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=-1.3 required=5.0 tests=BAYES_00,INVALID_MSGID
	autolearn=no autolearn_force=no version=3.4.4
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,9e2776c05028676e
X-Google-Attributes: gid103376,public
From: dewar@merv.cs.nyu.edu (Robert Dewar)
Subject: Re: Why Ada is not the Commercial Lang of Choice
Date: 1997/06/21
Message-ID: <dewar.866898018@merv>#1/1
X-Deja-AN: 251571694
References: <33A7FBFF.29D2@mitre.org> <EC471t.Ltp@thomsoft.com>
Organization: New York University
Newsgroups: comp.lang.ada
Date: 1997-06-21T00:00:00+00:00
List-Id: <comp.lang.ada>


Keith said

<<GNAT follows this advice (the full type is modular, but that doesn't
make literals available).  ObjectAda makes System.Address an access
type.
>>

That follows an old custom in Alsys compilers, where Address has always
been an access type. What I do not like about this choice is that it means
that the compiler always wants to initialize address values to the null
pointer. This default initialization seems appropriate for access types,
but I usually find it to be unwelcome junk overhead for addresses.

The modular representation seems more natural for modern machines.
Historically one of the reasons for the Alsys choice was that the modular
choice is NOT a good choice for the old junk segmented 286 architecture,
and indeed if we still had to fight that machine, I would probably follow
the access type decision, since it is indeed more portable in this sense.

(yes, yes, I know, the 32-bit x86 architecture still has segmentation, for
gruesome details, see my book on Microprocessors - when Hennessey reviewed
this section of the book, he wrote a large note asking "does anyone use
this nonsense?" -- the answer to his question is pretty much no :-)