overriding object equality tests
brendan at mozilla.org
Thu Dec 20 23:20:47 PST 2007
On Dec 20, 2007, at 3:42 PM, Brendan Eich wrote:
> I wrote "may" in my reply about mutable objects, because IE's DOM is
> different. I'm not sure what's mutable and what isn't (I never
> learned after the early horror in the IE4 days). Mark, any hope
> through immutability to avoid an identity hash?
Note: "mutable" here means "can add expandos", not "can call mutating
methods or attributes that vector into some native vtbl and mutate C+
+ state shared for all wrappers". The issue is whether setting .foopy
on a wrapper should be visible to the next guy to get that wrapper
(e.g., for the document object).
More information about the Tamarin-devel