From:	Phil Levis <pal@**.BROWN.EDU>
Subject:	VR 2.0
Date:	Thu, 28 May 1998 17:07:58 -0400

Something tells me that this was discussed a long time ago, but I'd like
to hear the thoughts of the present denizens of the list.

In my opinion, Virtual Realities 2.0 is an excellent book; the rules and
systems described are robust and detailed, and the wealth of information
provided allows for campaigns set entirely within the Matrix, in which
every character is a decker.

One aspect of the book bothers me, however: the methods by which deckers
obtain their programs. Either they buy them, or they write them. Writing
them takes excessive amounts of time: the example given of a decker with a
computer skill of 8 (damn good) taking a base time of 128 days to write an
Attack-8S program. Buying them is extremely expensive: Hacker House sells
computer programs for three million nuyen. That Attack-8S could be sold
for 128,000 nuyen. That's a pretty amazing living, 1000 nuyen a day.

Both of these systems seem at odds with one of the aspects of deckers
which I have always thought to be important: the idea that deckers write
their own programs on a regular basis, indeed, use their own programs
almost exclusively.

Given good programming techniques, it seems reasonable to me that one
should be able to 'upgrade' existing programs that you own. For example,
one might buy an Attack-6M utility from some software house, hack it for a
while, upgrading it to Attack-8M, use it a bit, then hack it some more to
add DINAB capability. Eventually, the program will need an overhaul as the
additions begin to tax the design of the original program, but it I find
it quite reasonable that competent deckers should be able to write
programs which are not utterly from scratch.

I've been working on a rules system to allow this, but I'm wondering what
people on the list feel should be important considerations. For example,
when should upgrading a program be a losing proposition as opposed to
rewriting from scratch? Which options should be the most difficult to add?
For example, transforming a one-shot program into a full utility should be
very difficult; the one-shot option has a tremendous amount of
optimization and little tricks which allow it to fit in the smaller memory
space. Additionally, if it were not difficult, software houses couldn't
offer one-shot test programs. Upgrading a program to have an Area option
should be much more difficult than upgrading the Area to a higher rating.

Thoughts?

Phil