Ian Collins wrote:
> Roland Mainz wrote:
> >Ian Collins wrote:
> >>Roland Mainz wrote:
> >>>Garrett D'Amore wrote:
> >>>>Roland Mainz wrote:
[snip]
> >That will not work because the |#ifdef <symbol> /*64bitinclude*/ #else
> >/*32bitinclude*/ #endif| comes before all other cpp statements (or
> >better: The "#ifdef <symbol>" part comes before all other cpp
> >statements), including "#include"s which are done in the single AST
> >headers.
>
> I see. How horrible.
Why ? The header generation code just did not forsee platforms which
consists of both 32bit and 64bit binaries in one operating system (the
fix would be to make sure that the public headers in
arch/$(PLATFORM)/include/ast/ are identical on 32bit and 64bit).
If you look at the differences you'll see that they are small and it's
technicially possible to fix it in the future. It's just not possible to
do it ad-hoc because there is a large software library which depends on
libast and ksh93 is just a tiny piece of it.
> I've had to face code like that before and it's a bit of a no win
> situation trying to modernise it.
What do you mean here ? The AST code isn't "old" ("ksh93" means "korn
shell, spec=93" ... current version is ast-ksh.2007-01-11).
----
Bye,
Roland
--
__ . . __
(o.\ \/ /.o) roland.mainz at nrubsig.org
\__\/\/__/ MPEG specialist, C&&JAVA&&Sun&&Unix programmer
/O /==\ O\ TEL +49 641 7950090
(;O/ \/ \O;)
