Lennart Poettering: BSD Isn't Relevant Anymore

[Aryeh Friedman]

On Mon, Jul 18, 2011 at 7:30 AM, Jerry <jerry at seibercom.net> wrote:
> On Mon, 18 Jul 2011 15:47:24 +0700
> C. Bergström articulated:
>
>> I wish people would spend as much time solving problems in *BSD as
>> they do trying to defend an irrelevant OS ;)
>
> Personally, I wish they would spend more time in developing fully
> functional wireless drivers as opposed to simply bumping the major
> version number every 18 months +/-. I have two new laptops ion front of
> me that I cannot use FBSD on simply because they don't support the
> wireless (N class obviously) installed in them.
>
> I simply refuse to purchase a machine to accommodate an OS; nor will I
> attempt to change the wireless network card/chip for the same reason.
>
> OK, now the usual group of "blame the manufacturers", blame Microsoft",
> blame everyone else for the problem are free to chime in. I was
> seriously considering hiring a professional programmer to write drivers
> for devices for me; however, it was then I remembers something I
> learned in business school, class 101. I weighted the cost of
> developing the drivers as opposed to simply purchasing an OS that all
> ready had those drivers readily available. Guess which was "many" times
> cheaper? Cost analyses proved that developing my own drivers was not
> cost effective.
>
> I suggested several years ago, and I will re-suggest that FreeBSD start
> a program that would allow programmers to be paid to write code that
> either the regular contributors do not want to write or are not capable
> of writing. Other OS's are currently working on that model. No one
> would be forced to contribute. This would prove beneficial to everyone
> and should satisfy both capitalist who don't mind paying for quality
> products and socialist like Poly who want everything for nothing. It
> would be a win-win situation.
>
> With the advent of the next version of FBSD soon to be upon us,
> this would be a propitious moment to start such a project. FBSD has
> never been considered a dreadnought in the driver development field and
> this might work to change that. At the very least, it would
> create a brouhaha among others although the pigeon-livered average
> FreeBSD user would probably abstain from support this project either
> from a lack of need or indifference to others or basic socialist/fascist
> concepts.

The issue your talking about is actually caused by a fundamental flaw
in *ALL* pure open source projects namely in return for the freedom to
look at the code and stuff we give up market forces.    If there where
real market forces then *SOMEONE* in the larger freebsd community
would find it profitable to write such drivers (and other needed
unglamorous but necessary tasks).    A model I proposed (with about 3
others from different FOSS backgrounds) a few years ago is still as
relivent now as it was then despite the lack of reconition that it
allows for all the freedoms of open source but without the neglecting
of user demands (i.e. market forces).   The model is actually really
simple: the source code is freely available to *ANYONE* for
study/research/evaluation/educational *BUT* the minute you compile it
becomes economically valuable to the user (assuming that there is no
value to the above free uses [it is not a bad assumption if you look
at it]) and thus *MUST* be paid for.   Now the one small twist this
has over any other model is that with basic (but careful) planning it
always for anyone who has contributed to get their fair share of any
revenue.    Think of it as the idea that everyone must contribute to
the project either with money and/or work to improve it (let their own
enlightened self interest dictate what mix they choose).     The only
legal hurdle to this is that the OSD definition of open source does
not allow for licenses to "descriminate" between classes of users
(people who only read the source code but do not use it and those that
do use it).