Re: [FT] Robot workersRe: Why superships cost more per mass

David Rodemaker wrote:

> > However, this is still mostly guesswork, and YMMV is still highly
> > operative. I'd have to say the current thread on corruption in ship
> > construction is VERY speculative, unless I missed ANYONE declaring
their
> > expertise in the area.
>
> Since I started it I'll comment... <g>
>
> I was mainly viewing it as less "corruption" and more "pork" to tell
the
> truth, I think that that point is valid more so than as "corruption".
> However, there is also a fine line between the two <g>
>
> Once again, there is a less fine line between "maximizing
profitability" and
> "charging the market price" to "sticking to the government because we
can"
> (remember those $900 toilet seats, and $500 wrenches back in the 70's
or
> 80's?) I would guess that it happens more than we know but less often
than
> we are afraid of...

Construction of starships in tuffleyverse will probably be different
from
shipbuilding right now because the great powers are faced with the same
problems
as the British Empire in the nineteenth century.  The only way to have a
presence in a backwater is to actually be present.  This is especially
true if
the only FTL communication is by courier ship.	The four major powers
need a
large number of hulls in space just to keep the mail flowing.

The timeframe of tuffleyverse is after the amortization of a lot of the
components would have been paid for.  The beam-4's of the Komarov are
not
inferior (in game terms) to any other beam-4, so the ships are designed
for the
weapons, not the other way around.  An old ship like the Komarov implies
that
combat systems are  mature systems.  The life support systems for these
vessels
are no more complex than for a nuclear powered submarine of the same
size, and
only the Jeanne d'Arc carrier pushes this limit for a submarine (28,000
tonnes,
crew of 300).

If you doubt that the prices of the high-tech systems will go down, just
look at
the cost of photocopiers before and after the Xerox patent expired.  The
complexity of the combat management systems will hit a plateau, as the
system
designers will actually reduce the feature set of the hardware and
software, so
that the system falls within that subset of all programs for which the
halting
problem can be solved. [Bug free computer systems have been designed and
implemented, and I know of one example that even does something useful.]

[Off-topic tangent]
Ontario Power Generation, then Ontario Hydro, decided to go with an
ultra-modern
computer controlled reactor safety system.  However, the major snag was
that
they not only had to prove that the software design was correct, but
also that
the implementation was correct.  It took months just to come up with a
way to
describe the software in terms that could be verified, but after a year
and a
half, they were able to prove to the satisfaction of the Atomic Energy
Control
Board that a license could be issued for the untried technology. 
Previously,
the safety systems were electro-mechanical in nature, with hard wired
relay
logic, but the components were expensive and required diligent
maintenance.
Should Ontario Power Generation ever build another nuclear generating
station,
they will use the old style safety system, to save the high verification
costs.
But it should be noted that there is at least one computer system that
has been
proven mathematically (not statistically) impossible to crash, yet it
still does