Difference between revisions of "Talk:Portability issues"
Line 13: | Line 13: | ||
Of course, if you are doing any real floating point in your roguelike, you may be unhappy with the performance on 386's that lack the 387 coprocessor... | Of course, if you are doing any real floating point in your roguelike, you may be unhappy with the performance on 386's that lack the 387 coprocessor... | ||
I'm not assuming anything about processor. What I'm saying is that the memory problem is connected with the hardware, not the operating system. | |||
It has nothing to do with the DOS. It's just the way the hardware works. It works the same under other operating systems on the same hardware. | |||
It's not OS-specific. It's portability to hardware, not portability to OS. |
Revision as of 11:19, 5 February 2005
The memory real-mode/protected-mode is not really tied to DOS, but to the mode in which the processor is working. You can have DOS programs working in protected mode, as well as some operating systems (like CP/M?) working in real-mode.
That doesn't change the fact that when you write for DOS and want to run on every computer that runs DOS, you have to support the 8088 and 80286.
Isn't this page about maximal portability? Don't assume all the world's a Pentium.
The crime isn't assuming all the world is a pentium. It is assuming all the world is a 386 or higher.
Of course, if you are doing any real floating point in your roguelike, you may be unhappy with the performance on 386's that lack the 387 coprocessor...
I'm not assuming anything about processor. What I'm saying is that the memory problem is connected with the hardware, not the operating system. It has nothing to do with the DOS. It's just the way the hardware works. It works the same under other operating systems on the same hardware. It's not OS-specific. It's portability to hardware, not portability to OS.