Mostly good news on the NeXT front.

[Ben Lindstrom]

I'll be able to test on next 4.2 after the initial 3.3 port.  

I have a vsnprintf/snprintf that compiles (works?) on NeXT.  Not sure how
portable nor if it truely solves the security issue (it's from the
comp.sys.next.programmer fokes).  Plus NeXT 3.3 (not sure about 4.2 yet)
lacks a putenv().  I found one that compiles cleanly from the GNU libc
fokes.  Which brings up licensing issues.  I'm not sure if we can get a
wavier or if we will have to write our own.  It looks simple enough to
rewrite if we have to.  I personally hate reinventing wheels.

But things are slowly compiling. It does look like I am missing poll()
(I believe older versions of Linux has this problem so I may be able to
find sample code) and the whole mem*() functions are very poorly
documented so it should be interesting.  (It exists in the libposix.a, but
no manpages)

The "ssh" program compiled, but on attempting to connect I get a coredump
and a nice "Floating exception" error.  (It would not suprise me if my
vsnprintf/snprintf are incorrect, but I've not bothered to trace it yet)

But hey.. We have least progress.. And after finding -posix and
-D_POSIX_SOURCE cc flags I have less #ifdef and things are compiling
much cleaner.  So maybe next week sometime I may have patches.  (Looks
like 2 or 3 #ifdef now instead of 10 =)