Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!cbatt!cbosgd!ucbvax!hplabs!pyramid!csg
From: csg@pyramid.UUCP (Carl S. Gutekunst)
Newsgroups: net.unix-wizards
Subject: Re:  Rumor about AT&T Validation
Message-ID: <566@pyramid.UUCP>
Date: Sat, 9-Aug-86 01:42:14 EDT
Article-I.D.: pyramid.566
Posted: Sat Aug  9 01:42:14 1986
Date-Received: Sun, 10-Aug-86 19:47:49 EDT
References: <2846@brl-smoke.ARPA>
Reply-To: csg@pyramid.UUCP (Carl S. Gutekunst)
Organization: Pyramid Technology Corp., Mountain View, CA
Lines: 17

In article <2846@brl-smoke.ARPA> ron@BRL.ARPA (Ron Natalie) writes:
>Well, one point is that fstat doesn't work right on systems that use
>sockets to support the PIPE system call.

Um, not necessarilly. Pyramid OSx implements SV pipes using sockets, and the
SV fstat() works perfectly. (Believe me, I know; the 4.2bsd fstat() doesn't
distingush between pipes and sockets, so I had a BSD program calling the SV
fstat() so it could determine which was which....)

I really don't believe the rumor that AT&T will reject machines that have
sockets *anywhere*, but I *DO* believe the validation suite will check for
correctness on fstat().

Waving a red flag in front of the bull again, :-) I'd bet that Sun's Unix 3.2
does the fstat() correctly for SV, too, and they also have pipes on sockets.

<csg>