Date: Wed, 10 Mar 1999 00:23:01 +0100
From: Bertho Stultiens <bertho@panter.soci.aau.dk>
To: wine-devel@winehq.com
Subject: Re: Corel-Cygnus deal - More info

Gavriel State wrote:
[snip]
> Now for the bad news: The work that's being done is in Cygnus' GNUPro
> toolchain, not directly in egcs.  The changes will be available as
> patches to GNUPro, but may need some work to be integrated into egcs.
> Once someone does do the work to integrate them into egcs, it will be up
> to the egcs maintainers to decide if they want to incorporate these
> changes into egcs itself.

And that is a major issue for wine. Wine cannot use any of the suggested
changes because it needs to be compiling with the current tools and not
something that will come in the future. This means that there cannot be
compromises to use Windows-compiler specific constructs in wine itself.
IMHO, it would be a bad move to code as badly as windows has shown us.
Even the suggestion to recode the COM objects using C++ is not an option
unless you enforce the use of a future compiler extension. Not a nice
idea for portability. Wine can exist, run and be compatible with smart
constructs and we should not compromise here if possible.

-- flamewarning on --
I do see the advantages for corporate winelib developers. They also have
the infrastructure and resources to take on huge porting efforts (as
Corel has), but, speaking for myself here, I am not working on
wine/winelib for big companies. I am doing it for my own fun and all who
want to use wine/winelib. They must not be cut off because there will be
coming a new compiler when stuff gets ported.
-- flamewarning off --

Greetings Bertho