- From: Ian Hickson <ian@hixie.ch>
- Date: Wed, 2 Aug 2006 22:55:02 +0000 (UTC)
- To: Jim Ley <jim@jibbering.com>
- Cc: www-html@w3.org
On Wed, 2 Aug 2006, Jim Ley wrote:
>
> >I would, however, like to comment on the assertion that Bjoern makes
> >that "...Users want web applications to respond to their actions before
> >they have been fully loaded..." While you have the right to your
> >opinion, can you actually back this statement up with factual data that
> >*proves* this statement?
>
> Proves the statement? I could show you (with appropriate NDA's and
> agreement from 3rd parties) plenty of user testing that shows users
> start interacting with any control as soon as they are able, it matters
> not if "the whole document has downloaded" - that's simply not a concept
> the users understand, they interact with things as soon as they are able
> - so as soon as they are shown it, no matter how they are shown it.
It's not often that I agree with Jim, but on this point, he's right. Every
usability study I've seen has underscored the fact that users don't know
if the page "has finished loading" or not, they interact with it as soon
as it looks like they can, and get rapidly annoyed at the computer if
things don't work.
Browser vendors have no choice here -- Web pages have to incrementally
load and work on the fly.
This implies one of two things -- either:
a. XML has a processing model that allows for incremental parsing,
loading, and interaction with incomplete documents, with scripts
running before the document has finished parsing, or
b. XML won't be used on the Web.
Admittedly, option 'b' does appear to currently be the case.
--
Ian Hickson U+1047E )\._.,--....,'``. fL
http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,.
Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'
Received on Wednesday, 2 August 2006 22:55:13 UTC