[DGD] Re: Recompiling object trees

Jason Cone jcone at cs.tamu.edu
Sun Feb 22 01:06:23 CET 1998


-----Original Message-----
From: Felix A. Croes <felix at xs1.simplex.nl>
To: dgd at imaginary.com <dgd at imaginary.com>
Date: Saturday, February 21, 1998 5:21 PM
Subject: [DGD] Re: Recompiling object trees


>The reason for handling it this way is that DGD can recompile objects
>without destructing them, but only if the object is not inherited.
[...further explanation...]

I'm hearing 2 different things, here.  Let's go back to the original issue
presented by Mr. Carpenter:

>1) If I made a dramatic change to something fundemental, like the
>auto object, it seemed cleaner to restart the mud from scratch.

Mr. Harte implied that dynamically updating the auto object was possible,
yet the above explanation by Mr. Croes would counter that.  I'm truely not
trying to spark a debate here, but I think this is important enough to
undestand and I think there are more people than just I who don't fully
grasp the implications of this issue.  Here are my assumptions and
conclusions based on the previous few posts:

* To update an object that is inherited, you must destruct every inheriting
object.


* You cannot update/recompile the auto object via compile_object() because
it is inherited by every object.  You would theoretically have to destruct
every loaded object for this to happen - this is not possible.

* The only way to "update" the auto object is to perform a state dump and
restart DGD.  I'm assuming here that, even if you perform a state dump and
reboot from it, the newest version of the object will be used.

Is this right?  If not, here's what I'm having trouble grasping... I
understand how to keep track of inheritance trees.  I understand how you can
update a whole tree.  I -think- I understand how the auto object can be
upgraded, which is the original issue and issue in question.



||==================================================================||
||                         ||  "Hard work spotlights the character  ||
|| Jason H. Cone           ||   of people:  some turn up their      ||
|| Dept. Computer Science  ||   sleeves, some turn up their noses,  ||
|| Texas A&M University    ||   and some don't turn up at all."     ||
|| jcone at cs.tamu.edu       ||             - Sam Ewing               ||
||==================================================================||





List config page:  http://list.imaginary.com/mailman/listinfo/dgd



More information about the DGD mailing list