Re: HTML5 object element - classid or not classid?

Ian Hickson:

>
> The only way for an <object> element in an HTML5 document to have the
> classid="" attribute at the moment is for the doucment in question to be
> non-conforming.
>
> HTH,

Ok, would be fine, if the definition section would reflect this ;o)
As already explained, it is bad style to talk about an attribute 
not defined at all (would be maybe enough to reference the HTML4
section with an additional comment, that this attribute is considered 
to be outdated and non-conforming in HTML5, therefore authors 
should not use it) and to explain in the definition section of the
HTML5 object element, what happens if it is used anyway.
Else if someone mentions to a 'copy and paste' author, that this
is non-conforming in HTML5, such an author will point to exactly
this part of the object definition to show, that it is defined for
object and therefore still a desirable use case for HTML5, 
leading to end- and useless further discussions, thousends of
readers looking, where this attribute is defined, the section
is talking about to decide themselves, whether this should be
used or not or what to do with this.
Maybe the strange use of several attributes of object on 
many webpages resulted from poor/vague descriptions in the 
object chapter of HTML4 and corrupted therefore a lot
of content in the web. It would be an improvement,
if this could be avoided somehow in HTML5.
 






Boris Zbarsky (about the declare attribute in HTML4,
currently missing in HTML5):

>>> Probably because no one actually implements it?
>> 
>> Why not? 

>Sorry, I can't read minds very well.  Especially when there is temporal 
>separation involved.  You'll have to ask the implementors from 10 years 
>ago or so.

Well, this is not really a historical question, because HTML4.01, 
XHTML1.0 and XHTML1.1 are the current recommendations for
HTML and HTML5 is still an early draft, there should be enough
time, maybe several years for implementors to implement more
features from HTML4.01 as they currently have - therefore it is still
a question at least for several content authors why this is not even
implemented in current state of the art and commonly used 
browsers, because documents claiming to follow the current
recommendations will be still in use for many more years and
many new documents will be written following these versions and
not HTML5 in the next years (assuming that they understand 
the HTML5 draft/future recommendation, else 'HTML4 style' will
be alive much longer).
But ok, it is more a question to implementors, not to everybody in 
the list ;o)

Received on Tuesday, 5 August 2008 14:54:13 UTC