tsarna@endicor.com (Ty Sarna) wrote:
> [blurb about NetBSD's config mechanism and flame snipped]
   
   Mr./Ms. Sarna, have you ever actually looked at the arch/vax subtree of
the NetBSD kernel source? Until you do examine these sources (I mean
examine intelligently, with a decent degree of VAX hardware knowledge, not
just stare at the C code through the eyes of an i386er) this discussion is
pointless. Sure, NetBSD's config mechanism can do miracles, but the uVAX
portion of the present-day vax port is a conglomeration of hacks that allow
it to run on the (very limited) set of machines that the developers
happened to have at the time it was written, but make any expansion
virtually impossible. Note that I'm talking about the uVAX portion only
here, Ragge's vax port in general is better on average. Also what Ragge is
currently doing seems to me like a total restructuring of this code, so
there is a good chance that it will soon improve dramatically.
   
> You go around badmouthing NetBSD here and on other lists [...]
   
   I am NOT badmouthing NetBSD. It is true that it is not my OS of choice,
but this is a technical discussion list about NetBSD/vax, not about choices
and preferences, so I keep mine to myself and believe that everyone here is
mature and responsible enough to make the choice for him/herself. When I do
have to make a negative remark about some aspect of the present-day NetBSD
port, it is not because I want to bash NetBSD, but because it would be to
everyone's benefit if I say the truth, rather than a sweet lie that would
please NetBSD devotees. Also note the "present-day" part. No one and
nothing prevents you from doing your port of your favorite OS right.
   
   Also keep in mind that I'm not denouncing NetBSD in comparison to any
other OS. Presently there is NO version or flavor of UNIX that properly
supports all uVAXen. Ultrix's uVAX code is not any better than NetBSD's,
although the exact way in which it's screwed up is quite different. As for
Berkeley UNIX(R), it presently has practically no uVAX code at all (the
only supported ones are KA630 and KA650, the earliest and the closest to
big VAXen).
   
   I have decided to give this task a shot myself, writing the uVAX code
the way I believe it should be written almost from scratch. Given that none
of the existing VAX OSes have any satisfactory uVAX code, from the
technical perspective there is no real difference which one to use as the
base. I have chosen 4.3BSD-* simply because it suits my personal
preferences the best, and in the world of volunteer-written software the
one who does the work sets the rules. But then no one and nothing prevents
you from porting my work to NetBSD. NetBSD is a Berkeley UNIX derivative in
the first place. Oh, and since my work will be 100% my own, it will be
totally free.
   
   Sincerely,
   Michael Sokolov
   Phone: 216-368-6888 (Office) 440-449-0299 (Home) 216-217-2579 (Cellular)
   ARPA Internet SMTP mail: msokolov@blackwidow.cwru.edu