Re: [Fractdev] Update to ManpWIN and Problems WINFract
Paul, You will find bignum_seg defined in biginit.c. The assembly of bignum.asm shouldn't be a problem, but it does need to be linked to C code that initializes everything. The memory model of the assembly code is not correct for what you are trying to do. I'm not sure what it needs to be in order to work in a Windows environment. I'm still trying to track down the problem with passes=orbits. It doesn't work for Fractint or WinFract, unless running a PAR. It works fine in Xfractint. Jonathan --- pdeleeuw@deleeuw.com.au wrote: From: "Paul" <pdeleeuw@deleeuw.com.au> To: "'Fractint developer's list'" <fractdev@mailman.xmission.com> Subject: Re: [Fractdev] Update to ManpWIN and Problems WINFract Date: Mon, 4 Jun 2012 10:59:21 +1000 Hi Jonathan, Sorry to pester you, but you seem to know what you are talking about. I downloaded MASM32 and when I tried to assemble BIGNUMA.ASM, I got a large number of errors because the assembler couldn't find bignum_seg. Somehow I still think this is using the old 16 bit memory model. I have played with various options in BIG.INC and still haven't made any progress. I suspect that if we can crack a few simple ideas up front, we can create a fast 32 bit bignum library to link into all our windoze code. Any further tips are greatly appreciated. Thanks, 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@mailman.xmission.com [mailto:fractdev-bounces@mailman.xmission.com] On Behalf Of Jonathan Osuch Sent: Sunday, 3 June 2012 8:30 AM To: Fractint developer's list Subject: Re: [Fractdev] Update to ManpWIN and Problems WINFract Paul,
I've been trying to get orbits to work in the latest versions of WINFract and get a floating-point error when I try to use "Orbits" on the Basic Options dialogue box.
Does anyone have any ideas how to prevent this in the executable?
This appears to always have been broken in WINFract. I'll have a look at it.
Has anyone created a Bignum library for Windows 32 bit using assembler? My Bignum c code is too slow and I only have the sources for 16 bit assembler. I have no idea how to convert it to 32 bit.
Shouldn't be needed. Look in big.inc and comment out BIG16. Since I have never tried this, I don't know if breaks anything. Jonathan _______________________________________________ Fractdev mailing list Fractdev@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/fractdev _______________________________________________ Fractdev mailing list Fractdev@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/fractdev
Hi Jonathan, I can't get it to assemble as I have no idea of how to tell Bignum.ASM that there is an external variable. I thought that it would be a memory model thing. I have no knowledge of how to change this. I think I'll have to let it go. Thanks for trying. Thanks for trying to get orbits to work. Best regards, 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@mailman.xmission.com [mailto:fractdev-bounces@mailman.xmission.com] On Behalf Of Jonathan Osuch Sent: Monday, 4 June 2012 11:05 PM To: Fractint developer's list Subject: Re: [Fractdev] Update to ManpWIN and Problems WINFract Paul, You will find bignum_seg defined in biginit.c. The assembly of bignum.asm shouldn't be a problem, but it does need to be linked to C code that initializes everything. The memory model of the assembly code is not correct for what you are trying to do. I'm not sure what it needs to be in order to work in a Windows environment. I'm still trying to track down the problem with passes=orbits. It doesn't work for Fractint or WinFract, unless running a PAR. It works fine in Xfractint. Jonathan --- pdeleeuw@deleeuw.com.au wrote: From: "Paul" <pdeleeuw@deleeuw.com.au> To: "'Fractint developer's list'" <fractdev@mailman.xmission.com> Subject: Re: [Fractdev] Update to ManpWIN and Problems WINFract Date: Mon, 4 Jun 2012 10:59:21 +1000 Hi Jonathan, Sorry to pester you, but you seem to know what you are talking about. I downloaded MASM32 and when I tried to assemble BIGNUMA.ASM, I got a large number of errors because the assembler couldn't find bignum_seg. Somehow I still think this is using the old 16 bit memory model. I have played with various options in BIG.INC and still haven't made any progress. I suspect that if we can crack a few simple ideas up front, we can create a fast 32 bit bignum library to link into all our windoze code. Any further tips are greatly appreciated. Thanks, 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@mailman.xmission.com [mailto:fractdev-bounces@mailman.xmission.com] On Behalf Of Jonathan Osuch Sent: Sunday, 3 June 2012 8:30 AM To: Fractint developer's list Subject: Re: [Fractdev] Update to ManpWIN and Problems WINFract Paul,
I've been trying to get orbits to work in the latest versions of WINFract and get a floating-point error when I try to use "Orbits" on the Basic Options dialogue box.
Does anyone have any ideas how to prevent this in the executable?
This appears to always have been broken in WINFract. I'll have a look at it.
Has anyone created a Bignum library for Windows 32 bit using assembler? My Bignum c code is too slow and I only have the sources for 16 bit assembler. I have no idea how to convert it to 32 bit.
Shouldn't be needed. Look in big.inc and comment out BIG16. Since I have never tried this, I don't know if breaks anything. Jonathan _______________________________________________ Fractdev mailing list Fractdev@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/fractdev _______________________________________________ Fractdev mailing list Fractdev@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/fractdev _______________________________________________ Fractdev mailing list Fractdev@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/fractdev
participants (2)
-
Jonathan Osuch -
Paul