I wondered exactly the same thing about this: bsky.app/profile/shas.... Will there be people caring enough, for the 30-40 hours I worked on it? Probably not. Was it fun for me to do it? Absolutely. And as a byproduct, now I can watch this demo that most people don't care about, but matters to me :)

I have more stuff queued to release, but I need permission from original authors, which I don't remember if I have (I wrote most of the ports code ~20 years ago).

Final one is a high quality capture of Myth of Orion / Anaconda. www.youtube.com/watch?v=8oWe.... The previous available capture was 720/30fps, this one is 4k/60fps, which due to youtube's bitrate at 4k, makes it look pretty nice.

Third one is Insane Elements / Anaconda. Code: github.com/shashClp/ins.... Capture: www.youtube.com/watch?v=Dj-N.... This not a port, just fixed the windows version to not crash, be framerate independent and a bunch of small enhancements.

Second one is The Requiem / Anaconda. Code: github.com/shashClp/req.... Capture: www.youtube.com/watch?v=s4NC...

First one is Eunectex / Anaconda. Code: github.com/shashClp/eun.... Capture: www.youtube.com/watch?v=cTlL...

The source code has been uploaded to github (github.com/shashClp/pla...) and the binaries to scene.org (files.scene.org/view/demos/g...)

Yesterday I released all of the work to the public (with permission of the original authors). The 4k video capture can be seen here: www.youtube.com/watch?v=mRvM...

And just in case you are asking about why I am not using DX12 or Vulkan, they are a bit too much for what the project needs, this is a fixed function based RHI :)

I ported it to DX9 a few days ago because it was closer to the original API, but then I hit too many hard to debug issues, and today I decided to port it again to DX11, which is one of the options to get RenderDoc/PIX support. With the RHI being ~800 LOC, it is not that hard and pretty fun.

I'm porting someone else's RHI, written 1999 for DX7, which I don't have source code for, and that doesn't work on Windows 10 for many reasons. I have information of the entry points, so I can infer things. I ported it to OpenGL ~15 years ago, because that's what I knew by heart at the moment.

I have some other projects to fix that use DDraw indeed, but the bugs are on the Window setup code, they don't switch to fullscreen correctly :P

In this particular instance I don't think it will matter much (albeit, who knows), the original code was DirectX7, I did a reverse engineered OpenGL 1.x implementation, and now bringing it to DirectX9. Yes, I am doing really advanced stuff :P

I miss RenderDOC / PIX already, so probably porting to D3D11 once it works to a certain degree

Only 4096 bytes :P