home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!mcsun!corton!seti!goudurix.inria.fr!edelson
- From: daniel.edelson@inria.fr (Daniel R. Edelson)
- Newsgroups: comp.lang.c++
- Subject: Re: template static data members
- Message-ID: <4089@seti.UUCP>
- Date: 20 Aug 92 08:30:45 GMT
- References: <Bt5FDH.DFK@apollo.hp.com> <4086@seti.UUCP> <Bt8K2o.Azy@apollo.hp.com>
- Sender: news@seti.UUCP
- Reply-To: Daniel R. Edelson <daniel.edelson@inria.fr>
- Organization: INRIA -- Institut National de Recherche en Informatique et Automatique -- Rocquencourt, France
- Lines: 22
-
- In article <Bt8K2o.Azy@apollo.hp.com>, vinoski@apollo.hp.com (Stephen Vinoski) writes:
- |> In article <4086@seti.UUCP> Daniel R. Edelson <edelson@sor.inria.fr> writes:
-
- |> Regardless of my error, the real problem underlying all of this is
- |> that the compiler cannot tell a legal program containing an
- |> uninitialized definition for each static data member from an illegal
- |> program containing only declarations, no definitions, for each static
- |> data member, since in both cases a bss symbol is created. It never
- |> used to matter until templates and automatic instantiation came along.
-
- |> Thanks for making me go back and try this again, Dan, and also thanks
- |> for embarrassing me in public. :-) You should already know that I am
- |> perfectly capable of doing that without anyone's help. :-)
-
- You'll just have to return the favor next time I open my mouth too
- wide. :-O
-
- Ciao,
-
- |> -steve
-
- d-
-