S on 4.2BSD ?

[mikem at uwstat.UUCP]

I've seen several questions about S recently so I'll follow-up rather
than reply:....
> 	Has anyone installed S on 4.2BSD ? We have S basically working on
> 4.2BSD, but a number of crucial primitives are still broken. Not only did
> we have to hack S's signalling code, but a few other rough edges have made
> us wonder whether something more fundamental is wrong. Is 4.2BSD's F77
> getting in the way (by being its buggy self) ? Have there been new, patched
> up releases of S in the past year (or two) ?

Yes to both of your questions.  THere was a new release of S in about
January of this year.  As I recall making the old S work under 4.2
looked like it was going to be a major job so we didn't try.  We waited
for the new S before we changed to 4.2 .  Note there are some minor
things I had to do to install the new S, mail me and I'll send details.

The standard 4.2 f77 is totally %$#^#^%.  I have yet to see a any of the
fixed version which is satisfactory either.  We run the 4.1 f77 compiler
with the 4.2 libraries and everything seems to be happy.  Specifically
we have S, matlab, linpack, eispack, imsl ,... compiled and apparently
correct.

Lets continue this discussion in net.math.stat.
-- 

Mike Meyer --  Phone (608) 262-1157

EASY ARPA:	mikem at statistics
CORRECT ARPA:	mikem at wisc-stat.arpa
UUCP	...!{allegra,ihnp4,seismo,ucbvax,
	     pyr_chi,heurikon,uwm-evax}!uwvax!uwstat!mikem