From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me
X-Spam-Level: 
X-Spam-Status: No, score=-0.8 required=5.0 tests=BAYES_00,INVALID_DATE
	autolearn=no autolearn_force=no version=3.4.4
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,ffce418d7a49585f
X-Google-Attributes: gid103376,public
X-Google-ArrivalTime: 1994-10-04 09:12:34 PST
Path: 
 bga.com!news.sprintlink.net!howland.reston.ans.net!newsserver.jvnc.net!netnews.summit.novell.com!netnews.summit.novell.com!not-for-mail
From: jls@summit.novell.com (-mlc-+Schilling J.)
Newsgroups: comp.lang.ada
Subject: Re: C++ bashing (was Re: Vendor bashing? Sort of.)
Date: 4 Oct 1994 12:03:23 -0400
Organization: Novell, Summit
Message-ID: <36rucbINN4gr@marble.summit.novell.com>
References: <35nlhb$6u0@gnat.cs.nyu.edu> <85BA331C6A0@annwfn.com>
NNTP-Posting-Host: marble.summit.novell.com
Date: 1994-10-04T12:03:23-04:00
List-Id: <comp.lang.ada>

>>In <35nlhb$6u0@gnat.cs.nyu.edu> dewar@cs.nyu.edu Robert Dewar writes:
>>
>>I would not quite put C-Front in the same category as GNAT. The big difference
>>is whether you are generating source code as an intermediate step. The
>>word preprocessor is normally reserved for translator systems that do
>>generate intermediate source. 

If cfront had only done syntax and semantic checking for C++-only constructs,
and passed C constructs on to the C compiler for checking, then it could
rightfully be called a preprocessor.  But in fact cfront does full checking
for the entire C++ language, and no diagnostics should ever appear from 
the underlying C compiler (except due to glitches in either cfront or the
C compiler).  Thus cfront must be considered a legitimate compiler, as would
any similar translator that compiled Ada through C.  (Of course GNAT is also
a legitimate compiler, I wasn't trying to imply it is a preprocessor!)

See also Stroustrup D&E, Sec. 3.3.1, for a similar defense of what cfront is.

Yes, there are certainly limitations to this approach, especially when it
comes to areas like exception handling and symbolic debugging support. 
But the limitations falls under the category of "tradeoffs in compiler 
architecture", not "this is not a compiler".

-- 
Jonathan Schilling
Novell, UNIX Systems Group (UNIX System Laboratories)
jls@summit.novell.com