On Jun 21, 10:47*am, Michael Ash wrote:
In rec.aviation.student Nomen Nescio wrote:

From: Le Chaud Lapin

Very nice, and applies well to software too. Of course it's not true that
software has 0 marginal cost. There are support costs, which can be
significant.

I said material cost, not marginal cost.

But let's say that software really does have zero marginal cost. Well,
this is extremely *bad* news for the use of software in GA, not good news
as has been presented.

Why? Because software costs a *lot* of money to make. And with zero
marginal cost, the price is effectively the development cost divided by
the size of the audience.

GA is a pretty damn small audience. Why do you think you can buy a
perfectly capable car GPS, with a database full of every road in the
country, for under $200 but you'll spend ten times that much on something
that's significantly less capable for your airplane? Certification and
liability come into it, of course, but even ignoring those you would spend
what seems to be an unreasonable amount of money. This is just because the
development costs are fixed but the audience is microscopic.

To keep costs down, you want something with low development costs, even if
the material cost is significant. This mean proven designs, simple
mechanical linkages, etc. And guess what, that's what we have. Software
isn't going to save you any money unless you either find a way to make
multipurpose software that the public can also use, increase the GA pilot
population by an order of magnitude, or create a magical software-making
machine that can cut your development costs by an order of magnitude.

It is my belief that a software-controlled PAV, with the features
outlined by NASA/CAFE/PAV would allow an

increase the GA population by an order of magnitude (at least)

because the machine would be easier to fly, etc.

This is what the FAA, NASA, CAFE, DARPA, and aero/astro departments
all over the United States and elsewhere would like to see, not just
me.

-Le Chaud Lapin-