Alistair G. Crooks writes:
 > >    Has anyone managed to get a fully functioning set of Tcl7.4 and 
 > > Tk4.0 running on NetBSD1.1 ??
 > >  
 > > I am using an Amiga 3000 running NetBSD1.1 with gcc2.7.2 and can 
 > > get both packages to compile but I get errors running the test 
 > > programs.
 > >  
 > > The Tk4.0 test programs give a large number of errors, none of 
 > > which seem to be fatal. The Tcl7.4 test programs come up with the
 > > following errors which cause problems later when trying to run the
 > > TclX7.4a extensions.
 > 
 > I put both of these down to NetBSD's lack of a unified buffer/VM cache.

*BEEP*, Wrong answer, thanks for playing.  While NetBSD does not have
a unified buffer/VM cache, it's not responsible for this problem.  In
fact, since he's running on NetBSD-Amiga, it's not even clear that the
unified buffer/VM cache from FreeBSD would even be applicable for his
machine.

The real cause of the Tcl failures here is that NetBSD is one of the
few Unixen with a max file size of 64 bits.  The code that converts
the file size from the stat call into a ascii string assumes that the
value is a 32 bit entity.  When patched, it provides the correct
answers.

On the other hand, the TK errors are probably due to differences in X
servers.  I don't know Tcl/TK yet, but when I built them a month ago,
the test instructions indicated that it was probably OK to get errors
unless it dumped core.  If you're still having problems, I'd suggest
that you post to the tcl newsgroup.