From: dewar@merv.cs.nyu.edu (Robert Dewar)
Subject: Re: Function Pointers
Date: 1998/04/15
Date: 1998-04-15T00:00:00+00:00 [thread overview]
Message-ID: <dewar.892641366@merv> (raw)
In-Reply-To: ErEsxx.9F@jvdsys.nextjk.stuyts.nl
Jerry said
<<-- As GNAT allows only one compilation unit in a file
-- we nest the functions within main;
>>
Surely a bad reason for doing anything. It is bad practice (see for example
AQ&S) generally to put more than one compilation unit in a file anyway.
But the issue of whether to write a single compilation unit or many
compilation units is surely a design issue that should not be affected
by low level representation issues (i.e. how compilation units are
represented in files, whether you want to use gnatchop if you are using
GNAT etc.)
next prev parent reply other threads:[~1998-04-15 0:00 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
1998-04-13 0:00 Function Pointers Barry L. Dorough
1998-04-13 0:00 ` Matthew Heaney
1998-04-14 0:00 ` Jerry van Dijk
1998-04-15 0:00 ` Robert Dewar [this message]
1998-04-16 0:00 ` Jerry van Dijk
replies disabled
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox