[DGD] Re: My idea for the DGD driver - validate

[Felix A. Croes]

Josh Dady <jpd at indecisive.com> wrote:

> On Feb 12, 2004, at 7:07 PM, Felix A. Croes wrote:
>
> > Regarding calling security, I think one mistake was made during the
> > inception of LPC: functions (and variables) should be private by
> > default.  Unfortunately it is now too late to change this.
>
> I don't know about that -- the idea isn't that much different from the 
> semantic changes that happen when you change the typechecking config 
> setting.  If you reallyl wanted to, you could even introduce a mode 
> where declarators sprinkle member-access magic dust on the object, 
> similar to how strict types happen at typechecking==0.

Yes, MudOS has something like this.  You can configure the driver
(before compilation) to make functions private by default, and in
that case 'public' presumably becomes meaningful.

But -- that's definitely changing the language.  Typechecking modes
merely make mandatory what is already possible in typechecking mode 0.
Increasing the typechecking level doesn't change the meaning of code
which is valid at both levels (all right, there are some borderline
cases because of the difference between 0 and nil).  A change that
would be more in line with typechecking modes would be to require a
'public', 'static' or 'private' class specifier for each function.

Regards,
Dworkin
_________________________________________________________________
List config page:  http://list.imaginary.com/mailman/listinfo/dgd