[comp.windows.x] Difficulties getting X.V11R3 going on a Vaxstation 2000

pr@computer-lab.cambridge.ac.UK (Peter Robinson) (02/11/89)

Summary: Is the revised sg for Ultrix 2.2 only available as binary?
References: <8902041633.AA04521@EXPIRE.LCS.MIT.EDU>

As advised on xpert, I picked up the new sg.o from the xstuff server.  
However, the person trying to build a new Unix observed that all three 
of
	- the standard distributed binary
	- the result of compiling the distributed source
	- the binary from xstuff
are non-trivially different.  He also expressed some reservations about 
splicing an arbitrary and unrecognisable bit pattern into the middle of a 
Unix kernel (:-), but tried it anyway and it does not seem to work.  It 
gets as far as printing the Worksystem banner and then stops.  The disc 
still chatters occassionally, but the system does not respond to the 
console or rlogins.

It may be possible to reverse-engineer the changes, but obviously it would
help if we could get hold of source or source diffs.  I checked the index 
for etc on xstuff and could not see anything obvious;  might you be able to 
release the diffs?

- Peter Robinson.

converse@EXPO.LCS.MIT.EDU (Donna Converse) (02/11/89)

> Summary: Is the revised sg for Ultrix 2.2 only available as binary?

Yes.

> .. it does not seem to work

I can only think of stupid mistakes: the file from the xstuff
server should be uudecoded on the target machine, 29946 bytes.
	?

maj@computer-lab.cambridge.ac.UK (Martyn Johnson) (02/13/89)

> > Summary: Is the revised sg for Ultrix 2.2 only available as binary?
> Yes.
> > .. it does not seem to work

Local changes to the header buf.h affect the compiled code sg.o; this
could be the reason that the binary from "xstuff" does not work. Trying
to build kernels out of binaries compiled against different headers is
a waste of time.