> Thanks for the interaction. I'm learning a lot.
As in all intellectual pursuits, the interaction helps both of us ;-)
> Let me try to resume this, so I can avoid my tendency to digress. Sorry
> for the repetition of somethings, but I think it's needed.
Yes. The repetition keeps us focused.
> Your premisses, as I see them:
> (1) Free software development is more efficient than proprietary
> software.
Yes.
> (2) Software should be free because it can enforce controls on people.
Yes, but as with RMS, I would put (2) first, since the moral reasons are the
most important.
> (3) To become the only kind of software development, free software
> development needs to find a way to get money to developers.
In line with putting (2) first, (3) is mostly true, but not completely,
because if you have enough moral people who will fight for freedom, then you
don't have to pay them to win. However, in the real world most people are
utilitarians, not philosophers, so paying them gets us the needed critical
mass. Obviously, it also makes it easier for the moral ones, too, because why
suffer for the sake of doing what's right, if you can do what's right and be
paid for it too?
> (4) The market isn't working right now because the companies are gaining
> money from free software are not using that money to reward developers.
The major reason the market is not working is because proprietary development
is the current paradigm. Most of the world's machines are run by proprietary
code. Given (1) and (2), this is both morally wrong and inefficient. Being
morally wrong was less of a problem until the Internet made all the machines
interconnected and therefore a powerful force of control. But at any rate,
the paradigm must change to free development going forward for reasons of (1)
and (2).
Now, the problem with going to a completely free paradigm is that 4 million
developers working in proprietary need to be paid, because they (like current
free software developers) need to support themselves and their families, and
because they will fight the paradigm shift otherwise, since the change will
adversely affect them financially if they are not paid.
Also, it is hard to believe that even the current free software developers
would not be better off being paid to develop free software. If they are
working somewhere else to personally fund their free software development,
they at the very least can work on free software full-time instead of having
to take another job or have more personal free time.
> (5) The way to correct the market is to create an organization that
> sells free software and rewards developers. This company would have a
> worldwide scope.
It doesn't have to be worldwide in scope, though there is no good reason for
it to not be worldwide. The structure could work regionally, too, because
there will be great efficiency gains just going to that size. But it is best
worldwide, because the community is worldwide.
> The premisses I think are not correct are (3) and (4).
> In (3) I'm not
> sure if I want a single type of software development, I believe
> diversity is beautiful (I believe this is another discussion).
It is not clear whether you object to free software being the only paradigm
or that you object to developers being paid for free software. I assume that
it is the former by your statement of diversity.
Remember the statement, "moderation in all things, except moderation". If
free software is superior for reasons of (1) and (2), why have a different
paradigm? Diversity is a beautiful thing, but not in all things, only in
things that ought to have diversity.
Do you want diversity in justice, so injustice is allowed? What about
diversity in government systems - is tyranny a good thing, because you want
diversity with democracy?
These are superstructure items and you don't want diversity with them. In
fact, they are paradigms and only one can exist at a time. But good
superstructures allow diversity within them. The diversity works within them,
not with them.
> My real
> problem is with (4). What I've seen of the market is that more and more
> companies hire free software developers either letting them work on or
> with the purpose of working on free software projects.
Companies are of the Industrial Age, so they see everything in terms of
exclusion and property. If they are adopting the free paradigm, it is only as
a strategic, competitive move, because free software is a means for them, not
an end. If you look carefully, most companies are not freeing their crown
jewels, only their also-rans. They do this to compete with the market leader,
hoping to topple the market leader. The issue is that if they are not
philosophical about free software, in that it is a means not an end, they
will close the code once they achieve their purpose. This is a real threat
that only becomes apparent after they achieve their purpose.
We already see some companies saying that they need to close code for
profitability's sake.
> As I see it, then there's the market. FSMC (Free Software Marketing
> Company) will have to be extremely effective at marketing if it wants to
> be a world-wide software supplier that, according to you, will compete
> with a lot of companies that don't have that cost. Simply put, what will
> be this company market advantage ? After all all the system depends on
> the success of FSMC.
I think you misunderstand the FSMC. The FSMC is the balancing point between
the developers on one side and the marketers (assuming independent
consultants) on the other. Both will be large networks. In fact, probably
they will be mostly overlapping networks, where most independent consultants
will be on both sides, as marketers and as developers.
The point of the FSMC is to maintain the balance, so a strong marketer (which
you were correctly concerned about in a prior email) does not get the market
power to control the rest of the marketing function, and most importantly,
get control of the developers, who are the most vulnerable. So the FSMC is a
collective solution to protect themselves. This is the same as a democratic
council protecting itself from strong leaders, who would otherwise try to
overcome them.
A correct structure balances off all the forces, so that no one person or
force overwhelms the others. Without the FSMC, there will be marketers who
become too strong and then will overcome the other marketers and once that is
done, will overcome the developers. At that point, you are back to a
Microsoft, which is a natural outcome without a properly balanced,
postindustrial structure to countervail the natural tendencies of the
marketers.
> To me it seems that, as described, FSMC depends on getting a
> distribution monopoly on some free software projects. This limitation on
> distribution reduces the effectiveness of development by separating
> users from developers, putting marketers in the middle of that
> interaction. This introduction causes delays in the feedback loop that
> we agree to be an important feature of free software development. This
> means effectively reducing free software development efficiency.
Again, I think you misunderstand. The marketing will be done locally. The
FSMC is to keep the local marketers contained so they don't get too much
power to take advantage of the rest.
I suppose the confusion is also because the FSMC does 2 things which are both
related. One is ensure that the costs of development is included in the price
of the marketers, so that revenues flow back to the developers as salaries.
The other is to maintain another balance, so that no marketer uses market
power to get control of the developers and the code again.
There are other forces that need to be balanced too, like investors/financial
markets and hardware companies. But these are the same forces for traditional
companies and are on the diagram.
> The reason why I don't think the market is as gloomy as it seems is that
> the number of companies that have been working with free software is
> increasing, and most of them are not playing by corporate rules, but by
> the community rules. Those who don't have less support by the community
> and end up paying more for their product development, so they have to
> ask more money from their customers.
Yes. But the traditional corporate structure has built-in conflicts, because
the interests of shareholders has to be paramount, under law, to all other
groups, including the community. So as long as helping the community helps
the shareholders, they can do it. But once the interests of the shareholders
conflict with the community, the officers and directors are duty bound to
look after the shareholders, not the community. If they resist, they will get
sued personally by the shareholders. Even if some refuse and quit, there will
be others who will come forward and do what is in the best interests of the
shareholders. So the community should expect to lose in that contest with any
traditionally structured company.
The place where the conflict will show up is in closing the code, because
proprietary code is always better for the one company once its code is the
market leader. Closing the code is bad for the community, but good for the
one company. So the company will be obligated to close it at some point,
since there is not a countervailing force.
In the CommCo, the countervailing force is that the community owns the
company, so the community can resist the temptation, because the interests of
the community and the shareholders now mostly overlap. If the officers and
directors want to close the code, the community can elect new ones that look
after the interests of the community. That is why it is called a
CommunityCompany.
> An example of this, as far as I can tell, has been Caldera. They had one
> of the first "graphical" linux distributions, but they've fall beyond
> most of others in market share. My point is that this happened because
> most of its value-added software was not free and not even
> distributable.
They are also the ones that prove my point, since Love has always said that
free software is not the way to go and is now moving more towards closed code.
If and when we see major proprietary companies releasing all their code
including their market leaders under the GPL, I will take another look at
this analysis (which I believe will never happen). But until then, my
presumption is that all traditional companies will use the free/open
community to dislodge the current leaders with their second and third tier
products and then lose their religion at some point and bring us back to the
same proprietary paradigm with only the names of the market leaders at the
top changed.
In a way, this is exactly what Microsoft did in the last cycle, but within
the proprietary paradigm. It used the developer community to dislodge IBM and
then consolidated (and is continuing to consolidate) its power once it became
the market leader. Now the community is trying to find someone else to topple
Microsoft. We can either repeat the cycle or understand that the problem is
the structure not the people working at the top of the structure.
If a system has a throne, someone will mount it. It is not the king who
creates the kingdom. It is the kingdom that creates a king.