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: 103376,80e8e0df8032d89e
X-Google-Attributes: gid103376,public
X-Google-Thread: 109fba,fa07350fd81f7563
X-Google-Attributes: gid109fba,public
X-Google-Thread: 1108a1,fa07350fd81f7563
X-Google-Attributes: gid1108a1,public
X-Google-ArrivalTime: 1994-11-01 14:32:38 PST
Newsgroups: comp.lang.ada,comp.lang.c++,comp.object
Path: 
 bga.com!news.sprintlink.net!howland.reston.ans.net!math.ohio-state.edu!cis.ohio-state.edu!news.sei.cmu.edu!ajpo.sei.cmu.edu!riehler
From: riehler@ajpo.sei.cmu.edu (Richard Riehle)
Subject: Re: Is C/C++ the future?
Message-ID: <1994Nov1.131914.14904@sei.cmu.edu>
Sender: netnews@sei.cmu.edu (Netnews)
Organization: AdaWorks Software Engineering, Palo Alto, CA
References: <1994Oct25.234705.26530@sei.cmu.edu> <85DF1879046@annwfn.com>
Date: Tue, 1 Nov 1994 13:19:14 EST
Xref: bga.com comp.lang.ada:7437 comp.lang.c++:34991 comp.object:8118
Date: 1994-11-01T13:19:14-05:00
List-Id: <comp.lang.ada>

In article <85DF1879046@annwfn.com> merlin@annwfn.com (Fred McCall) writes:

>Doesn't this problem (not being able to write 'real' programs, according
>to the respondent, without requiring all that machine-specific support)
>pretty much shoot the much-vaunted 'portability' of Ada code in the
>foot?  It seems to me that this leaves Ada with the same problem that so
>many Ada advocates want to 'bash' C/C++ for -- non-portable code is
>non-portable (somewhat solipsist, but that seems to be the complaint).

  This is a problem one would expect from high-level languages that try
  to be portable.  COBOL, an early attempt at portable HOL was never
  portable because each compiler vendor realized the need to include,
  embedded in the targeted-edition of the compiler, a set of platform-
  specific options.  

  An intelligent designer of an Ada application can create platform-independent  wrappers (as packages) to effect the design of fully portalbe designs. But
  implementations will frequently be non-portable because of the variations
  in operating environments.  An Ada designer has good tools available to
  push platform dependencies to the lowest possible level of abstraction,
  thereby enhancing the portability of the product.  

  Even though Ada 9X provides some relief in this area, the continued progress
  in new hardware designs will necessitate retaining the capability for 
  interfacing high-level code  to low-level code in some applications.   

>
>Should the language include platform independent ways of doing most of
>those 'platform-specific' things?  It would seem to follow the Ada
>philosophy of maximal safety; after all, if pieces of configuration
>control are required to be part of the linker, why not define an 'Ada
>windows' interface (for example) and then require compilers for
>platforms with window support to map their functionality onto the
>interface? 

   In principle, I agree. Much of this lies at the feet of the compiler 
   vendors, who have chosen not to go beyond simple validation with some
   of their products.  And validation requires nothing more than the bare
   essentials of the language.  I provides no guarantee that the compiler
   will be useful for real work.  

   All in all, a language such as Ada is the lesser of two evils when
   compared against assemblers (native or C).  We can, at least, improve
   portablility, even if we do have to resort to some non-portable code
   from time to time.   

   Richard Riehle