On 2007-03-23, at 14:31 EDT, Henry Minsky wrote:
The original code looked like this
var LzRPC = Class( "LzRPC", null, function () { } );
//--------------------------------------------------------------------
----------
// @param Number num: number to pass down to explicitly cast to a
double
//--------------------------------------------------------------------
----------
LzRPC.DoubleWrapper = function(num) {
this.num = num;
}
So was this trying to define a subclass of LzRPC?
The modern equivalent would be to make it a static function of the
class (i.e., a class method). But you will have to update the code
that is testing for LzRPC.prototype.DoubleWrapper to
LzRPC.DoubleWrapper, too. The __proto__ hacking needs to be replaced
with instanceof too.
On 3/23/07, Henry Minsky <[EMAIL PROTECTED]> wrote:
I might've converted this wrong when I updated to use the new class
system.
On 3/23/07, Pablo Kang < [EMAIL PROTECTED]> wrote:
>
> Oops. Thanks, Tucker. :)
>
> On Fri, 23 Mar 2007, P T Withington wrote:
>
> > I posted what may be a work-around.
> >
> > It looks like this mistakenly was spec-ed as an instance
function and
> > should be a class (static) instead (and the encoder would have
to be
> > fixed too).
> >
> > Seems to me the encoder should be using `instanceof` not
comparing
> > prototypes. Not all javascripts support __proto__. (It never
> > occured to me that we had components that tried to do __proto__
> > hacking!)
> >
> > And, I wonder if Double should be a subclass of Number? (Not
sure
> > that will work in JS1.)
> >
> > On 2007-03-23, at 09:02 EDT, Max Carlson wrote:
> >
> > > http://forum.openlaszlo.org/showthread.php?t=8757
> > >
> > > Do we have a testcase that verifies DoubleWrapper works
properly?
> > >
> > > --
> > > Regards,
> > > Max Carlson
> > > OpenLaszlo.org
> > >
> >
> >
>
--
Henry Minsky
Software Architect
[EMAIL PROTECTED]
--
Henry Minsky
Software Architect
[EMAIL PROTECTED]
