[246] in Coldmud discussion meeting

clear_param() again.

daemon@ATHENA.MIT.EDU (Tue May 17 22:42:40 1994 )

Date: Tue, 17 May 1994 20:39:09 -0600 (MDT)
From: the Lyncx <BRANDON@cc.usu.edu>
To: coldstuff@MIT.EDU
X-Vms-To: COLDSTUFF

<>     clear_parameter($defining_object, 'parameter)
<
<> Which is called from the object that it is currently residing on.
<> This would not break encapsulation, and it would fix what I see as a
<> gaping wound in the side of ColdMUD.
<
< This would break encapsulation: it allows you to set the value of a
< ancestor's parameter.
< 
< If you want to implement this, make it a $sys-only method that takes
< an object, a definer, and a parameter name.  The system object can
< then have a method which does clear_parameter(sender(), caller(),
< name).

I started to think about this, and I dont know if I was upmost clear.  Perhaps
have:


        clear_parameter('parameter)

Which is defined on a method on the defining object, and called by the object
which wishes to be cleared.  Therefore:


   $objA has .clear_param()    (a thunk)

   $objB calls .clear_param to clear the foo_parameter...

Would this follow encapsulation?