Re: [ox-en] Step-change by Free Software
- From: "Niall Douglas" <s_fsfeurope2 nedprod.com>
- Date: Wed, 04 Feb 2004 01:07:24 -0000
On 3 Feb 2004 at 1:08, Rich Walker wrote:
The use of
electronic communication to organise a loosely collaborative
volunteer-led project was a step-change innovation.
I think that you'll find that IBM sites were doing this earlier. Look
up some of Lynn Wheeler's stuff on the internal network, as well as
the stuff about SHARE and so on.
Nowhere near as loosely as Linux. I mean, most of the development
isn't even sanctioned or organised by any central authority at all.
Anyone previous to Linux would have thought it highly unlikely to
work at all, but there it is.
One guaranteed way of robbing all step-change innovation from a free
software project is to add volunteers. If you get enough step-change
innovation in before they arrive then usually they'll refuse to
Depends on what you're doing. If you fork an existing project to
innovate, *and it works*, people will come in.
Only if they think it works. The most frustrating thing is to improve
something significantly and for no one to agree with you or see what
you're on about. Stackless python is a good example here.
ObAcorn: If you release
an interesting but bizarre variant of an existing stable OS managed
with novel weird, ill-documented coding constraints, brag like a
skript-kiddie about the superiority of your model and diss the press
for reporting it without consulting you, then they'll refuse to
Well, I'm guilty of that too - remember my ill-fated Tornado idea,
one I'm still working on eight years later on a different platform,
processor, language and well, just about everything else :)
When a senior elderly scientist tells you something is
possible, he is usually correct. When he tells you something
is impossible, he is usually wrong.
A very apposite quote.
On a more personal note, I have been hawking around my ideas to form
a software venture for 18 months now and if I've learned anything,
it's that you pitch the idea as being as conservative and unradical
as possible. Investors simply will not bet unless they have at least
a 50/50 chance of success from the expert.
Umm; a VC with a 10% success rate is spot on the mark as far as
risk-reward goes. (Which tells you something about the reward ratios
they are looking for)
It's 50% chance the company won't go bankrupt, but it's a 10% chance
it'll pay back more than you invested.
But if you can do your project in <4 man-years, and it will be of
benefit to Man (as opposed to just finding a new way to make money
from people) then there are funding bodies that will fund you.
Strangely, they aren't VC's or bankers.
Quite reasonably, they want a working demo in front of them and
plenty of market validation. Which lets the cat out of the bag. But
then good ideas are plentiful, their good practice is not.