Re: [XBL] "Semantics"

Hi, Ian, WAF WG-

Here's a 2-for-1 special.

Ian Hickson wrote:
 > On Wed, 6 Dec 2006, Doug Schepers wrote:
 >> The word "semantic" is used several times with (as far as I can tell)
 >> at least 2 different meanings, but is not defined.  If you must
 >> use this term, please define it for the purposes of this document and
 >> stick to that definition.
 >
 > Defined.

Thanks.  The definition is clear, and your use of it seems to be pretty 
consistent within the document.

Using such an overloaded term, though, please consider changing the text 
a bit, something like this:
"In this specification, the term 'semantics' is used to refer to the 
processing model of elements, attributes, events, and DOM interface 
members, as defined by their indicated specifications; for example, this 
specification defines the semantics of XBL elements."

The phrase "intrinsic meaning" just opens up a huge can of worms, and I 
think that the specification is clearer without it.

Please read on... (not for those who don't want to read another blasted 
screed on semantics.)


Ian Hickson wrote:
 > On Mon, 8 Jan 2007, Jon Ferraiolo wrote:
 >> A couple years ago, I disliked the use of the term "semantics" in the
 >> sXBL spec (http://www.w3.org/TR/sXBL/#introduction, paragraph 3) and
 >> today I still dislike it in the XBL2 spec (3rd paragraph, still). I
 >> think it is too ideologic for a W3C spec.
 >
 > I don't understand. Why can't a W3C spec be ideologic?

I actually don't think that it's wrong for a W3C spec to be 
technologically ideological, as long as it is a shared ideology.  As 
hotly debated a term as "semantics" is unlikely to reflect the views of 
most or all of the membership of the W3C, which this specification 
represents.

I would encourage finding another term that meets your needs, or a more 
qualified term such as "processing semantics".


 >> clever developers might use XBL for other
 >> purposes, some of which might alter some of the "semantics".
 >
 > I don't understand how (short of having script actually go in and
 > mutate the DOM, of course). The whole point is that the
 > underlying meaning isn't changed, just like CSS doesn't
 > change the underlying meaning.

If I use XBL to change a list into the seeming of a paragraph, or a 
<rect> into a <circle>, I have not changed the way it is processed by 
the UA... but I have changed the way it is understood by the user.

Such content is still subject to the rules laid down in its spec, so 
it's accurate to limit the scope of its processing (as you've done), but 
not the manner in which it is used or the effect it achieves, and 
certainly not the meaning as interpreted by users.

My comment will be satisfied if wording functionally equivalent to my 
example at the top is used, and the term "semantics" is qualified in use.


 > The current state of the spec *is* the watered down version.
 > I don't want to water it down any further, it might become
 > homoeopathic! :-)

Okay, that was really pretty funny.  (True story:  A co-worker was 
chatting with her friend, who had to rush off to her homeopathy class. 
I advised her that if she was late, she could just swallow her watch. 
She didn't like that for some reason.)


-- 

Regards-
-Doug

Research and Standards Engineer
6th Sense Analytics
www.6thsenseanalytics.com
mobile: 919.824.5482

Received on Wednesday, 10 January 2007 10:10:11 UTC