Things that should probably be looked at

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

This is actually something that has irked me about event info all the
way back to DOM2, and odds are if it's not noted at some point, wide
deployment of XBL will make it very obvious.

At this point in time, mouse events only return two types of
coordinates; coordinates relative to the screen, and coordinates
relative to the document (well, in general).  However, what happens if
neither position can be anticipated (for example, through an XBL binding
that adds in new event handling)?  For situations requiring relative
positioning, this won't be a big deal, but if absolute positioning is
needed, it becomes a huge annoyance because one would have to either
anticipate the positioning of an object, or one would have to get the
width, height and relative positioning of every tag between the object
in question and the object the person wants to take coordinates relative
to.  Considering the amount of work it takes to get that kind of info
from DOM3 Views, most people will likely miss it and ask why you can't
do absolute position coordinates relative to a certain parent.

I can't say what would be a good solution to this problem, but it IS a
problem worth pondering.

Oh, and in terms of the window object, I believe the navigator object
could use some form of standardization.  It's so completely random in
implementation right now, it's downright unusable for all but the most
simple of tasks.  Not to mention most of the uses for which it was
designed for can't be done with it because of bad implementation and
lack of a standard for it's behavior.  I know for a fact
navigator.plugins would be very useful if it were decently supported...
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.4 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iD8DBQFEuzm3vCLXx0V8XHQRApM8AKCknvcgmWIjqV4av36EmLKRcMTThgCePr8s
oo1WHnaUU6yhfbryNJSfoDY=
=SbJf
-----END PGP SIGNATURE-----

Received on Monday, 17 July 2006 07:18:29 UTC