RE: [tcltk-perl] Why is Tcl a subclass of Tck::Tk?

[Jeff H. <je...@Ac...>]

> I think it would be valuable to actually have the Tcl namespace 
> "empty".  All I think it should have is a AUTOLOAD that turns 
> all method calls into a tcl procedure call.  I.e. that:
>     $tcl->foo($a, $b)
> is always the same as
>     $tcl->call("foo", $a, $b)
> is now.  That has the potential of simpilifying the Tk stuff I think.

Yes!  This is definitely what I want out of the API.  This
is what I was expecting in some ways (more natural to me
from a Perl and Tcl point of view), I just didn't know the
magic to make it happen.

There are a few calls that we want to reserve.  We do need
to keep around a few core things (like Eval, call, etc), but
we might make a standard of how to name them.

> For the Tcl::Tk stuff I would really like to be able to move 
> all that deals with perlTk compatiblity to its own module.  
> The API to trigger it could be the same, but as this grows in 
> complexity you don't want to load it when all you need is to 
> use the Tk interface as it is.

We are all in agreement - my perl fu is just not up to the
task of doing this quickly.  I was thinking that Tcl::Tk and
Tcl::pTk should be separate modules, instead of the odd (to me)
	use Tcl::Tk q(:perlTk);
that is there now.

Jeff