But the hacked games with the translations and interesting selfmade stages often don't work, but I don't think that will change.
Yea. I wouldn't count on it. It's not impossible though. The new params for the Switch footer format bring options that have not existed before, and those params are not all understood yet. One could quite possibly lead to enabling extended memory access, and that's the biggest thing preventing a lot of hacks and translations from working. Being expanded beyond 4 MegaBytes into ExHiROMs.
I do still look through the asm, and try to make new sense of things. Param63 may allow ExROMs to work, and its just that we havent combined it with the right other param so far. For example, it may be primarily used just to tell the emu where to look in the ROM for the internal header, but not do anything for telling the emu how much ram to allocate for the game. 6A is odd and is being hard to pin down. At times it makes me think it might be for size allocation, but the values just don't match up. The value 0x38 gets converted by the emu to 0x3800, and added with other numbers... and the whole function sets variables based on similar, but different numbers from the preset ID. IDK. I haven't been able to make sense of it yet. 53 is checked in the same function, but its purpose seems redundant and unclear. 50 is a collection of bit flags that could have a wide array of effects... some of which could be for allocating memory. Then there's 74 too, which seems to be a control byte for usage of the params. In a way I cant define and document yet. I just know its values from 1-7 are all individually checked at times, and based off that other params may/may not be checked.
There's a lot going on here to still be figured out. I just need to find a footing so I can have a hope of proposing tests that look for specific/distinct results. Until then, please, I encourage people to experiment and mess around with those params in SFROM tool. Find some interesting results from setting those params to whatever!
Like for example, RadMcFist found that if you set flag 21 of Param 50, it causes an audio glitch with Aladdin. Interesting? To me, yea. Useful for anything specific? No. But combined with other such results... it may help start painting a picture...
This is where I really wish I had a Switch. With the SNESC I was able to test every little idea I had, or run hundreds of tests to find and note results. With the switch... it's just not remotely reasonable to require others to run tests for me and report results. Its WAY slower, confusing, and wastes more than just my time.
If anyone has a spare switch lying around that they don't want... =p