This is either a bug in v8, or something I deeply misunderstand about
how JavaScript is supposed to work. Please consider the following:
d8> Class1 = function() { this.foo=1; };
function () { this.foo=1; }
d8> Class2 = function() { this.foo=2; }
function () { this.foo=2; }
d8> Class2.prototype = new Class1();
[object Object]
...so here, we've declared a base class Class1, and a subclass Class2.
d8> c1 = new Class1();
[object Object]
d8> c2 = new Class2();
[object Object]
...created an object of each class.
d8> c2.constructor
function () { this.foo=1; }
d8> c2.constructor === Class1;
true
d8> c2.constructor === c1.constructor;
true
...wtf?!? c2's constructor property should point to Class2, shouldn't
it? Yet the 'foo' property was properly initialized:
d8> c1.foo
1
d8> c2.foo
2
...and other tests show that all inheritance is working as it should;
properties added to Class1.prototype appear on both c1 and c2, while
properties added to Class2.prototype appear on c2. But how can this be,
when c2's constructor points directly at Class1?
I've tried this with deeper hierarchies as well: if we define a Class3
which subclasses Class2, any objects we create also have their
.constructor property pointing at Class1. It's as though the
.constructor property always walks all the way up the prototype chain
and returns most "base" constructor.
I've tried hard to understand why this might be correct, but I just
can't see it. If this isn't a bug in v8, can someone explain why?
Thanks,
- Joe
--~--~---------~--~----~------------~-------~--~----~
v8-users mailing list
[email protected]
http://groups.google.com/group/v8-users
-~----------~----~----~----~------~----~------~--~---
