Hi Edwin, If someone can get the code to compile with Microsnot visual c++, I can do the rest. I already have developed a greatly cut down version of FRACTINT called MANPWIN which has true colour support and true windows 32 bit operation. The problem with FRACTINT code is the massive number of globals which makes porting so difficult. Most of them are hard to follow. If only fractint was made of stand alone subroutines, it would be so much easier to port. I understand why it was set up that way due to the speed requirements on 8086 machines, but today, it is not an issue. Fractint is just too big. My 2 cents worth, Paul. ---------------------------------------------------------- Paul de Leeuw Computers NSW Central Coast, Australia Email: pdeleeuw@deleeuw.com.au www: < http://www.deleeuw.com.au> ABN 72 360 822 562 ---------------------------------------------------------- -----Original Message----- From: fractdev-bounces+pdeleeuw=deleeuw.com.au@mailman.xmission.com [mailto:fractdev-bounces+pdeleeuw=deleeuw.com.au@mailman.xmission.com]On Behalf Of Edwin Sent: Thursday, 22 July 2004 12:29 PM To: fractdev@mailman.xmission.com Subject: [Fractdev] Fractint vs XP, redux I've read with some interest the discussions on the difficulty of getting the DOS version of Fractint running under Windows XP. While there are obviously solutions around using DOS, FreeDOS, bootable CDs, etc, I can't help thinking that the hassle involved in rebooting is too great for this to be viable. Long-term, some sort of rewrite to make Fractint more environment-neutral so it can have DOS, Windows, X11, etc, flavors built from a largely common code base, would seem to be ideal. But, as has also been pointed out numerous times, this would be a huge undertaking and no-one seems to have the time. I wonder if the more limited goal of getting Fractint and Windows to 'play nice' is more achievable? That is, keep the existing interface and make the minimal changes possible so Fractint will run under XP. It sounds like the major problem is graphics card support - are there any other major gotchas people know about? If that's the main problem, the simplest approach I can think of is: 1) Get Fractint compiling with a modern C compiler toolchain (eg Visual C++) to produce a Win32 executable. 2) Modify the display-driver code to use DirectX, so that Windows supplies the video-card handling. Otherwise leave as a full-screen, character-mode app. (This would of course be IFDEF'ed to leave the existing code intact). That still sounds like a lot of effort, but perhaps a bit more achievable than the 'major rewrite' approach. Does this sound interesting? Plausible? Can anyone think of anything simpler? Would it be simplest to start with the DOS code, or XFractint (or one of the other branches I've heard mentioned)? Regards, -- Edwin _______________________________________________ Fractdev mailing list Fractdev@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/fractdev