georgel wrote on 2021-09-03, 18:53:
digger wrote on 2021-09-02, 20:34:
If you read my quote again, you'll see that I mentioned compatibility with Microsoft's port trapping API, not QPI (QEMM). From what I found on-line, 386MAX is compatible with the Microsoft API, including the port trapping stuff, and can even be considered a superset of it. To be fair, though: I haven't actually tried that myself yet. It'll be fun to experiment with this in the weekend. I was just wondering if someone here had already tried SoftMPU with 386MAX already, and if so, what the results were.
I am aware of the misleading info that you quote. A rumor is a weak word for it -- it is 100% lie.
If you are referring to this article on Programmer's Heaven, I would consider it a bit harsh to call it flat-out "misleading" or a "100% lie", since the writer does clarify that he hasn't tried it himself and doesn't know for sure.
You seem quite sure that it doesn't work, though. Have you confirmed this by trying this out yourself?
Also, through some more googling/duckduckgoing on this topic, I stumbled upon the source code of usbjstik, a DOS legacy joystick driver/emulator for USB joysticks, which uses the same "INT 2Fh, Function 4A15h" API and contains multiple references to 386MAX support in the code. Also, in this forum post, Bret Johnson implies that 386MAX is supported by (and verified to work with) the usbjstik TSR.
@elianda Do you think the usbjstik source code that I linked to above would also provide a nice reference for the SoftMPU developers to add 386MAX support to SoftMPU, if that doesn't yet work out of the box (other than the EMM manager detection routine not yet taking 386MAX into account)?