[racket] Clarification on licensing of Racket code?...

From: Rüdiger Asche (rac at ruediger-asche.de)
Date: Sun Apr 8 06:42:26 EDT 2012

I regret that the discussion has taken this strange turn and the ridiculing 
aspects, and I apologize for in parts being responsible for this 

This is not about me (or anybody else) having spare money to spend and 
wanting to finance somebody's home theater out of questionable moral or 
altruistic feelings. It's about business realities and the real world 
outside of academia and its ivory tower surreality.

I look at it this way (and now, after I did receive a number of helpful 
comments in private conversations, which I am grateful for, violate my own 
plea NOT to make this into a debate about open software): Writing software 
is a creative skillful task, and everyone who *is* in a computer program in 
some University actually prepares to make a living of applying this skill, 
generally in the real (business) world which (yet) provides paid positions.

A few years ago, the embedded market (which is what I am working in) still 
worked like this: You designed hardware and specified that the flash or 
Eeprom memory that this hardware provides must accomodate all the software 
that you needed to realize the task this hardware was set out to do. This 
(aside from your own code) covers things like RTOS, connectivity (networking 
and/or serial drivers), GUI libs (if the hardware had a display), self 
programming routines and so on. Also, you had to invest in tools that helped 
you glue everything together - compilers, IDEs, debuggers, and so on, some 
combined with special hardware (ICEs/emulators..), but all of this contained 
software that somebody had to write somewhere. Consequently, there was quite 
a diverse aftermarket of commercial companies who offered all of this - 
needless to say under the condition that the company could make a living off 
it. I as the person who glues everything together made the choice which 
tools/libraries to use - or even provide those myself (I used to do that, 
provide a TCP/IP stack for one of our systems), and the decision was 
strictly a gain-loss type of thing: Which provider has the least expensive 
solution that best suits your needs and can still survive on the agreed 
price. You would then buy all the necessary licenses and price your final 
product by your own expenses and the revenue you expect to make from the 

Now. A lot of the third party companies that provided exactly those 
components have gone out of business because of high-quality open source 
competition - it's kind of a vicious circle; when our competitors use 
Embedded Linux or FreeRTOS along with gcc and other open source tools, we 
just can't compete anymore when our product has to finance in the licensing 
cost for a commercial OS and/or compiler and consequently must reduce the 
costs and stop buying software and instead also go open source components, 
which in turn puts more companies out of business.

So everybody who contributes to Open Software that may be used by commercial 
developers at the end of the day ruins the price of the very skilled work of 
software developers and thus saws off the branches of the tree that 
otherwise may have provided him a job outside the world of academia (whose 
finances mostly depend on arbitrary decisions of funding institutions like 
the NSF instead of real money paid by real people for real products).

I'd rather support the "old" system in which I get paid for things that real 
people really use. That's why my preference is for commercial products where 
developers get paid for something that generates real and genuine revenue. 
Of course I CAN get fantastic support from people like Matthias and Matthew, 
and of course I could "simply" fix bugs myself in OS products by wading 
through the sources, but the question is always whether me doing it would be 
more expensive than somebody else doing it who I pay to make sure that a 
closed part of my system works (and who is so deep into it that he'll find 
the problem much faster than myself). Those two reasons (more efficient 
support by people whose job that is AND working with a company that pays 
their developers) is why I am hoping for some commercial version of 
Racket/Scheme that can sustain its developers and provide me with an 
environment that I can use on commercial terms.

So to all of those who make jokes about me wishing to pay somebody's home 
theater out of thankfulness: Make your jokes now but remember that in 10 
years when you DON'T have a job because 90% of all software will be open 
source by then and the rest is expected to be free as well. There won't be 
THAT many jobs in academia by then that allow you to write software and 
still get paid for it from someone.

----- Original Message ----- 
From: "Todd O'Bryan" <toddobryan at gmail.com>
To: "Neil Van Dyke" <neil at neilvandyke.org>
Cc: "Rüdiger Asche" <rac at ruediger-asche.de>; <users at racket-lang.org>
Sent: Saturday, April 07, 2012 5:14 PM
Subject: Re: [racket] Clarification on licensing of Racket code?...

Alternatively, draw up a contract with the authors of the parts of
Racket that you really want, which guarantees a level of service you'd
be comfortable with. You could probably get by with Eli, Matthew, and
Robby, and they could probably guarantee quick service for sufficient
amounts of money.

Chances are you'd get service that's only a teeny bit faster than what
they already provide and they'd get a new home theater system. :-)

On Sat, Apr 7, 2012 at 6:46 AM, Neil Van Dyke <neil at neilvandyke.org> wrote:
> It sounds to me like you want to use Racket, and that the barrier is that
> you are uncomfortable with open source software in general.
> A lot of organizations have been very comfortable building upon open 
> source
> software for around a decade (a few, much longer).
> Also, I think you'll find that the core Racket developers want to 
> encourage
> intelligent industry use of Racket, and that the license reflects this
> intent.
> You can think it over however long you like. Open source will still be 
> here
> whenever you're ready.
> Or, if you really want, NeilvRacket Professional 1.0 is available for
> $20,000 per Development Kit seat. (We find that the substantial price tag
> confers credibility with the Fortune 500.) NeilvRacket Runtime licenses 
> are
> available for as little as $5 per installation, in quantities of 100,000 
> or
> more. You'll also need NeilvRacket technical support and maintenance
> contracts, since anyone who mentioned "NeilvRacket" on this email list 
> would
> receive more looks of exasperation than the usual assistance. Ask about 
> our
> corporate training junkets.
> Neil V.
> --
> http://www.neilvandyke.org/
> ____________________
> Racket Users list:
> http://lists.racket-lang.org/users

Posted on the users mailing list.