Re: [ox-en] Germ of a new form of society or germ of a new form of business?
- From: "Niall Douglas" <s_fsfeurope2 nedprod.com>
- Date: Mon, 26 Jan 2004 01:40:11 -0000
On 25 Jan 2004 at 18:51, Robin Green wrote:
We're nearly off-topic here but I think discussion of untested forms
of free software development will be of interest to the list. If it
goes much further off-topic, I'll move it private.
It still surprises me that people find the phenomenon of free
software surprising. It would be more surprising if it didn't
happen.
Agreed - _very_ surprising. It was the original form of software,
according to Stallman.
Heh. On old Acorn RISC-OS many programs were written in BASIC and so
came with source. It was understood you would delve into them to
extend or fix bugs within them, exchange copies etc. and it was a
wonderful environment to work in.
Certainly silly issues like deliberately obfuscating code came much,
much later towards the end.
Hmm... what about Red Hat, for example?
You might say they are a special case, and I'd agree with you. Not
many companies can maximise profits for their shareholders while
following a Red-Hat-type business model.
They're not a special case - they make their profits because so many
volunteers donate their work to RedHat for sale for free. Therefore
RedHat have the weird economics of selling something which isn't
theirs but don't have to pay any royalties upon.
Economically similar would be a magazine staffed by slaves, though
ethically there's no similarity.
However, I have been thinking a lot about how to make a profit or at
least a living while writing free software, and I believe that a
privately-held company could take a subtly different approach to Red
Hat and still make a profit (although again, there would not be much
space for many companies of this sort, probably).
RedHat must walk a very fine rope. They can't piss off those who
admin the projects they collate to form their releases and so often
when those projects are doing really stupid things, RedHat must prod
very gingerly indeed. For example the hitherto loathsome state of
thread support in the Linux kernel was one of the few times RedHat
just gave up on prodding and took it into their own hands to produce
NPTL.
One thing RedHat cannot do is start smelling unfree. To do that would
lose them all their karma and suddenly many hurdles would magically
appear. OTOH, their shareholders like all shareholders want profits
and they don't care who you must crush to get them.
I would very much like to be paid to work on a programming project of
mine, a new programming language incorporating a number of "step
changes" which would put it head and shoulders above Java and C++.
Is that like the "D" language?
Hmm, I have ideas here too but my language would surpass Haskell in
terms of what a single statement can achieve! But that comes much
later.
However, I am only willing to enter into such an arrangement on
condition that it all be released as free software (possibly after a
fixed, short time of being proprietary software, say 3 years - but
ideally immediately). Call me crazy, but that's my terms.
I feel all software should become under a LGPL style license after no
more than two business cycles (roughly five to seven years).
Typically, investors will only invest in a company on the
understanding that the company will attempt to maximise profits -
within certain regulatory or contractual - or risk-averse -
constraints. However, occasionally what one might call an "angel
investor" (if the phrase didn't already mean something different!!)
can step in and agree to the company being allowed to follow various
ethical constraints. Such as releasing everything as free software.
Even if those ethical constraints would reduce the expected
profitability of the company.
Very, very few investors will agree to the GPL. It permanently locks
you out of flexibility which you might suddenly need later on when an
opportunity arises. Of course if you prevent anyone else altering
your GPL code this isn't a problem. However, why then not use the
LGPL which can be converted to the GPL by clause 3 yet keeps all your
options open?
However, it could in principle work with a for-profit company, or a
not-for-profit co-operative aiming for a return on investment (as
opposed to charity-type-things like the OSAF or the Mozilla Foundation
which just spend money and do not expect to sell anything
significant). It's not unthinkable for an investor to come along and
say "I'm willing to do a part-philanthropic, part-capitalistic, deal
here: give me a cut of your profits as set out in this here contract,
but in return I won't prevent you releasing anything as free
software."
I speak from experience when I say that persuading rich people to
part with their money is like squeezing blood from a stone. You must
jump through loads of hoops first such as market validation,
assembling execution intelligence (ie; a well-rounded team) - all
this is good, but it's bloody frustrating when you'd just like to get
on with the coding! (But if that's all you want to do, you don't
actually want patronage this way - better go become an academic). See
"New Business Road Test" by Mullins or "A Good Hard Kick Up the Ass"
by ... arse, I forget the name. It's easy to find on Amazon with a
title like that.
And of course the profits can come from the familiar area, services:
support, training, etc.; and trademark licensing, potentially.
Try selling free software to a mass market. You will fail. You can
only sell free software to a business market profitably and therein
lies a major problem - how to improve the lot of the home user who is
lumbered with Microsoft.
Indeed and it's why I am a post-capitalist. I think I know how free
software - all software - gets written and I think I can base a
revenue model upon a system which combines the best parts of the
volunteer system with a touch of entrepreneurship. Entrepreneurship
far predates capitalism, and in my mind any successful post-
capitalist economic model will perhaps place entrepreneurship even
more at its heart than it is now within capitalism.
The result I would still call free software though it won't fit the
OSI definition. Put it this way - it's close enough and I don't
think anyone other than the GPL zealots will complain.
Fascinating - care to share the general idea, or is it top secret? ;)
Hardly - ideas are cheap, practice is scarce. If you read
"Information Rules" by Shapiro and Varian it gives an ideal economic
strategy for making profit in an ideal world. Because they assume
ideal conditions eg; people not going to astalavista.box.sk and
downloading a crack, the book's proposals are flawed.
However, they still have good points to make. For example, an
information good is worth only that which the customer is willing to
pay for it. Note it's their *perception* of worth. So how do you
maximise their perception of worth?
Easy - you make them think it's more valuable than it is by widely
disseminating it. For example the Bible is valued precisely because
it *is* so widespread - just as Windows compatibility is so valued
because everyone else has a Windows box or can emulate one.
Since this is the case, you should aim to charge very little for your
software and maximise the purchaser's ability to find value in their
purchase - and most importantly of all, their ability to *create*
value in their purchase of use to others. Thus you get a networking
effect whereby the value or perceived value of the product escalates
due to the users sharing improvements and the feeling that you're not
alone on this.
This dovetails in nicely with my views about maximising step-change
innovation. The ideal programmer should not concern themselves with
creating the next big thing - they should instead lay the foundations
for the next big thing and let others build it for them. Linus
unwittingly did this.
Under this model, the customer would pay maybe five euro for the
convenience of downloading the software right there and then instead
of having to bother searching for a crack. Advertising is important
for this kind of customer as they can't escape generating profits for
you (and thus you must structure your software so as many people as
possible view your site). You should supply no direct support, no new
versions - use the community model so successful for conventional
free software. If they want more support, bug fixes then they must
pay a support contract - which is something most businesses will do
and here you can make bucks.
But what's important is that the user must be able to modify their
purchase and distribute those modifications to others. Here's where
the non-compliance with the OSI definition comes in - it has all the
freedoms of free software except that you cannot redistribute the
original software itself. You accept that people will ignore this as
they do now with ?100 software but since five euros is a token amount
for most people, if they like the software they'll tend to pay that
for an immediate download rather than expend the hassle. If they
creates mods for your software, they will probably exchange them via
the community forums you will have set up and needless to say, you
can scavenge good ideas from those thus ensuring a frequent stream of
upgrades.
If you work on this basis for your language, you could make a lot of
money and improve the world while you're at it. Unfortunately you
will need to get the thing written to a usuable level first.
Cheers,
Niall
_______________________
http://www.oekonux.org/