Hey
Nick,
This
was checked into svn on 29th March...
Are
you using the library version that was written much before
that?
If
not, then this might still not address the problems.
Thoughts?
-Mandy.
-----Original Message-----OK, scrap this, check out the latest in svn. this issue got raised in trac: http://dev.rubyonrails.org/ticket/4465
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]On Behalf Of nick hemsley
Sent: Thursday, May 11, 2006 8:24 AM
To: [email protected]
Subject: Re: [Rails-spinoffs] prototype: leak with Element.extend
On 5/11/06, nick hemsley <[EMAIL PROTECTED]> wrote:Disclaimer: Im still getting used to dynamic scoping. & _javascript_ as well.
Which part clobbers window.property? If this _is_ the case, then suiltably unique name would be a temporary, but if you are talking about the for() loop itself, that is unchanged from the original source... actually looking at the latest svn, Element.extend appears to have changed. Too late for me to look at it properly now, but it appears it uses some tricky scope checking to decide what to return. (would this avoid all memory leaks?) ... nice. That mofo is just too fubar for me to come up with.
On 5/10/06, Todd Ross < [EMAIL PROTECTED]> wrote:On 5/10/06, nick hemsley <[EMAIL PROTECTED]> wrote:
> for (property in methods) {
for (var property in methods) {
ensures that you're working with a locally-scoped property variable.
The way it's written, if window.property existed, then you just
clobbered it.
Todd
_______________________________________________
Rails-spinoffs mailing list
[email protected]
http://lists.rubyonrails.org/mailman/listinfo/rails-spinoffs
_______________________________________________ Rails-spinoffs mailing list [email protected] http://lists.rubyonrails.org/mailman/listinfo/rails-spinoffs
