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

[I've boldly changed the header rather than have this intertwined with
an unrelated discussion. IMO it's worth the community periodically
confronting this as a integrated part of language issues, since the
question always exists at the end of developing a piece of software:
how do I deploy it, not just technically but commercially. As such,
I don't think this thread is off-topic. If you do, feel free
to ignore the thread.]

Ken Tilton <kentilton@xxxxxxxxx> writes:

Rajappa Iyer wrote:
Pascal Bourguignon <pjb@xxxxxxxxxxxxxxxxx> writes:

...Stallman is an idiot? Nothing wrong with that, I am glad for the
company. Just can't believe so many people think they are getting
something for "free" when in fact all the FSF does is cripple the
economics of the software industry and cost all of us better tools.

Perhaps we'd need hard data backup (anyone has the IP of the gateway
to parallel universes?), but it seems to me the software industry is
much better today with freedom software than it was twenty years ago
with only proprietary software.

I don't know who said that, nor why they think this.

The "software industry" is a commercial enterprise. A commercial
enterprise makes money on something one competitor can do that another
cannot. If Food Company A has seven ovens Company B has five, then
Company A has a strategic advantage. But if both have five, and you
buy five more for each compahy, then probably Company A and Company B
cannot make a dime more than they could before. The price of food may
come down, but Company A and Company B are not any better off. That's
how it is with free software. Free software adds to what programs can
do, but it doesn't add to the "industry" at all. The industry simply
can't afford not to use every piece of free software offering it's given,
and consequently will
(a) mostly never develop any other way of solving that problem
[because the cost of developing an alternate solution is non-zero
and the economic system already models the cost of dealing with
that problem as "zero" so has no more money to devote to that]
(b) mostly will require more complex solutions to problems in order
to make money (this will put uneducated workers out of business
since they can't compete, will force re-education of moderately
educated workers just to keep up with the economy at current levels).
Does it benefit the end-user? Yes and no. It kills competition on the
underlying issue, so if they like the free component that everyone now
depends on, then it helps them. If they wish someone would develop
an alternative, they may wait forever. And if someone ever creates a
better one just to be nice, will that person be rewarded? No, only the
people who deploy it will get the credit, and they'll have to pass
none of it along to users. Is that freedom? Not in my book.
Is it an engine for innovation? Not in my book. It works really well
at the beginning when the people motivated by a mistaken view that
the pre-existing system (the one that rewards innnovators) is still in
place are still cranking out new products... but when those people fall
off and everyone realizes the only place to make money is working for
big business, and that the little guy can no longer make anything on
his own, I don't think anyone's going to like it. But by then it will
be too late. If it's not already. Mostly, I think it's too late already.
The damage is done and I don't see forces in place to repair it.

Quite. Kenny is hypothesizing about the displacement effect, but the
evidence so far seems to be against his thesis. If one were to find
several promising tools and/or systems that vanished solely because
they could not compete with free software, then he might have a
point.

How exactly do you find vanished tools? Kent will occasionally speak
up about AServe and maybe cl-http (I forget which) hurting his chances
with a similar tool. How will you identify the others, or (vastly more
challenging) those never attempted because something like FTGL or
ImageMagick or SuperCollider is freely available?

My sense had been that CL-HTTP didn't hurt my chances because its
intellectual property pedigree was questionable enough that no one I
knew was willing to incorporate it into a commercial product. Also,
CL-HTTP was more heavyweight than many wanted, and so even if the
IP issue wasn't the key problem, it didn't seem to be catching on
in the commercial world enough that I felt it was a threat.

I actually just wanted to write a book on using Lisp in commercial
settings but I felt there was a major gap in that there was no web
server. Rather than write the book, I made what I now regard as a bad
business decision to spend my time and money on the web server
instead. Just as I was nearing completion, with no hint that it had
been on the horizon, AServe was published. This meant my valuable
time that could have been spent on a book had been squandered, and my
time that had been spent on the web server was no longer valuable.

I would not have begrudged any entry to the market in which the web
server had sought to recover its development costs. I don't know what
anyone else would have spent developing a server, but I figured they'd
either spent more and gotten more power (in which case I could hit a
lower price point and offer less) or they'd spent less and gotten less
(in which case I could occupy a higher-end). But zero has the special
multiplicative property that it's hard to compete with. Free software
is the Limit Commodity.

Moreover, as a result of this, HyperMeta was effectively dead in the
water as a software developer. It still exists, but it turned to
consulting, but the consulting never achieved sufficient revenue to
produce interesting software. What Stallman's story about commercial
software overcharging the community neglects is: (a) normal
competition already brings prices down pretty fast to a reasonable
price and (b) some degree of profit is necessary to a business in
order to allow people to live (health care, vacation, etc) and to
yield new offerings (you have to invest not only in things that might
profit, but things that might fail).

So, you're free to conclude that there's nothing interesting I might
have ever produced if I'd been given enough money to do so. But if
you think I've ever had an interesting idea that it might have been fun
to see me pursue, then it's not a huge stretch to say that the
introduction of Aserve led directly to my not getting to explore the
other ideas I might have had for how to go interesting places with
web-based Lisps in future releases. Why should I bother? The culture
has spoken clearly: We don't want innovation--not if it costs us a dime.
We want to pay ONLY for specific solutions to specific problems that
business wants solved, and as far as individual personally driven
creativity, we will operate at every point to sabotage the chances
of anyone making a buck. ... all in the supposed name of freedom.

And you're free to conclude that I'm just whining about lost money, but
I'm not. I didn't make the money I thought I should have, and indeed
that annoys me. But what really annoys me is my lack of ability to BUY
the software I want, because now I can only buy from BIG BUSINESS.

I can HOPE that free software will develop what I want, but I can't
make it happen without funding the entire development--something I
can't afford.

I can HOPE that big business will develop what I want, but they mostly
care about things the community at large wants, and they won't invest
in niche markets like, say, productivity tools for smart programmers.
There's more money in "yet another realplayer clone" because there are
ways that the intellectual property channels or advertising channels can
be leveraged.

I can develop it myself at enormous expense since I have to do all the
work that I probably don't have time for. But then I have to give it
away and still have no hope of getting the next thing I need because
no one will reward my effort "in kind", or else I can try to sell it
and wait for a "free software sniper" to re-implement it and try to
undercut me for no other reason than to make sure I don't make back my
investment or... shudder... make enough profit that I could afford to
try inventing something else that's cool.

So what really bugs me is not that I can't sell innovative things,
it's that I can't buy them. Because the things I want are not being
produced any more because everyone I know who'd like to do these kinds
of things knows it's a losing proposition any more.

And who says the tools have to be vanished? An open question is where
Lisp environments now would be now if so many of you were not spinning
your wheels on free Lisps. Hell, a few of them are scraping by
anyway. If certain Lispniks had half a brain and simply bought a
commercial license, maybe franz could stop pricing ACL like hen's
teeth and ease up on their whacky licensing fees.

I agree completely.

(Which makes it a bit ironic in the context of my grumbling about AServe.
But I don't think Franz meant to sink my business with AServe--I think they
just didn't understand the full implications of what they were doing.
I like to think they've learned from that, although I don't know if they've
actually changed any policies on free software. My company certainly changed.
It started out hoping to create and license interesting software, and has
largely abandoned the idea that it can make money that way.)

I used to have a neighbor who would sometimes cut my lawn for free
because my lawn was small compared to him and he didn't mind doing me
the favor. But I used to insist on paying him anyway. The issue
isn't the favor--favors are well-intended. It's the bookkeeping. He
can't trade a favor he's done for me to get something from someone
else. I told him, if I do something for you, I'll bill you back and
you can give me back the dollars. But if I don't, you've got the $10
that says you did something for me, and you can trade on that with
anyone, not just me.

And there's nothing wrong with making a profit because if others charge
back, they'll make their profit, too.

What free software really rewards is people who don't contribute. It
gets them a ton of stuff but requires nothing from them. Moreover, it
fails to compensate those who have contributed a great deal from people.

But nooooooooo, you geniuses have to save all that money and stagger
along with crappy tools at a third my productivity, losing all
that... oh. no, I forgot, Stallman says your time has no value. Or
that you are evil for expecting anything for that value. Super.

Indeed.
.

Relevant Pages

  • Re: the free software paradigm [was Re: Amazon used lisp & C exclusively?
    ... Suppose your answer is "you can't get money" or "don't expect enough ... a free software sniper instantly de-value innovation and claim its ... And if you don't like my inspiration, ... Free software can't preempt you there, because its programmers don't have access to the hardware before you do. ...
    (comp.lang.lisp)
  • Re: Lets talk about GUI and sound libraries
    ... as you write Cells, which is free software. ... Put another way, if SBCL and Lispworks cost the same amount of money, ... If Linux and OS X cost the same amount of money, ...
    (comp.lang.lisp)
  • Re: my opengl progs - ZX & nonZX themed
    ... I pay for my Free Software so I'd appreciate it if you laid ... Who did you pay, and for what? ... more complicated support issues than the suse support covers anymore. ... > a brief moment, about what money is and why we have it, you will see ...
    (comp.sys.sinclair)
  • Re: ILC2005: McCarthy denounces Common Lisp, "Lisp", XML, and Rahul
    ... and what you mean by "free software". ... staff and offices and computers, ... that same military money was enabling their contribution. ... notice that essentially none of these innovations ...
    (comp.lang.lisp)
  • Re: My mite (2 small coins)
    ... *acquiring* a software package for no money, ... The total purchase cost of software ... represents about 2-3% of my business expenses each year. ... depends on which free software and which non-free software you are using for ...
    (borland.public.delphi.non-technical)