Clarifications on license

[aronson]

Is GPL-3.0 appropriate? I've seen people want to link this to decomps, and those aren't GPL-3.0.

By including this and linking it into any source decompilation project or binary ROM hack it would have some pretty severe implications.

Have you considered dual licensing this under MIT or zlib? It's effectively unusable in serious projects in this state, AFAIK. IANAL.

AFAIK there are no FOSS, clean-room reimplementations of sappy.

[ipatix]

Huh, that's kind of true actually. @easyaspi314 Would you be fine with changing the license?

[ipatix]

I'm currently considering the following: Either MIT or MPL 2.0, what would be the preferred option for you?

[aronson]

MIT is well vetted and seems the best to me

[DaforLynx]

There has reportedly been talks to merge this into the pokeemerald-expansion project, which apparently requires such a license switch. MIT is the simplest and most ideal, I would think.

[aronson]

Honestly if you all have a good social relationship, the state of pret and the expansion projects was never cleanroom... it's unlikely anything worse would happen. You're already in hot water with GF as a whole. Merging it in would just mean closed-source modification source trees with a released binary (through .IPS/.UPS patches I hope) would be subject to GPL inquiries which doesn't really change anything from the state of things now (those developers are still going to ignore these inquiries, and no one has money in that scene to act on that most likely, though IANAL). Perhaps ipatix can remove the other contributor's changes for a dual-licensed branch in the source tree including MIT if you're wary? Without a response from the other contributor that's the only option at this point.

Bringing in the HQ mixer has other consequences, such as dooming any hope of a port of those closed-source mods to GBARunner 3. I know last I discussed that problem with the "Jedi council" there wasn't much interest in supporting the hypervisor anyways, though.

[ipatix]

Is there anything in gba-hq-mixer which prevents gbarunner 3 from working?

[aronson]

Self-modifying code

[ipatix]

I guess it would be possible to patch that out at the cost of more memory. But as you might guess by the opening date of this issue, this probably won't be my first priority.

[aronson]

It's possible to patch in GBAR3 by listing in its game-specific configuration all memory locations where the self-modification may happen. However, this has to be derived for every mod and those that share the same title codes/IDs can't be distinguished out of the box. Doing it within the application is too expensive AFAIK