[Home] [Groups] - Message: [Prev in Group] [Next in Group]

nu.kanga.list.mud-dev

3442: Re: [MUD-Dev] MUD Design Fundamentals (Was: Looking for

[Full Header] [Plain Text]
From: Ola Fosheim Grøstad <olag@ifi.uio.no>
Newsgroups: nu.kanga.list.mud-dev
Date: Mon, 1 Sep 1997 17:15:00 +0200 (MET DST)
References: [1]
Organization: Kanga.Nu
[J C Lawrence:]

>Rather than get into definitional arguments, its probably easist to
>comment on what I do, which is pretty close in this regard to ColdX:
>
>  Everything is persistant.  That means all code, all objects, all
>methods, all attributes, all relations, all references, everything. 
>The easiest way to think of this possibly would be to consider battery
>backed RAM or core memory (magnetic rings et al).  
>
>  Persistance is hidden from the user.  Its automatic, hidden, and
>implicit.  The user never actually deals with it or sees the
>mechanics.  

But this is a taskspecific system right?  That's something different,
because the realtime-problems of connections and external events and
systemclock can be hidden "under the hood".

I figured we were talking about general "make no assumptions"
programminglanguages with persistence (if they exist at all,
general languages, that is).

If we are talking userprogramming (which I don't even consider as
I don't want to give programmers an advantage over other people) then
you don't have much choice, but to go for a persistent objectstore.
If you want a sane system..  But then again, with userprogramming you
are giving up some control over the object-system, and major changes
are not an option anyway.

Ola.