Xref: utzoo comp.mail.misc:4102 comp.mail.uucp:5314 comp.sys.ibm.pc.misc:2317
Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!clarkson!news
From: help@kendra.kew.com (Drew Derbyshire - UUPC/extended Help Desk)
Newsgroups: comp.mail.misc,comp.mail.uucp,comp.sys.ibm.pc.misc
Subject: Re: UUPC 1.08a delay bug (MSC)
Message-ID: <1990Oct4.234839.23948@news.clarkson.edu>
Date: 4 Oct 90 23:48:39 GMT
References: <1990Oct4.084647.29558@hq.demos.su>
Sender: ahd@clutx.clarkson.edu (Drew Derbyshire)
Reply-To: help@kendra.kew.com (Drew Derbyshire - UUPC/extended Help Desk)
Organization: Clarkson University, Potsdam, NY
Lines: 26
Nntp-Posting-Host: clutx.clarkson.edu

From article <1990Oct4.084647.29558@hq.demos.su>, by ache@hq.demos.su (Andrew A. Chernov):
> "delay" is standart library function for Turbo-C, but not for MSC.
> Some "delay" emulation code for MSC in file "dcplib.c" contains very
> unplesant bug:
> 
>        ...
> 	 milliseconds =- (t.millitm - last); /* Yes --> mSecond delta*/
>        ...
>          milliseconds =- 1000 * (int) (t.time - seconds)
>        ...
> 
> But in MSC (like in all ANSI's standard compilers) "=-" means "= -",
> not "-=". I spend more then 6 hours to localize this bug.

and the bloody thing loops forever.  I fixed that ... I thought.  I know the
system hung on me when testing it.  Only took me an hour, but *I* knew
what I had changed.

The next release will use DosSleep instead.  I hope to get the OS/2
version done soon, which requires use of the MS C compiler; hence the
next release will be tested better  ... I hope.

Drew Derbyshire

Internet:  help@kendra.kew.com           Snail mail:  108 Decatur St, Apt 9
Voice:     617-641-3739                               Arlington, MA 02174