Path: utzoo!utgpu!jarvis.csri.toronto.edu!torsqnt!tmsoft!masnet!canremote!jharkins@sagpd1.uucp
From: jharkins@sagpd1.uucp@canremote.uucp (jharkins@sagpd1.UUCP)
Newsgroups: comp.lang.c
Subject: Re: typecasting constants
Message-ID: <89122504120812@masnet.uucp>
Date: 21 Dec 89 05:04:00 GMT
Organization: Canada Remote Systems Limited, Mississauga, ON, Canada
Lines: 22

From: jharkins@sagpd1.UUCP (Jim Harkins)
Orga: Scientific Atlanta-GPD, San Diego

Our customer is trying to force a coding standard rule down our
throats that we don't like, but can't give an argument on why we hate
it.  The rule is "All constants will have their types explicitly
specified." This means we'll have

#define FRED ((long) 4)
#define WILMA  ((short) 9)

Outside of the fact that nobody else ever does this, and we've never
had a problem omitting the typecasts, whats a good argument against
this? Note we don't have a problem with things like '0x42L' when we
know it has to be a long.

jim
"Only dead fish go with the flow"

---
 * Via MaSNet/HST96/HST144/V32 - UN C Language
 * Via Usenet Newsgroup comp.lang.c