Ok so, when I installed the drivers on OS/2 Warp 3 (original Millennium 2064w, not anything later), it appeared that the icons transparency is somewhat broken, for example, when the icons are not clicked, they have a strange blackness around them instead of the normal pattern-like appearance seen in...

Are we even sure the Fbitblt command is implemented correctly in the Matrox Millennium? I mean, it shows upside down windows under win9x while moving them, mainly in the login screen and after. I know it can be reproduced on any machine, both i430fx and non-i430fx, win9x (Win95/98/ME), using known d...

Why is it not like the trio64 on the github repo? I mean, the manual of the virge clearly says the same as the trio64 about double word mode.

Just tried NT, first part of installation can be done normally, but during boot, I get gibberish during ntldr or unknown hard error on the dynamic VHD I'm using, this doesn't happen with fixed size VHD or plain RAW.
Note that the file system is FAT16, not NTFS, of the dynamic vhd I'm using.

Well, try generic atapi drivers post-1994.

I believe your driver is asking for a SCSI card in this case, especially that atapi wasn't a thing in 1993.

Just a heads up, the tgui9440 is affected too for future reference.

Glad to see it works

Starting Windows 9x setup (be it prototypes or finals) from DOS is normal practice... And no, betas don't suck because they are buggy, they show you the evolution of what was in development, if you say that they suck, you'll likely say that finals suck and don't try them as well.

Try a, e.g.:, the 430VX socket 7 board.

Probably bad cpu emulation, also what board? Disable apm and acpi if needed.

Well, to start it fresh, on any machine (including the 430fx), this crashes when the drivers are active with the command "line close" being unimplemented, plus, the fact that there are many graphical glitches when using a stub line close implementation (obvious, I know, but that's for test...

You're likely using a tampered copy.

Any news for a 1mb 1024x768x8bpp capable 28800-5/6 bios support? I mean, support for mode 64h.

Have you used the reference disk and the right bios?

I've used the AZT1605 one.
Edit: got news, it tries to use command 0x08 with data 0x55.

NT 3.1 drivers seem to not work at all, even if I select the right settings (Nova 16 Extra drivers), in the sense on the read data available read register it issues a 0x80, then a write data available read with 0x00, then, a read data available one with 0x00, then says the audio board is not the rig...

That "SCSI" boot sequence in the cmos means that the bios in question has got SCSI support only for a NCR/LSI 53c8xx adapter.

I'm greatly interested into this!

resle wrote: ↑Tue 28 Jan, 2020 3:48 am I had a S3 Virge and experienced exactly the same bug. It was a driver bug and yes, it was related to pointer shadow rendering.

If memory serves me right it was specific to a certain driver and reverting to an earlier one, fixed it.

Driver version of the earlier one? I'm curious.

Bump.
This is getting annoying I guess.

I guess this would work.

Yeah, no longer crashes, but now it throws glitches (white parts almost everywhere).

Well, for another bug report of this, 8MB is supported on the Matrox through an add-on module but let's get into the details now. AMIDIAG for DOS, when it's in the part "Initializing Multiprocesor test", it crashes with the following log: "mystique_start_blit: unknown blit 0effffff&qu...

Well, I know, WIP, so it's still early, but are the official and latest Matrox Mystique PCI (1064SG) drivers from 2000-2002 compatible on PCem under Win2000/XP? In the sense, here upon reaching the desktop, it fatals with Mystique busy in the blit.
Happens on any PCI-based machine of course.

Well, whenever I click and drag the cursor over the desktop or inside a folder or whatever, it results in glitches in 16bpp and 24/32bpp modes. Happens in anything that is able to handle a S3 ViRGE or a Mach64GX/VT2 card regarding the PCem configuration used under Win2000 and WinXP, and in version 1...

And I hope it won't have that annoying glitch like the mach64 and virge have in win2000. Pheraps I should open a bug report in the appropriate section about this.
Edit: Bug report done.

Pardon my question, but does the currently in development mystique (on Sarah's end) work on win2k or something outside DOS? I know it's a bit early to talk about this.

Following that after a while, there's this:

As said above, anything that uses the dynarec, even Pentium boards.
Screenshot is here: