Gould Computers - UTX/4.2

[Doug Gwyn <gwyn>]

> 	1) How stable is a Gould computer? (e.g. Uptime, maintance, etc)

We have several PN6080, PN9050, and PN9080 computers acquired during the
last year or so.  The hardware seems no less reliable overall than DEC's,
although fewer third-party add-on peripherals are available.  (A Selbus-
to-Unibus converter that is rumored to exist may help with this.)  Since
we have been beta-testing UTX/32, it would not be fair for me to comment
on the reliability of the software (I mostly use my own anyway).

> 	2) How many problems have your had with the Unix 4.2bsd/System V
> 	operation system, UTX/4.2, as compared with just a  straight
> 	4.2bsd or System V system? (How portable is code from SYS V or
> 		4.2bsd systems?)

Some of us feel that neither the 4.2BSD nor the System V camp will be
fully satisfied with the hybrid.  The System V camp can at least use the
BRL UNIX System V emulation, an older version of which is shipped on the
"D4" tape from Gould as user-contributed software.  UTX/32 is definitely
based on 4.2BSD and can probably be made more compatible with 4.nBSD
than it presently is.  It ought to be one or the other..

> 	3) How do you rate their machines to other companys? (DEC, IBM, AT&T, etc)

We got ours because they were cost-effective at the time (not factoring
in software conversion costs, alas).  Technically, the weakest points
of the systems are the segmented base-register architecture, large page
size (get all the main memory you can!), poor compiler quality (which
appears to be improving with new releases), and tendency for changes
("improvements") in the functioning of system utilities.  The strongest
point is a lot of MIPs for the buck.

Of course, the above is my personal opinion and is not to be taken as
official word of any government agency.