jak-project VS pokeemerald

Github page

If you want to try jak and daxter, try this first before emulating https://github.com/open-goal/jak-project

Jak and Daxter

But now there's NSA's Ghidra which has accelerated decompilations/reverse engineering a ton, it means that within a few years any active community can fully reverse engineer into human readable code and make major changes, it's why now there's a Driver 2 PC port and a Jak & Daxter PC port with 2 and 3 being in the works despite these games having basically no community before it, Nintendo games have a bigger modding community and a lot of games are on the way to being fully decompiled (bonus in that some games share a lot of stuff, so as one project completes it also makes it so that multiple games also advance faster

When you start a new game, the game generates the Dewford phrases (and Feebas tile seeds) from EasyChat words by making some calls to RNG, shown here in InitDewfordTrend and SeedTrendRng. To summarize (assume H16 is the upper 16 bits of the RNG seed on that advancement): It makes a call on Advance x to determine the first word, always pulled from CONDITIONS (H16 % 69) It calls RNG on Advance x+1 to determine whether the second word is from LIFESTYLE or HOBBIES (H16 & 1, LIFESTYLE if it's 1, HOBBIES if it's 0) Then it calls RNG on Advance x+2 to determine what that word will be (H16 % 45 for LIFESTYLE, H16 % 54 for HOBBIES) It calls RNG on Advance x+3 to determine whether the trendiness is increasing or decreasing (not important for this, but it's H16 & 1) It calls RNG on advance x+4 to try setting the maximum trendiness If this value is greater than 50, it calls RNG again (on advance x+5) to try setting the max trendiness. If this value is greater than 80, it calls RNG a third time (on advance x+6) to set the max trendiness (each of these calls are H16 % 98) Once it has the max trendiness (takes the result from above and adds 30, giving a max value of 127), it calls RNG to set the current trendiness (the value itself isn't really important for our purpose, but it's (H16 % (max trendiness value+1)) +30, where the max trendiness value is the value before adding 30. This value also has a max of 127) Finally, it calls RNG to set the associated Feebas seed (H16)

it's used for working with decompilation projects like pokeemerald

This is the main one: https://github.com/pret/pokeemerald/wiki/Adding-new-event-object-or-overworld-sprites

I strongly recommend you to switch your project to pokeemerald as it's more difficult to get those bugs. Also you'll get some tutorials in its wiki.

There are several reverse engineering projects out there which perfectly reconstruct 1:1 accurate C code and compile to the exact same bytes given the original compiler. Please see https://github.com/pret/pokeemerald . Of course you lose names and comments but it is not accurate to say no to this question here. It's perfectly possible to construct recompilable matching C code (purely in this narrow case, anyway.), it's just really tedious and a question of permutation through sets of C fast enough to find a matching member.

You could start by looking at the decompositions like https://github.com/pret/pokeemerald and seeing what resources they have, but it’s gonna require coding and it’s not as drag and drop.

Pokémon Ruby/Sapphire, Pokémon FireRed/LeafGreen and Pokémon Emerald.

I don't know about any sort of translation, but if you're trying to do something specific, you generally try to think of a place in the vanilla game that already does it and check out its script file to see how it works. You can see all the event functions (or macros as they're called in the files) here: https://github.com/pret/pokeemerald/blob/master/asm/macros/event.inc

With enough dedicated people it doesn't have to be, https://github.com/pret/pokeemerald