48:35 +0100, Marcos Caceres <w3c@marcosc.com>
>> wrote:
>>
>>> On Wednesday, 7 March 2012 at 10:35, Carr, Wayne wrote:
>>>> With HTML4 we have the formal W3C Recommendation for HTML, but no one
>>>> would want that as the basis for a UA. It has been superseded by  
>>>> HTML5.
>>>> But, people still write that html4 is the html spec and html5 is in
>>>> the future. Obviously, that doesn’t reflect the Web Browsers out
>>>> there which have embraced html5. But html5 is still years away from
>>>> going to REC.
>>>>
>>>> It seems in this (hopefully) unusual situation,
>>>
>>> Just wondering, was it not the same case with CSS 2.1 (and now with
>>> many modules of CSS3)? And with XHR 1 and XHR level 2 also (with
>>> neither XHR version going to REC, while XHR 1 was abandoned in favour
>>> of just "XHR"…)
>>
>> Yeah, I think the situation is actually quite common.
>>
>>> XHR level 2 was originally published under:
>>> http://www.w3.org/TR/XMLHttpRequest2/
>> ...
>>> What would be nice would be if HTML5 would dethrone the obsolete XHTML
>>> spec from the URI (and we dropped the "5" from the spec, as the WHATWG
>>> has done):
>>> http://www.w3.org/TR/html/
>>
>> Actually, I think that is a far too simplistic approach to be useful,  
>> except to a small
>> handful of the stakeholders. And I think removing the version token  
>> from the spec
>> is not a helpful solution, as nice as it might feel.
>>
>> Having a pointer that describes the status of HTML at
>> http://www.w3.org/TR/html/ might be useful - and it might need to be  
>> more
>> complex than "latest draft, last 'heartbeat' publication, last  
>> Recommendation".
>> There are plenty of people who use XHTML. There are people who use HTML  
>> 4.
>> There are people who use HTML5. And there are people who use HTML and  
>> don't
>> actually know what they are using. If they want to look up "What is  
>> HTML" and
>> they expect W3C (instead of Wikipedia, about.com, or their local  
>> library's CD-
>> ROM encyclopedia of the Web) to explain it, we should think about how to
>> provide different bookmarkable references for the different needs.
>>
>> (Which also feeds into the bibliographic references discussion Marcos  
>> kicked off a
>> while ago. Did that result in any issues, actions, or
>> conclusions?)
>>
>> cheers
>>
>> Chaals
>>
>> --
>> Charles 'chaals' McCathieNevile  Opera Software, Standards Group
>>     je parle français -- hablo español -- jeg kan litt norsk
>> http://my.opera.com/chaals       Try Opera: http://www.opera.com


-- 
Charles 'chaals' McCathieNevile  Opera Software, Standards Group
     je parle français -- hablo español -- jeg kan litt norsk
http://my.opera.com/chaals       Try Opera: http://www.opera.com

Received on Wednesday, 7 March 2012 12:48:39 UTC