[
The discussion below should maybe be taken offline, but as
usual I can't resist "one more online response" :-). Our
revered moderator should feel free to edit this or censor it
entirely, and I'll try to refrain from extending this discussion
further in public unless other TMLers express interest.
> Tom Harris
>> Me (Bart Massey)
]
First, let me thank Tom Harris for providing us with much new
detailed information on the DV! This was the kind of thing I
was hoping for in response to some of the earlier TML queries...
>>Speaking as a computer scientist of sorts
> Gee, I was speaking as a real computer scientist. Sorry to be so
> sarcastic, but your article has a attitude problem.
Sorry -- I didn't really want to get into my qualifications in
detail since I didn't think that they were that important to the
discussion, but since you sort of asked, let me just mention
that I've been programming computers for almost 15 years now,
I'm about 6 months from my MSCS here at UO, I spent 3 years as a
professional computer programmer and system administrator, I
spent two terms teaching undergraduate CS at Portland State
University, and I've done a lot of hobby electronics, both
digital and analog. Hopefully, this is adequate background for
understanding some of what computers do and can do.
You're right, I do have a bad attitude about the proposed
virus. Let me explain again the fundamental reasoning behind
my earlier letter:
Several people on the TML described a trivial method
(moving the setting to another part of the galaxy) of
obtaining a desired effect in the setting for SV (low to
medium technology, but with starships). They also
voiced what I feel were valid complaints about the
method GDW proposed to obtain this effect (a "Long
Night" caused by computer viruses). The responses I
saw to these complaints from GDW didn't seem to me to
evince *any* sentiment other than "we're smarter than
you give us credit for, and so you should trust us when
we say our way is superior" (although this may have
been a bad reading on my part) and I am frankly hesitant
to "trust" GDW after the MT debacle.
It was based on these perceptions that I wrote earlier.
> I'll grant that
> the idea of a virus that could do what was described falls totally in the
> realm of science fiction, but being allowed to make any pre-assumptions
> you choose how would you make the Traveller universe stop?
Distance, distance, distance! Just put the SV setting far
enough away that it could be reached from "known space" only
through *extreme* effort, if that. Then set it up however you
want.
If you want an analogy, consider 14th century Europe, 14th
century China, and 14th century America.
> Frank thought that computers were the way to go and I made
> some suggestions. If you don't think it will work then present
> something better in this forum with the same effects.
Well, assuming that I *did* want the effect you're intending to
achieve, I would probably just adopt the method used by one of
the all-time great sci-fi classics, Asimov's Foundation
Trilogy, and let the mysterious forces of sociological
"psychohistory" do the job. Heck, the post-rebellion MT
universe already looks enough like the fallen Roman Empire to
me to make the whole thing plausible anyhow. If you wanted to
be really faithful to the original you could use
specially-trained telepaths as psychohistorians.
But I sure wouldn't use this or any other idea without having a
bunch of independent and qualified people review it first for
obvious problems. And I'd sure listen closely to what they had
to say about it.
> Personally my choice of attack would be the jump
> drives, but I don't know the physics to figure out how to do it.
Well, for example, as several of us physicist sorts (my
undergraduate degree is a BA in Physics) have noticed
(particularly Dow Reider, if memory serves) it is possible to
build perpetual motion machines out of thruster plates -- I
believe I could build one with the jump drives as well under
the current rules. Assume for a moment that this is so.
Let's assume, then, that this energy being produced during jump
doesn't come from "nowhere", but from some source unknown to
the builders of jump-drives. Now, let it run out :-).
Note that this is an off-the-cuff idea which might not work out.
So what's the first thing I'd do with it (once I'd fully
elaborated it) if I wanted to use it? I'd *run it by the
physicists on the TML* and see what they thought of it...
> Rather than
> atacking memory it destroys the CPU.
....
> My assuption was that this would be feasable
> because those programs
> were guarding against memory assaults.
> Hey this is a big gimme, but then
> so is most science fiction.
OK, first of all, let's agree on some common terminology.
First, let's call this thing a "worm" instead of a virus --
maybe then we can both agree that this is what the DV actually
is. For the purposes of this discussion a "worm" is a computer
program which spreads by copying itself from computer to
computer.
A "CPU" (lit. Central Processing Unit) is a collection
consisting of "memory" together with a permanently hardwired
"control". "memory" is storage which will be altered during
normal execution of computer programs (thus, the CPU contains a
bunch of internal "registers" (both directly user-visible and
otherwise) which are a form of memory, but the CPU typically
also runs with some larger amount of "peripheral" memory
attached).
Now one of the more straightforward things to do in TL7-8 CPU
design, as I understand it, is to check that no possible memory
state can lead to the CPU hardware or control being physically
damaged. Why is this in principle so easy? Because one uses
existing computers to do "worst-case" analysis, and
over-engineers the hardware and control to handle the worst
case. Why is this almost always attempted? Because it's not
possible at TL7-8 to determine that there is no path from the
initial memory state(s) to this worst state. Indeed, there are
sound computer-theoretic reasons why this latter problem is in
general insoluble at *any* TL.
Which isn't to say that CPUs aren't built which are vulnerable
to the sort of worm described above. But I'd be willing to
wager that there's no "common architectural flaw" in current
CPUs which makes them all vulnerable to the *same sort* of
software attack, and that there never will be. (EXERCISE: find
out how to damage the CPU of the machine you're reading this on
from software.)
> At any rate, the lag time before the program
> takes effect is similar to a virus and is how it gets so widely dispersed.
> CPU failure should look very small when the virus is first introduced but
> increase exponetially as the virus takes effect.
Better yet, make the worm activate the destruct code everywhere
at once, using global clocks. ("real" viruses with long lags
must affect the host populations exponentially rather than
instantaneously because they have no global clock available,
and would probably "choose" to do so anyhow because destroying
their entire host population is an "undesired" side-effect
rather than a goal.)
> "Seeding" production
> facilities insures their shut down leading to the lack of replacement
> CPUs.
Seeding production facilities shouldn't be *necessary* for the
worm; I assume you just want to *make sure* the worm reaches
these spots, which is sensible.
> All of this spirals with the social collapse causing lack of
> transport to provide manufacturing materials and food for the people.
> >(2) If all the computers are destroyed, how the heck are the
> >Star Viking ships driving around? You want to jump a starship
> >under manual control? Good luck -- you'll sure need it!
> (3) is the partial answer to (2) i.e. not all technology dies
> and once the problem is known protection can be taken and some
> recovery done.
Ahh, but you can't have it both ways (at least not easily).
IMHO the most plausible "protection and recovery" scenarios
would lead to a mixture of TLs from the low to the high end.
If *any planet* in the region manages to retain TL13, it could
bootstrap the others to that level in a straightforward
fashion...
> Frank's assumption is that Traveller societies are pretty close to [totally]
> dependent on computers. The fall is not from TL15 to TL6, but from TL15
> to TL0 and then the problem is how far can the survivors come back.
OK, I didn't understand this (although see directly above).
> Also I
> have always interpreted Traveller TLs as being the TL that a planet could
> support on its own. I could go one step further and even say it is what
> the planet can support given the support it gets from other worlds.
OK, there's two distinct interpretations of planetary TL in two
sentences. We've had our own discussion on the TML about what
planetary TL means, and one thing I thought we pretty much
decided is that the planetary TL is some kind of average over
various planetary technologies, which raises additional
questions. One thing it would be nice to see in any new GDW
rules is a very detailed discussion in a single place of TLs
and what they represent.
> I wouldn't
> think most TL15 worlds are self-sufficient
Hard to say.
> and if space travel is
> interupted they will suffer economically. For lower TLs it is almost
> certain these planets import higher TL stuff to make their life easier and
> are vulnerable through these imports.
"`Is that what you're setting your hopes on, man? What
do you expect? A housewives' rebellion? A Jacquerie?
A sudden uprising of butchers and grocers with their
cleavers and bread-knives shouting `Give us back our
Automatic Super-Kleeno Atomic Washing Machines?''"
-- Isaac Asimov, "Foundation"
I've always wanted to use that quote somewhere :-). But I agree
that if you could destroy all the computers in a TL15 culture
simultaneously you might well send the culture back to TL0.
Seriously, reread the Foundation Trilogy -- it'll probably give
you some good ideas about more *subtle* ways of doing what you
want.
>>(4) Do you really believe the "common architectural flaw"
>>argument above? If so, then answer these: Do you believe that
>>all TL7 bridges contain a common easily exploitable
>>architectural flaw?
>They do. Use of Aluminium superstructures as shown in the Falklands.
>We are currently in the process of changing them all back to steel.]
>>Is this more or less likely than that all TL4 bridges did (HINT: think
>>resonance)?
>They did. That's why we made TL7 bridges with so much electronic garbage,
>TL4 radios are so easy to jam once you know how.
Oops. I should have made myself clear -- I was using "bridge" in
the Webster's II sense (1) "a structure spanning and providing
passage over an obstacle, as a waterway", not (6) "a crosswise
platform above the main deck of a ship from which the ship is
controlled". An understandable misunderstanding, given the
Traveller milieu :-).
BTW, we knew how to jam TL4 radios at TL4, so I would call that
more of a "unavoidable technological difficulty" than an
"undiscovered architectural flaw", but maybe I'm just
quibbling...
> >Is this a fair analogy?
> >What other analogies of this sort might one draw
> >(IDEAS: automobiles, electrical devices, plumbing)?
> Haven't you noticed recalls of new ideas across auto corperations to fix
> such flaws e.g. the change to timing belt rather than the timing chain
> several years ago caused numerous engine failures which the manufacturers
> paid for when they broke. What about the Hinsdale fire several years ago
> Mother's Day and its effect on telephones service for the entire Chicago
> area. Plumbing how about Legionare's disease and its relation to the
> new vent system in the convention hotel. We are vulnerable to disasters
> the examples abover were just pulled off the top of my head.
> >Do they lead to similar conclusions (SUGGESTED ANSWER: yes)?
> Yes, you bet i.e. system normal all f***** up. Flaws like this exist in
> all human societies and most of what we produce. The ability to exploit
> them may well be in the realm of science fiction, but that's part of
> Traveller isn't it. We tend to be dependent on some bit of technology and
> if it fails or we streach it too far, our society will fold up and go
> away. For the Mayans this was slash and burn agriculture vs population
> pressures. For the Sumarians it was irregation vs salt buildup in fields.
> For our current society it may be the internal cumbustion engine/pollution
> /progress vs the atmosphere.
Here, I think, we reach the crux of the argument. For me, and
IMHO for many in the TML, a variant of Occam's Razor is an
important tool in SFRP game design and play. I'd phrase it
like this: "One should not multiply unexplainable entities
without cause in the design of SFRP games or adventures". So
far, there's really only been a few "Big Black Boxes" in
Traveller -- jump drives, antigrav, and thruster plates, for
the most part. Note that many TML respondents (count me in)
seem to hate the thruster plates for the very reason that
fusion rockets would achieve the same effect in an
*explainable* way. I oppose the BBB DV idea because I think
the idea of simply moving SV elswhere physically allows one to
achieve the effect desired much more "simply".
> >Finally, riddle me this: GDW is making noises about "listening
> >to customer input". I've seen at least 5 distinct messages
> >approving a suggestion that the Star Viking setting be
> >explained by simply putting it in another part of the galaxy
> >from Traveller -- an area with low max TL. This seems to me to
> >be reasonable and practical while still allowing the SV
> >universe adequate latitude for development. So WHY KEEP PUSHING
> >THIS VIRUS THING???
>
> You get the same amount of support either way. They get much more reuse
> by just leaving things where they are. So the decision is probably based
> on keeping R&D costs to a minimum. What I'm trying to say is the impact
> of the reuse of the Traveller universe is the same on you in either case.
> For example, even if they put it in another part of the universe then you
> won't see anymore updates for MT anyway so you will have to exist with
> what you have and if they put in the existing universe then you will have
> the same effect by ignoring it. Where you choose to go with Traveller is
> up to you, not up to GDW. None of my Traveller running was ever done in
> the Imperium or even the same universe by my choice and to avoid these
> problems with not liking someone elses dicision about the game.
> Personally I agree with some of your statements above, I thought 2300 +
> 1000 or more years or putting it so far in Travellers future that MT is
> unaffected might be an even better way to go. But if for business
> reasons GDW has decided to drop support of MT then you have to make a
> choice about whether or not you will continue your support or MT. Its not
> that bad of a decision either way you pick its just a matter of preference.
I think what you're trying to say here is that by setting SV in
the near-future in the "known space" of MT/Traveller, GDW gets
the ability to re-use much of the background material from
Traveller. I agree. I think they're also giving up an
important thing which hasn't been much discussed, namely the
ability to do Trav/SV crossover adventures, but let's assume GDW
doesn't care, for business reasons, ever to be able to do so.
They're also giving up the ability to *change their mind* and
start supporting MT again, which is not a trivial thing, but
again, assume they don't care, because they made the right
decision the first time.
Given all that, I don't think that it's the *idea* of setting
SV after a "Long Night" which bugs me so much, it's that I
still don't like or believe the whole DV *mechanism*. I think
others may still be hesitant as well (anyone?). So invent
another one! "Long Night" scenarios are millicredits per
kiloplot in SF, and I personally am much more inclined to
believe in Niven/Pournelle's "superconductor virus" or even
your own "oil-damaging virus" than in the DV. All I'm
suggesting is that whatever scenario you come up with, run it
past us first. We'd be happy to criticize it, and even suggest
ways to improve it. But once it's set in stone it's no fun for
us anymore.
Finally, assuming that you're still dead set on a computer
virus, here's an alternative scenario you may find appealing:
a disgruntled TL21 scientist writes a complicated collection of
programs which spread through galactic computers in spite of
being detected and understood, because they simply don't cause
enough trouble to be worth the effort (e.g. a program which
broadcasts jokes on a ship's intercom in the evenings -- or
imagine software Tribbles, if you will). These programs *very
slowly* get more troublesome (e.g. the program is replaced by
one which sometimes starts broadcasting jokes just before a
complicated jump), but by the time people begin to realize how
annoying and dangerous the whole thing is getting, the
debuggers, repair robots, everything is full of them. Further,
the higher the TL, the more troublesome and common the programs
are.
As a result (insert complicated sociological forces here)
everybody begins to use higher TL computers only in "safer"
situations. Eventually, *usable* computing ends up at some low
TL, and the rest of society follows.
The prettiest bit of the above scenario is that you can still have
TL18 computers, robots, etc. They will just act strangely and
unreliably, making it iffy whether it's better to work with them
or without them. Endless fun for the Ref. (See e.g. a TML
description in an earlier volume of a repair robot which tended
to go batty, or the classic Terry Carr SF short "The Robots Are
Here" ...)
But is the above scenario at all plausible? I'm sure the TML
members will let me know, and if they say not, I'd sure invent
something else...
At any rate, it'll be interesting to see how the whole thing
turns out. Thanks again for your feedback, and if I've
offended you or anyone else who's been reading our exchange, I
hope that you will accept my humble apology. Please take my
word for it that I have only my own best interests at heart :-),
and that it is very much in my best interests for GDW to do the
best SFRP games possible. Take care,
Bart Massey
[email removed]