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=0.2 required=5.0 tests=BAYES_00,INVALID_MSGID,
	REPLYTO_WITHOUT_TO_CC autolearn=no autolearn_force=no version=3.4.4
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,dc94fe39f71093ec
X-Google-Attributes: gid103376,public
From: kilgallen@eisner.decus.org (Larry Kilgallen)
Subject: Re: The revolution will not be standardized
Date: 1999/12/13
Message-ID: <1999Dec13.102812.1@eisner>#1/1
X-Deja-AN: 560111545
References: <82p7hu$l1q$1@nnrp1.deja.com> <82ppc9$1u6$1@nnrp1.deja.com>
 <385252E8.FF140CD2@acenet.com.au> <82upp5$dt8$1@nnrp1.deja.com>
 <3854EBD4.6B100D52@acenet.com.au>
X-Trace: news.decus.org 945098896 20203 KILGALLEN [216.44.122.34]
Organization: LJK Software
Reply-To: Kilgallen@eisner.decus.org.nospam
Newsgroups: comp.lang.ada
Date: 1999-12-13T00:00:00+00:00
List-Id: <comp.lang.ada>

In article <3854EBD4.6B100D52@acenet.com.au>, Geoff Bull <gbull@acenet.com.au> writes:
> 
> 
> Robert Dewar wrote:
> 
>> This is at least partially in place, GNAT supports a form of
>> with type, precisely for this purpose.
> 
> I had seen this in the JGNAT bindings for Rapid, but there is
> no mention of this in the documentation, not even obscurely
> placed under the heading "With Type Clause" :-)
> 
> I had thus assumed it was JGNAT only. In retrospect this was a
> silly assumption, why would you make the front end different?
> 
> Through the wonders of open source, I find:
> 
>       ----------------------
>       -- With_Type clause --
>       ----------------------
> 
>       --  This is a GNAT extension, used to implement mutually recursive
>       --  types declared in different packages.
> 
>       --  WITH_TYPE_CLAUSE ::=
>       --    with type type_NAME is access | with type type_NAME is tagged

If your comment about the "wonders of open source" means finding
features that are not documented, I "wonder" about the wisdom of it.

Often a feature that is not documented is also not supported (although
in this particular case, Robert indicates it is supported).  Typically
the documentation, however, is the indication of what is supported and
what is not.  I recall discussions of some GNAT features provided only
for building GNAT.

Open source would be useful if you were going to support yourself on
a feature, of course.

Larry Kilgallen