I'm right now building X on my MicroVAX II.  According to NetBSD.cf,
NetBSD/vax is *not* supported. Nevertheless I didn't expect or face any
major problems.  I had to make changes to Imake.cf and NetBSD.cf so that
NetBSD/vax is recognized and things configure properly.  This allowed me
to build all of the client programs in xc.  They seem to work fine so
far (I haven't done too much testing).  Even the performance is pretty
good.  The only problem is the executable size.  Without shared
libraries it is totally ridiculous.  xclock is 926225 bytes (759808
stripped).  A 4.3BSD/VAX X11R4 xclock binary is only 249856 bytes.  Has
X grown (bloated) that much?  (An XFree86 3.3 xclock binary on my x86
Linux box is 18324 bytes, showing the immense benefit of shared X
libraries)

Xprt failed to build because Xserver/include/servermd.h failed to define
some macros.  This happened because gcc with -ansi does not define vax. 
It seems to me that this should be added:
#if defined(__vax__) && !defined(vax)
#define vax
#endif

Some other vax checks are somewhat questionable.  The X sources seem to
assume that if it is a VAX and it is not running Ultrix then it must be
running plain old BSD.  In particular Xosdefs.h then sets X_NOT_POSIX
and X_NOT_STDC_ENV.  I don't know about the first one, but the second
one seems very wrong (maybe it is fortunate vax is not defined).  I
don't know about Xserver/PEX5/dipex/swap/floatconv.h, but that doesn't
really matter at this point.

BTW.  The VAX port is really stable now.  It only crashes when you do
certain well-defined things, like set the clock or fsck (or is it fsck
-p only?)  Otherwise it can take heavy use without any problems.  Great
work!


-- 
|  Boris Gjenero <bgjenero@undergrad.math.uwaterloo.ca>              |
|  Home page:  http://www.undergrad.math.uwaterloo.ca/~bgjenero/     |
|  "Luke, you're going to find that many of the truths we cling to   |
|  depend greatly on our own point of view." - Obi-Wan Kenobi, ROTJ  |