From: Ludovic Brenta <ludovic@ludovic-brenta.org>
Subject: Re: tagged primitive operation and freezing
Date: Tue, 16 Nov 2010 00:53:03 -0800 (PST)
Date: 2010-11-16T00:53:03-08:00 [thread overview]
Message-ID: <bc21a278-46a4-44f3-b0d1-e20d115e8a5a@a37g2000yqi.googlegroups.com> (raw)
In-Reply-To: 6d5b4fec-ac32-4efe-a9b8-02cbaca625b9@u25g2000pra.googlegroups.com
troll wrote on comp.lang.ada:
>> But only if it is tagged. untagged type, even though frozen, DOES
>> compile.
My initial reaction to this was: of course, because, if type A is
untagged,
type Flight_T is new A;
freezes Flight_T immediately since its representation is known,
therefore
procedure C is new Blah (Object_T => Flight_T);
is not primitive because not dynamically dispatching.
After re-reading 13.14, I see it does not discuss non-tagged types or
types derived from non-tagged types, so I'm not so sure about the
first part anymore. And, per 3.2.3(6), C seems to be primitive after
all (primitive subprograms are not necessarily dynamically
dispatching).
While I intuitively agree with Adam's explanation, I can't find the
chapter and verse in the ARM that backs it; it seems like a gray area
to me.
--
Ludovic Brenta.
prev parent reply other threads:[~2010-11-16 8:53 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-11-13 3:24 tagged primitive operation and freezing troll
2010-11-13 10:24 ` Ludovic Brenta
2010-11-15 14:23 ` troll
2010-11-15 16:30 ` Adam Beneschan
2010-11-16 8:53 ` Ludovic Brenta [this message]
replies disabled
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox