Harry,
In regard to the debugger dump you included in your last post, it won’t really be of much use. There are disassemblers that turn program binaries back into machine code, but not much more. The area from your debugger image shows disassembled instructions in the lower part under the -u command. It is not as much source code as it is a giant intellectual puzzle. (Frustrating, too.)
When a compiler takes the machine instructions, (or code from higher-level languages,) it not only turns it into runnable binary code, it merges in other bits of code from macros and libraries. After that it relocates the code into runnable program segments. And perhaps even worse, the compiler is often set to obfuscate the original code by moving the compiled segments around to further deter would-be hackers and dissemblers; a modern day spin of the old shell game. But worst of all, actual source code is commented, so a programmer can get some idea of what a given subroutine or module is intended to do, the meaning of passed-in arguments, etc. Disassemblers don’t do that at all.
Reverse-engineering the program up from a binary program file would be a really tough way to get back to code.
And I loved CP/M. It was actually an adaptation of the original Digital Equipment Systems operating system, ported to the 8-bit world, right down to the tiniest PIP command. It was really fun writing the BIOs to bootstrap and run CP/M on your system. – I still have the boards from one of my many S-100 systems, huge floppy drives, including one with an early copy of MS Word on it…
Sent from Mail for Windows 10
Sent: Wednesday, April 25, 2018 3:49:45 PM
To: xywrite@xxxxxxxx
Subject: Re: Kenny Frank and TTG
And (maybe this is up Phil's alley) how hard is it to get from the binary back to something that could be intervened in (to establish page swapping for expanding the memory)?
Back in the day, I did some modification of the CP/M operating system, using only Assembler and DDT (the Digital Research predecessor of Debug). Even Debug has an Unassemble command, you know.
I'm not talking about extensive changes, just something that would add a jmp to some added routine that would swap in memory (I'm talking through my hat here, but . . .).
Don't know if the below will come through, but it's a screen shot of a debug U command applied to the first chunk of editor.exe. It has instructions like: MOV, JMP, ADD, and INT--which are pretty familiar.
Okay, CP/M was 2k and editor.exe is 681k. We're not going to let a little thing like that stop us, are we?
Surely, Harry, since it seems he was short of $170, if he had anything to sell he would have sold it long ago?
By the way - - what makes anyone think that the source code for any version of XyWrite is still in existence? Source code is a thing that can get lost. Macromedia lost some important code about ten years ago . . . these things happen.
And no, XyWin was never usable. Very promising, but unusable.
At 23/04/2018 19:11, you wrote:
We haven't tried buying the source from TTG. Has anyone here had friendly communication with Kenny Frank however long ago, so that he might recognize the name if he or she emailed him?