Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site bu-cs.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!allegra!mit-eddie!genrad!panda!talcott!harvard!bu-cs!root
From: root@bu-cs.UUCP (Barry Shein)
Newsgroups: net.unix
Subject: Re: Not checking printf's result causes another news bug
Message-ID: <606@bu-cs.UUCP>
Date: Sat, 24-Aug-85 02:44:25 EDT
Article-I.D.: bu-cs.606
Posted: Sat Aug 24 02:44:25 1985
Date-Received: Sun, 25-Aug-85 13:30:55 EDT
References: <11@brl-tgr.ARPA> <1288@eagle.UUCP> <15908@watmath.UUCP>
Organization: Boston Univ Comp. Sci.
Lines: 18


>Re: use SIGUSER, use SSIGNAL (actually don't), use vprintf (tho you
>probably don't have it.) [followup by Guy Harris]

Ok, maybe it's just late and I'm grumpy, but aren't we being a little
distracting here? The point was a much broader one about introducing
an interrupt mechanism for alerting a program to I/O errors, something
distinctly not in UNIX, likely by choice as the designers were well
aware of the IBM SYNAD handling. I wonder if FSIGEOF and FSIGERR should
be added to fcntl.h alongside FASYNC to handle these situations, especially
where a large piece of software (as noted in the article, most programs
that use printf) is suddenly discovered to be deficient in write error
handling.

I dunno, so much of this stuff seems to reduce to 'you forgot a semi-colon'
and the point gets lost, even though I am sure what you said was true.

	-Barry Shein, Boston University