I find it a great resource but don't trust it 100%, some problems only seem to appear on LCD, some are very card implementation dependent, not chipset, and some red fails have workarounds where minor and continuously irritating problems don't. But I guess that's what he saw on the cards that he tested, on the system he tested on with the monitor he had. YMMV as the saying goes.

Hm. I went from S3 1MB to Millenium 2 4MB back in the day. If Matrox was that bad at running DOS games, I would have known. I am 99.9% sure I beat Quake 1 on it, and that Jazz Jackrabbit worked fine.

In the past couple of years, I have been interested in researching different sources of VGA adapter incompatibilities for the SNOOP project: SNOOP.EXE: (S)VGA adapter info tool and CRT Terminator config utility

This is because with CRT Terminator: CRT Terminator Digital VGA Feature Card ISA DV1000 it is possible to read back information from the video data that the VGA adapter outputs, in order to come up with programmatic test cases that find if a particular VGA adapter has a certain known VGA bug or not.

Based on that functionality, I have tested several VGA adapters. Hercules Stingray Pro is one of them. In my test lab I have this card:

which tests out as follows:

There are a couple of hardware bugs that I find:

First bug is not actually a bug of the card, but a side product that the card is a fast PCI card. So it uncovers a programming bug that the authors of Commander Keen games wrote: Keen code assumes that the ISA bus should be "slow enough", or animation can stutter: Why does Commander Keen 4-6 hardware scrolling glitch on ATI (Mach) PCI video cards?.

Second bug is that when running the Kukoo 2 demo, part of the VGA palette colors are missing. This has something to do with internal timing of the demo, or the internal synchronization of palette updates. Hard to say. When using CRT Terminator to view the Kukoo 2 demo, this bug does not manifest. (so it is some kind of an internal RAMDAC thing on this card)

Kukoo2 on digital output (correct):

Kukoo2 on VGA (missing some colors on background effects, i.e. background is black):

Third bug is that the video mode needed by Warlords 2 is not available. I haven't had time to figure out what mode exactly Warlords 2 is trying to initialize (if it's VESA or VGA bios SVGA extended mode number), but the screen stays in the text video mode, and game audio starts to play, then the game hanging:

(if someone is knowledged in the inner life of Warlords 2 code to come up with a disassembly or small test code to reproduce this bug, I'd be eternally grateful)

There is also some kind of unfortunate issue that occurs with interlaced 1024x678 and 1280x1024 16 colors modes. They get signaled somehow incorrectly shifted (cropped), I'm not quite sure why that is. (could be just a CRT Terminator thing). Progressive modes don't have this issue.

Apart from those issues, I find that Doom, Pinball Fantasies, Jazz Jackrabbit, Commander Keen 5, Sim City and Windows 3.1 work just fine on this adapter. In particular, the 1024x768 256c progressive video mode is supported, which is just fantastic!

Higher refresh rates work as well. Here is Word Perfect 5.1 in 800x600 72Hz progressive:

and my personal Tetris record scores: (Win 3.1 800x600 progressive)

This card is certainly one of my favorites, as it also supports 15bpp, 16bpp and 24bpp video modes. Here is 640x480@24bpp:

Coming back to the gona.mactar.hu page - it is possible that different copies of a card, or different motherboards or mobo BIOSes might find different set of compatibility in their testing. Also, if a digital LCD panel is used to test instead of a real VGA CRT, then it is possible that slightly different signal modes could cause synchronization issues to digital LCDs with VGA inputs. That is why in some of those tests results could be a bit surprising: gona.mactar.hu doesn't really dissect the root causes of each issue, so there is some "YMMV" factor possible.

Thanks for that info clb, I'll test my adapter with same procedure as you. This - https://www.youtube.com/watch?v=DfUJjNkNdI0 - is definitely not happening in my case.

My model is a shorter one - https://dosdays.co.uk/media/hercules/stingray … pro_1_small.jpg. Computer is https://theretroweb.com/motherboards/s/trigem-milano.

Warlords 2 Deluxe work normally. I will check Warlords 2.

The back and forth stuttering on Keen 4 in that video is exactly the programming bug that Keen authors had written: they mistakenly assumed that the ISA/PCI bus would always be "slow enough" so that reading the bus a hardcoded 5 times (or 10 times with the Fix Jerky Motion fix) would take long enough to cover a hblank period of the card. But with PCI cards, that is not the case, and one can complete some 20-30 bus reads while in a video hblank. As result, the scroll registers got updated in an incorrectly synchronized manner, leading to the stuttering seen on the video. (This stuttering is exaggerated on some PCI VGA adapters more than others due to a different design convention around register latching)

I also have a TGUI9680-1 on board and I've tried it out. For that chip it's all green on the list. But there is absolutely no visual difference in Commander Keen.