Re: [OT] mothers, free software

[Jonathan S. Shapiro]

On Thu, 2005-11-17 at 20:43 +0100, ness wrote:
> Hi,
> 
> This day I had a really enervating discussion about free software with 
> my mother. Finally I got her to claim the following statements:
> 
>   o  One cannot make money with free software
>   o  Free software cannot provide (data) security
> 
> She also called called it fraud if a company sells support for a free 
> program.

The first statement is entirely correct. You cannot make money with free
software. You can make money with services, support, reputation, and
relationship management around a free software base, but you cannot make
money with free software per se.

However: you cannot make money with proprietary software either. The
only issue that changes with free software is whether a competitor can
turn your NRE costs against you. In the overall scheme, this is a very
small issue.

The fraud statement is just silly. Counterexample: many proprietary
vendors sell software and separately sell support contracts. This is
because the cost of the software is basically fixed but the cost to
support depends on the number of users. This is considered standard and
accepted practice. It certainly would not be fraud if that company
decided to give the software part away. Free software is no different.

In order to respond to the data security challenge, I'ld need to
understand more precisely what it is that she means by data security,
and why she believes that it is impossible.

> Oh, and she doesn't believe you get money for building coyotos, 
> Jonathan. Maybye it would help if there was sth. official that showed 
> your research contracts (especially those you get/got money for and that 
> result/resulted in sth. free).

While the software is free, many clients don't want the work disclosed
until they are ready to disclose it. Some clients *never* want their
involvement or the dollar amounts disclosed.

For EROS, the following contracts involving Johns Hopkins are a matter
of public record:

  Panasonic Research Collaboration (Exploratory: $40,000)
  EROS-Based Confined Capability Client  (DARPA: $852,596)
  Secure Development Environment (Air Force: $796,848)

Outside of the US, Panasonic is known as Matsushita.

I can also say that The EROS Group, LLC (my company) is currently a
subcontractor for one of the phase-I SBIR awardees on DARPA solicitation
SB052-011. The second awardee for this solicitation is also using an
EROS derivative, and is contemplating cutting over to Coyotos. The EROS
Group is not a subcontractor for the second awardee.

Since most people don't know, the SBIR process goes:

   Phase I:   Feasibility Study     < $100,000 total
   Phase II:  Prototype             < $750,000 total
   Phase III: Production Contract   (varies, typically $millions)

So financially speaking, getting a Phase I is not a big deal (even less
for me, since the primary contractor cannot subcontract more than 30% of
the work). I do expect, given the people involved, that this one is
likely to go to a phase II. Phase III remains to be seen when we get
there.

But the interesting part about the SBIR requirements is that you cannot
get one of these unless you have a credible *commercial* customer for
what you are building. That part I can't say much about at this point.

>  Quotes I found in short time:
> 
> > Yes, my money actually *is* where my mouth is. But I must say that
> > money does not taste very good, and I would prefer a normal meal.

This is a reference to the investment that I, personally, have made in
this technology on a private basis. Conservatively, that investment is
approaching $2M now.

shap