[DGD] Diferences... Thank you, and more cuestions.

[Felix A. Croes]

Arnau Rosselló Castelló wrote:

>[...]
> In the mailing logs i've read that dworkin doesn't want dgd to esta-
> blish outgoing connections, for security reasons, and i can understand
> why that functionality is kept in a separate package, but what about
> listening to many ports at a time ports? the security problem doesn't
> apply here, and it could be very helpful(say, you want one of those
> ftp servers, and one http server, and the mud in the binary port), so
> why is that it isn't in the official release?

It is something I may add in the future, though I have multi processor
support to deal with, first.


> Another question: What Is a tick exactly? It measures CPU capacity,
> but is it a fixed time length? a variable time length that depends on
> the power of the processor? it works with time at all or measures the
> number of instructions?

A weighted measure of the number of instructions and the size of the
datastructures manipulated by these instructions.


> YAQuestion: Is there any way to know the maximum and/or actual
> ticks and stack? I'm thinking on making that compile_object() and
> clone_object() assign a certain cuantity of ticks and stack to the com-
> piled/cloned object, who will be called everytime with that amount of
> stack/ticks, and substracting that many ticks and stack from the assigned
> to the object calling compile/clone. Do you think this is viable, or
> is my idea flawed in some way?

You can set the ticks and stack depth available to code with rlimits,
and query current values with the status() kfun.  Limiting the amount
of resources available to objects is a good idea.  In fact, the kernel
library does exactly that. :-)

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