Re: [Nolug] BSD vs GPL

From: Mark A. Hershberger <mah_at_everybody.org>
Date: Wed, 15 Jul 2009 20:39:56 -0400
Message-ID: <87d481ihpf.fsf@everybody.org>

James Thompson <james@plainprograms.com> writes:

> A lot of that has historical roots though. Linux came up as a "clean
> room" style implementation of *nix tech. While Linux was being
> developed in the '90s there was still a lot of concern about the legal
> status of the BSD codebase in terms of ownership. So while Linux was
> building momentum the *BSD community was dealing with legal questions
> pertaining to the ownership of their code.

Yes, Linux was “clean room” vs BSD's more organic development. But BSD
had a 12 year head-start on Linux, an established userbase, built
directly on a history of 25 years of UNIX. Throughout the 90s, BSD's
maturity and stability was touted compared with Linux poor memory
management and need to reinvent everything. People ran Linux servers in
the 90s, but that wasn't common. SysAdmins had a deep and abidng
respect for *BSD throughout the 90s. Yet, somehow, here we are almost
20 years since the birth of Linux, and the upstart has over-taken the
kernel with everything on its side.

A quick recap of history:

According to Wikipedia: “The (USL v BSDi) lawsuit slowed development of
the free-software descendants of BSD for nearly two years [starting in
1992].”

The lawsuit was settled in January 1994. Where was Linux at the
beginning of 1994?

Lets look at two areas: books and companies.

By 1994, a number of books had been written about Unix. One good read
was “A Quarter Century of UNIX” (1994, of which I own a copy).

But beyond that there are these:

    UNIX Papers for UNIX Developers and Power Users (1987)
    The Design and Implementation of 4.3BSD UNIX Operating System (1989)
    UNIX SYSTEM Readings and Applications (1987)
    The Unix Programming Environment (1984)
    UNIX, POSIX, and Open Systems: The Open Standards Puzzle (1993)
    Practical Internetworking with TCP/IP and UNIX (1993)

(and on and on and on)

Remember that for UNIX had been around in some form for 25 years. BSD
variants had been around for 12 years prior to the lawsuit.

Linux had exactly one book by 1994:

    Linux Installation and Getting Started (1993)

By 1994, there were a number of companies producing commercial BSD
variants. Remember, the lawsuit only affected people trying to use BSD
without a Unix license — 386BSD and BSDi. Most companies had no problem
purchasing licenses. A few of the variants:

    SunOS
    Dynix
    NeXTStep
    Concurrent

In contrast, there was a single commercial Linux distribution:

    Slackware

Red Hat was around, but was only selling “software accessories”
according to http://www.redhat.com/about/companyprofile/history/. They
didn't create a distribution until October 1994.

> And at some point the Linux kernel got big-business buy-in behind it
> which only accelerated its momentum even more.

But BSD had commercial buy-in before then (SunOS, Dynix, NeXTSTEP,
Ultrix, Tru64 UNIX) and still lives on in embedded devices (many
firewalls and routers) as well as OS X.

Perhaps when you say “big-business buy-in” you mean companies like IBM
started supporting Linux? They'll go where there customers are.

Really, if IBM was making a choice between freely-licensed kernels (BSD
v. Linux), they would have chosen BSD. The “big-business buy-in” was
driven by smaller companies and the free software developers they
employed, not big businesses. “Big Business” needs to make a buck and
goes where the customers are. Obviously they see more customers using
Linux.

Mark.

-- 
http://hexmode.com/
Every day, mindful practice.  When the mind is disciplined, then the
Way can work for us.  Otherwise, all we do is talk of the Way; everything
is just words; and the world will know us as its one great fool.
___________________
Nolug mailing list
nolug@nolug.org
Received on 07/15/09

This archive was generated by hypermail 2.2.0 : 08/06/09 EDT