[TUHS] Why Linux not another PC/UNIX [was Mach for i386 ...]

Larry McVoy lm at mcvoy.com
Thu Feb 23 07:34:05 AEST 2017

On Wed, Feb 22, 2017 at 02:35:00PM -0500, Clem Cole wrote:
> > I found routines
> > that were bit for bit identical in both in less than 5 minutes.  The one
> > I remember was bmap(), I found a couple of others that I don't remember
> > (just remember there were more) and I gave up in disgust.  I was pretty
> > disappointed that CSRG considered this not AT&T source, it was.
> >
> ???Agreed... the argument... I'm not saying it was correct... was that the
> code for bmap and like was the obvious code and any reasonable programmer
> would have written it that way.???  

So suppose you owned a software company and someone had a source license,
rewrote some of the code and claimed the rest was obvious so it didn't
need to be rewritten.  Now your code is out there for free.  Affecting
your revenue stream.  How well is that "any reasonable programmer
would have written it that way" going to play with with you?  

You are an experienced guy, you know that any reasonable programmer would
have written the very same initial version of your code but that's not
what you shipped.  You shipped debugged, tuned code.  It was debugged
and tuned through years of experience with users, that takes time.
I find it really really hard to swallow that any reasonable programmer
would come up with the same code in a vacuum.

I'll remind you I'm deep into the source management world and I've
repeatedly seen my own engineers want to rewrite code and what do they
want to write?  What was checked in as version 1.1.  All the warts are
from real world experience and they want to get rid of them (until 
they look at the history and understand why the warts are there).

I'd argue that what you'd get from any reasonable programmer is the
naive initial version that works in theory but fails in practice.

In my mind, the BSD guys cheated.  If it were my code, if I owned
that and they tried to make that argument, yeah, I'd sue them as well.
AT&T messed up the suit but I don't think they were wrong.  IP counts
for something, it's expensive to fix all those bugs, it's cheap to do
the initial naive implementation.

Morally, in my mind, BSD was tainted.  Whether it was proven so in a
court of law doesn't change my opinion.  It was AT&T's code to release,
yes, I wanted them to do so as well, but it's their code.  Just 
because we think it should free doesn't make it morally right to
make it be free.  Doing so is theft in my book.

And for the record, I've seen the same behaviour in Linux.  There 
was wholesale copying of BSD licensed drivers and other code into
the kernel, some mumbles about it being dual licensed, but eventually
it became GPL only.  That's theft as well in my opinion.

Perhaps my ethics are a bit too rigid, but they are my ethics and
aren't likely to change.

More information about the TUHS mailing list