Re: Shelby's Final Position Paper on XBL

Hello Ian, dear list members,

Am Montag, 6. Januar 2003 18:04 schrieb Ian Hickson:
> As you say, we (including myself) don't have enough experience to be
> sure of anything here. However, I personally find that "Behaviour" is
> more towards the Presentation side of your layer model than the
> others.
I agree.
If the behaviour of a form is behaviour, it's somehow also related to 
semantics, not just style, so yes, I would not consider behaviour to be 
presentational only.

> > I consider the a element of (X)HTML implying some behaviour.
> Is there any programmatic behaviour of an <a> element that is
> guarenteed to always be wanted?
>
> I'm not sure there is. For example, I can't think of any similarity
> between the behaviour of hyperlinks in Lynx and in Qube, except
> changing the "location.href", and even that isn't guarenteed to be
> always the case, for example I often open links in new tabs.
I don't know Qube. What is it?

I think opening a link in a new tab is altering the default object of the 
defaut behaviour by replacing the window/frame object with a newly created 
one.

I consider the behaviour of a href for user agents (simply, unqualified 
activation) and for spiders to be the same. That the spider threads is not 
part of the behavioural definition of a href but part of the behaviour of the 
spider.
So I think an interface between the behaviour and the user agent is required 
which takes arguments for the invocation of the method.

I think a href has a behaviour that effectivly consists of three steps:
The behaviour requesting the object from the user agent. (That makes it 
possible for a spider or a tabbed browser to thread / tab)
The user agent creating and delivering the object.
The behaviour invoking the desired method or altering the desired attributes 
of that object.

With such a picture in mind, could we agree on "a href" having a behaviour?

> > I also consider most or all of the elements of XSLT implying some
> > behaviour.
>
> True, but they aren't presentational, so it's a rather different kind
> of "behaviour" (not one which XBL is targetted for, I mean).
Yes, XSLT has behaviour but it's completely unpresentational and, as far as I 
know, cannot be exprressed in XBL.

Which creates an idea in my mind, a behaviour specification language that is 
capable of specifying how XSLT works. But that's just a weird idea.

> Basically, my point is that while I agree with you that Behaviour is
> something that crosses several of the levels, it seems to me that the
> parts of Behaviour that XBL addresses are most closely associated with
> the Presentation layer.
Yes.

Cheers
-- 
ITCQIS GmbH
Christian Wolfgang Hujer
Geschäftsführender Gesellschafter
Telefon: +49  (0)89  27 37 04 37
Telefax: +49  (0)89  27 37 04 39
E-Mail: Christian.Hujer@itcqis.com
WWW: http://www.itcqis.com/

Received on Monday, 6 January 2003 21:47:54 UTC