Wake up to reality

[Mark Linimon]

(current- removed since this is IMHO only of interest to small@)

On Tue, May 30, 2006 at 09:55:17AM -0700, Michael Sierchio wrote:
> VxWorks?  Every bit of userland code reveals its origins in BSD, esp.
> FreeBSD.  A scan of any of the older fiber channel switches identifies
> nearly every service as being from FreeBSD (and ancient revs, at that).

As someone who contracted at VxWorks in the early days (I have a 3.21
manual around here), I would say that yes, technically, early VxWorks
was based signficantly on the historical BSD codebase.  I'm sure there
were imports from BSDi when they owned it, and FreeBSD before and after.

But the market _perception_ is that VxWorks is not a FreeBSD derivative.
The best (IMHO) you could do in an argument was to make the case that
they share a common lineage and there has been much borrowing.  That's
not at persuasive a case.

Now add the fact that the "modern" VxWorks marketing idea is to push
Linux (and not BSD) and you have your work cut out for you.  Whether
or not that's because they reached a plateau with how many seats they
could sell based on "just like Unix", or their pricing strategy, is
irrelevant to FreeBSD's choices (also IMHO).

Lastly let it be noted that for reasons of performance and predictability,
the traditional codebase was designed to run on either the VRTX kernel
or one or two other choices (though 90% of the people used VRTX.  They
later bought its author.)  Due to licensing issues ($) they decided to
write the Wind kernel (John Fogelin's original code).

I have not tracked the progress over the years but at least in the early
1990s timeframe they were not using a BSD kernel; instead, some kind of
layer over the more strictly designed real-time kernels.  So for one of
the most important pieces, it was _not_ a derivative; it was aiming at
very different markets.

Summary: I think it's going to be pretty tough to make the case that
modern VxWorks is a "BSD derivative".

mcl