----- Original Message -----
From: "Ian Hickson" <[EMAIL PROTECTED]>
To: "Andrew Fedoniouk" <[EMAIL PROTECTED]>
Cc: <[EMAIL PROTECTED]>; "Benjamin Joffe" <[EMAIL PROTECTED]>;
<[EMAIL PROTECTED]>
Sent: Sunday, July 02, 2006 2:18 PM
Subject: Re: [whatwg] Canvas 2d methods
On Sun, 2 Jul 2006, Andrew Fedoniouk wrote:
1) if it is declared as
var foo = something; in current frame then
foo.bar();
foo.baz();
then execution is quite efficient.
But
with(something)
{
bar();
baz();
}
still requires scan of the whole universe in the worst case.
But this isn't the worst case! It's the best case -- you know exactly what
a 2D context's members are.
Consider this:
with(ctx)
{
bar( some_var_name );
}
runtime will scan full chain of ctx for some_var_name. What for?
It will scan *exactly* the same places as it would have if you didn't have
the "with", plus the context itself.
And that ctx is pretty heavy thing.
Again, "with" was considered harmful - that was the inital idea of
discussion.
Anyway, this is all a straw man -- this isn't the reason that the spec
doesn't allow this. It doesn't allow this because Safari didn't do it this
way in the first place, and changing it would likely introduce bugs (while
still not helping authors for some time anyway).
As far as I can see return value is not defined at all in:
http://developer.apple.com/documentation/AppleApplications/Reference/SafariJSRef/Classes/Canvas.html#//apple_ref/js/Canvas.lineTo
So the proposal is just to define them as returning 'this' instead of
nothing.
Andrew Fedoniouk.
http://terrainformatioca.com
