awesome-gbdev
awesome-gbdev copied to clipboard
Y2Dmg Emulator
Description notes that the emulator passes (all) of Blarggs test roms, it however only passes a couple. From the comments left in DmgSystem.cs, only 15 roms were tested out of the total 46.
Until we have a method of quantifiably measuring any particular emulator's accuracy (with a number for instance), I think we should avoid qualitative descriptions.
Only raising this as an issue as many devs are looking for a readable accurate emulator (including myself) and we shouldn't waste their time.
Hi @leefogg and thanks for the feedback.
At the moment, the EMULATORS.md does not require any minimum level of "quality" of the contents it lists, while the Emulators chapter in the main list provides a more accurate selection.
Should we try to outline some kind of criteria that resources have to satisfy to be on the EMULATORS.md list?
I think simple QC on it will do like:
- It builds.
- It runs {some popular game}
- It's description is truthful
How can you define "truthful"? Are we gonna check every documentation file and run tests on the proposed emulator?
If they mention they pass x tests, we'll have to prove that yes. Its only to stop proposals like "runs every game perfectly" from slipping in.
Nothing "runs every game perfectly", apart from the Game Boy itself, so that kind of proposal is excluded anyway. IMO we have to rethink about that list, and decide if we want to keep it as inclusive, or move to a more curated (similar to the main list) resource, providing examples of implementation for different frameworks and languages.
I think curating the list would be more useful- given the purposes outlined above. That said, given this and the list's size, wouldn't it be more relevant to make a separate "awesome-gbemu"/"awesome-gbemudev" list?
Why? What's the advantage of fragmenting something that has the same scope (curated list of resources related to gb/gbdev)? Can't we outline a process for different kind of resource in the same repositories?
GBDev and GBEmuDev actually have little interaction. Taking the Discord as an example, I see few users in both #emudev and other channels.