Re: the free software paradigm [was Re: Amazon used lisp & C exclusively?


Nathan Baum <nathan_baum@xxxxxxxxxxxxxx> writes:

On Sun, 30 Jul 2006, marc spitzer wrote:

On 2006-07-30, Nathan Baum <nathan_baum@xxxxxxxxxxxxxx> wrote:
On Sat, 29 Jul 2006, Kent M Pitman wrote:
The computer industry has changed,

Yes, it has. It seems like you'd rather it changed back. But at least
you're past denial, eh? :)

and what step are you at?

Mu. I'm not grieving the demise of the old computer industry. :)


Its still around, sadly enough. My dad uses a suite of for-profit
software to help him operate his observatory and take pictures and a
such a collection of unresponsive vendors selling bug-ridden, flaky and
poorly designed software its not been my misfortune to see anywhere else
yet.

There is generally little open source in the amateur observatory racket,
principally because most all the software is there to operate dedicated
hardware so there is little competition or incentive for the vendors to
be responsive to their customers.

As an anecdotal point, for 8 years or so up until late 2000 my father
ran a business building and selling fiberglass observatories to the
amateur market (many pro operations bought them too) and I did the
computer controls for the dome; it slaved the dome opening with the
scope, read weather sensors and closed up the dome when rain was
detected or the wind blows up enough and a few other things. We open
sourced all the client software, documented all protocols and interfaces
both software and electrical, and published the board schematics. The
only thing we didn't release is the firmware in the microcontroller,
though the control model was simple enough that a reasonably skilled
programmer could cook it up without much trouble. The microcontroller
code was a compiled dialect of Basic running on PIC16C77 devices with
just enough assembly to do the input sampling- nothing exotic in the
least.

Several thousand of the observatories and the computer controls were
sold worldwide (and continue to be sold since my father sold the
business). While I was there the company had 6 full-time employees, had
401k plans and grossed just under a million $ a year for several
consecutive years.

We thought about the possibilites of someone reverse engineering the
microcontroller and selling their own control system, in the end we
decided the risk was minimal for several reasons; a control system isn't
as easy as it looks, so a reverse engineer is going to have to do a lot
of work and our prices were not what the market would bear instead being
computed on the basis of SWAG estimates of the amortized costs of
development and assembly plus "reasonable" profit so it was unlikely we
would be considerably undercut on price. In fact we never bothered with
patents on the control system hardware or software much less protection
measures of the IP. I also felt and feel that there was little
motivation to reverse engineer the controller or sell a 3rd party
version of it because we made our systems so easy to interface with- and
perhaps more importantly, made responding to all users priority #1 no
matter how old their systems were.

This facilitated an excellent rapport with our users and we remain in
contact with several of them 3 or 4 years after the business changed
hands- my dad collaborates with a few of them doing long-term
observations of asteroids.

Our choice to make the system as fully open as we reasonably could was
greeted with considerable enthusiasm by the user community. In fact
several people exploited various of the control interfaces to set up
their own automated observatory systems. Other hardware vendors also
used our open source client software to integrate their own functions;
camera controllers and observation sequencers, for example. Since all
our source was open, there was no licensing barrier to interoperate with
our software so people did so. We received updates to our client
software on several occasions when other vendors added hooks into their
systems. I don't think we had students working on our client software
but I'm pretty sure a university prof did.

I will also observe that the closed source vendors we set up interfaces
to were without exception a pain to deal with because of the licensing,
software development kit purchases and no documentation. And given the
closed source, there was no opportunity to make the systems interoperate
better and no motive on the part of the vendors to do so since we were
doing the work to interface to their software.

Now that is anecdotal evidence of the operation of open source in a
market, but short of Kent's unfortunate example is now only the second
piece of actual data that has been presented. Perhaps if actual cases
were presented in this argument instead of rhetoric, we could develop
the theme on a more realistic basis.

Gregm

.

Relevant Pages

  • Re: Herds of cats
    ... Alex Martelli wrote: ... company wanted to acquire control over Delphi, they'd try and buy Borland; to acquire control over Python what are they to do? ... hiring Guido and Alex is probably a step in the right direction;-) but would it be enough? ... The open source projects remain outwith the control of the company; ...
    (comp.lang.python)
  • Re: Secret Source code ...
    ... And that's an application, not a control. ... There's a lot of really good code out there, open source and free in Windows. ... of a PDP-10 operating system whose acronym I've forgotten, ... but that was not an IBM product (IBM had a really good open-source library ...
    (microsoft.public.vc.mfc)
  • Re: Whats the story with the "end of XP"?
    ... your skills are as much on display as they are for an open source ... commercial SW simply because it was FOSS for all the usual reasons... ... I can only assume the FOSS programmers don't actually ... control over your own business. ...
    (comp.arch.embedded)
  • General advice for starting a new open source project
    ... I am thinking of starting a new open source project using PHP/MySQL to ... other applications such as Turbo Lister, Warehouse control systems, ... install packages, use of DEFINE files etc., upgrade packages when users ...
    (php.general)
  • Re: virtual observatory
    ... >of CCD camera is more suitable? ... A "virtual observatory" would seem to be one where the data exists inside a ... There are at least two approaches to control. ... In most cases the remote users access the control server via ...
    (sci.astro.amateur)