On Mon, Mar 9, 2009 at 8:35 AM, Jeff Watkins <[email protected]> wrote:
> Garrett, I gave a specific use case for why I'd like to be able to specify
> the name on an anonymous function: wrapping an existing function with pre-
> or post-call behaviour.
I think I get it.
It seems like just giving the function a name (not *anonymous*) would
fulfill that need.
Example:-
<script>
var getX = (function(){
var x = 10;
function getX() {
return x;
}
return getX;
})();
document.write(getX.name + ": " + getX());
</script>
> I don't think the desire to trigger a function on change of a value is at
> all ludicrous.
What would the type of the variable be? Number? A number with a
behavior? How would you modify ToNumber or ToObject to accomodate for
those changes?
If an object is wanted, an object should be used. Objects encapsulate
behavior; primitives do not.
I think we should cut Memolus some slack for not being at his
> most articulate.
Sorry, I think the code says enough.
Garrett
> On 8 Mar, 2009, at 6:51 PM, Garrett Smith wrote:
>
> Memolus does not say what he needs this for, but it sounds even more
> ludicrous than the recent discussion to allow [[Writable]] name to
> anonymous function objects created in a certain way because the
> "Objective-J" folks wanted it (what was that use case?)
>
>
_______________________________________________
Es-discuss mailing list
[email protected]
https://mail.mozilla.org/listinfo/es-discuss
