Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!wuarchive!gem.mps.ohio-state.edu!ginosko!uunet!pilchuck!dataio!bright
From: bright@Data-IO.COM (Walter Bright)
Newsgroups: comp.lang.c++
Subject: Re: Wanted: C++ Reference Manual
Message-ID: <2143@dataio.Data-IO.COM>
Date: 25 Sep 89 18:52:08 GMT
References: <21.UUL1.3#913@acw.UUCP> <110@dumbcat.UUCP>
Reply-To: bright@dataio.Data-IO.COM (Walter Bright)
Organization: Data I/O Corporation; Redmond, WA
Lines: 23

In article <110@dumbcat.UUCP> marc@dumbcat.UUCP (Marco S Hyman) writes:
 guthery@acw.UUCP (Scott Guthery) writes:
<    The only rigorous and reliable definition of the semantics of the
<    programming language C++ that I have been able to find is "Whatever
<    the latest release of Cfront does."

The problem with that is the inability to distinguish between a bug and
a feature. Lots of people may end up relying on an unintended quirk of
cfront, so future versions become forced to support it, and thus we all
get stuck with warts.

Cfront is no longer the only C++ implementation, others currently exist
and a large number of others are under development. Relying on undocumented
behavior of cfront will get you into as much trouble as doing things like
	*p++ = *p++;