Interest Check: Ymir Sega Saturn Emulator

General Development
#1

Hello everyone,

I’m writing to ask if there is any interest from developers in porting Ymir (Sega Saturn emulator) to libretro / RetroArch .

Ymir is a relatively new but very promising standalone Sega Saturn emulator , focused on accuracy, clean design and low-level emulation choices (including proper BIOS handling and CD block emulation). At the moment, however, Ymir is not available as a libretro core , which means it cannot benefit from RetroArch features such as:

  • CRT / post-processing shaders
  • Unified input, scaling and overlays
  • Frontend integration (arcade cabinets, Big Picture setups, etc.)

Many users (myself included) are currently using Ymir standalone and applying external solutions like ReShade, but a native libretro core would be a huge step forward for both usability and preservation.

This post is not a demand , just an interest check :

  • Would a libretro port be technically feasible?
  • Is anyone familiar with Ymir’s codebase interested in exploring this?
  • Are there known blockers (license, architecture, rendering backend, etc.)?

I believe Ymir could become a very strong Saturn core alongside existing options (Beetle, Kronos), especially for users focused on accuracy with modern tooling.

Thanks for your time, and thanks to everyone involved in libretro and Saturn emulation.

2 Likes
#2

While I agree with you, the topic below should clarify many of your questions:

1 Like
#3

warmenhoven got a port working really quickly and has opened a pull request to get it on the upstream repo: https://github.com/StrikerX3/Ymir/pull/746

Someone just happened to ask about ymir on the libretro Discord at the right time. warmenhoven saw that and when looking at the code realized it wouldn’t be hard to libretro-ize.

4 Likes
#4

Just saw that the ymir author responded to warmenhoven’s pull request. Of course, I’d like for there to be a libretro core but I’m fine with whatever ends up happening.

So for now tl;dr we’ll see, basically.

1 Like
#5

Would be nice but judging from the response i got in their server when i asked about a retroarch port (shot down like a criminal), he’ll never do it. Or probably won’t anyway. Like 1 in a thousand

1 Like
#6

Realistically, I think it’s a given that if there was to be a core (well, there already is one) it wouldn’t be made or maintained by StrikerX3.

Yeah, I know that there’s some hostility by some (not all) standalone emulator authors (often maybe more from the user base) to RA/libretro. From what I read, some of the MAME folks for example really don’t like RA it seems. But… at the end of the day, there still is a MAME core that is maintained and is up to date with MAME (and contrary to what anyone tries to say, i have not stumbled on a single case of something that could run on MAME and not the core, including old obscure non-arcade systems). And the libretro core is not breaching any license or anything like that afaik.

So again, some of the MAME folks might not like it but MAME is an open source project that can be freely forked (with, of course some caveats depending on the licensing and so on), which would include the libretro-ization core.

My point is, if someone really wanted to make and maintain a separate Ymir fork/core, there’s not much the Ymir author could really do.

After that, if the Ymir author was really against it, maybe warmenhoven would simply decide to respect the original author’s wishes and not pursue/maintain the core any more. I don’t know. Which was what my “we’ll see” comment above was mostly referring to.

1 Like