In article <[EMAIL PROTECTED]>, Magnus W wrote:
> Chris Hoess <[EMAIL PROTECTED]> wrote in
> [EMAIL PROTECTED]:">news:[EMAIL PROTECTED]: 
> 
>> In article <[EMAIL PROTECTED]>, Magnus W
>> wrote: 
>>>
>>> This is not crap. This will make it possible to use the wheel to
>>> scroll in DHTML scrollers, which is a quite useful feature...
>> 
>> ...for building even more ridiculously specific web pages.  "This page
>> best viewed in IE 6--with a wheelmouse!"
> 
> You don't seem to understand the issue, or rather, are blinded by your 
> anti-MS feelings. I say; this is a useful feature. I can provide examples 
> of why it is a useful feature.

Sure, it's a "useful feature".  So was <font> at one point, and SHORTTAG 
and OMITTAG, and so on, and so forth...but in the long run, all of them 
ultimately hinder the development of the Web.
 
> Features are not evil. Exposing hardware events is not evil. Implementation 
> may be bad, and maybe some more thought should have gone into that event 
> model, but the fact that a web designer can capture all kinds of mouse 
> events makes it easier to build advanced web applications.

In fairness, most of the blame should go to the W3C here, in dropping in a 
slapdash, mouse-centric model for event handlers.  Onmousewheel is simply 
a logical extension of the current ad hoc way of doing things; 
unfortunately, the current way is wrong.  It would be nice if MS said 
"Waitasec, this whole way of doing things is broken; why don't we look at 
developing a better framework," but also unrealistic; it provides (as you 
rightly pointed out) an immediate, cheap payoff to DHTML authors, and the 
consequences when the installed base of DHTML on the web slams head-on 
into device independence issues is not their problem.  So take this more 
as a lament that the W3C has screwed up again, and vendors have begun 
more-or-less innocently following, than a polemic against MS in 
particular.  (I have a box of those at home, anyhow. :-)

> More object-oriented event handling WOULD be nice, though.

Amen.  Rather than the current "hard-coded" event handlers, there should 
perhaps be a mechanism to create generic event handlers and bind them to 
specific events in a flexible way (so that an event handler might be bound 
to a mouseover for mouse-equipped devices and, say, "onread" for a 
speech-based device).  Unfortunately, most of the W3C's work on device 
independence has gone into CC/PP, which has always struck me as an 
unrealistic and unimpressive standard.

On a somewhat more upbeat note, while I was poking around looking for an 
official list of event handlers, I found 
<URL:http://lists.w3.org/Archives/Public/w3c-wai-gl/2002JanMar/0036.html>, 
which suggests that someone's looking into the situation; maybe 
event handlers will wind up being divided into generic and device-specific 
categories, much like logical and presentational elements/attributes in 
HTML.

-- 
Chris Hoess

Reply via email to