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: 1025b4,1d8ab55e71d08f3d
X-Google-Attributes: gid1025b4,public
X-Google-Thread: 103376,1efdd369be089610
X-Google-Attributes: gid103376,public
From: dewar@merv.cs.nyu.edu (Robert Dewar)
Subject: Re: what DOES the GPL really say?
Date: 1997/06/29
Message-ID: <dewar.867557676@merv>#1/1
X-Deja-AN: 253355237
References: <m2oh8ydzt8.fsf@devo.ridgenet.net> <dewar.867079112@merv>
 	<33B014E3.3343@no.such.com> <5oqp9s$7vj$1@news.nyu.edu>
 <m2soy3kebq.fsf@devo.ridgenet.net> <dewar.867497844@merv>
 <slrn5ra7oi.odk.Olivier.Galibert@iria.mines.u-nancy.fr>
 <dewar.867554739@merv>
Organization: New York University
Newsgroups: comp.lang.ada,gnu.misc.discuss
Date: 1997-06-29T00:00:00+00:00
List-Id: <comp.lang.ada>


I said

<<Also, note that actually the release that Ronald Cole wanted was 3.10a,           
which is offered under scheme (a), i.e. the sources are distributed                 
with the binaries, not under scheme (b) which we only use for                       
wavefronts. Wavefronts are distributed on a very limited basis by us,               
we do not even make them generally available to customers. We only                  
make them available selectively when it is absolutely necessary to fix              
a particular customer problem.                                                      
>>                                                                                  
                                                                                    
                                                                                    
Actually 3.10a is a very good example of why we hold off on making public           
releases. One of the new features in 3.10a is a more accurate treatment             
of delay statements and other details of the tasking implemen                       
tation. This works fine, but has a side effect that the tasking runtime
is always loaded for almost all non-tasking programs. 

Now this does not affect our customers in general, since nearly all of
them are using tasking in any case, or at any rate are building very large
programs where this is not a major factor.

However, we we know from many posts in this group, keeping student type
program executables small is pretty critical for a lot of the users of
the public version.

Consequently, we consider fixing this a prerequisite for the public release,
and indeed when 3.10 is released publicly, it will fix this problem. The fix
is not trivial, but we see it as important enough of an issue to hold the
public release for.

Robert Dewar
Ada Core Technologies

And now back to work, it so happens that I am busy working on precisely
this problem right now, as well as getting better debugging support in
place using gdb (we are very close to having gdb in Ada mode have 100%
knowledge about Ada data structures, including packed arrays, variant
records, variable length components depending on discriminants etc).