trying to get a second controller to work without breaking the first

Ok, I think I’ve almost solved everything. I have the hotkey buttons mapped to the autoconfig files to my taste. The only problem I have left is that the core remap happens across all controllers. This means if I have the buttons adjusted on my XB1 controller, they are incorrect on the 8bitdo controller and vice versa.

What I need is for NES “B” to be assigned to X on the XB1 and NES “A” to be assigned to A. On the 8bitdo, I need to keep B with B and A with A. And this needs to happen for all NES games (Nestopia), leaving the other cores alone.

Is there a way I can do that?

As a reference I found this discussion: Controller Override Per Core, Swap Buttons

Z3R0B4NG wanted the changes to happen across all cores, so , yes. The autoconfig file is best for that. However, in my case I just want per-controller remaps to happen only in Nestopia.

edit: Gah. I just can’t stop trying to think about making it work. Ultimately I think I’m at a roadblock. unless the remap file has the equivalent of an #ifdef, I’m stuck. Only one or the other controller will work. Not both in turn because the remap always has priority and overrides all controller settings. I was so close to figuring it out and having a truly automated couch experience.

I still want to see if I can get this working at some point, but without looking deeply into the repo, it’s not clear what I can/can’t do. I’ll clone the repo as soon as I get a chance to crosscheck, test, and submit the corrected modkit autoconfigs (busy for a bit).

After that, I’ll try to investigate the remap code and see whether there are any opportunities. My malloc-fu is very very (very) weak right now, but I could see if there’s a natural place to get controller-specific remapping to work. It would be a new idea/feature, however, and if it conflicts with how you guys want the retropad to work I won’t pursue it. I’d still, however, lobby for some kind of solution as this seems to be one weakness of the system.

I’m reporting back in on testing. I may have goofed again. It turns out that VID 045E and PID 02E0 is actually correct if you start the controller in Windows mode. The VID/PID I’ve been reporting has been the 8bitdo in Android mode.

So question. Both the N30 and SN30 share the same PID (02E0). Retroarch loads the N30 incorrectly as an SN30, probably due to loading order. Should I just go ahead and submit the files with the Android values and make a note about the Windows mode conflicting?

Hi.

I’m back after a hiatus. I apologize for not getting a pull request in. Life and travel took over in the interim. It looks like the repository has the numbers corrected for the Android values right now so everything’s good.

1 Like

I’m actually having this problem ATM. My Retroarch just can’t recognize my N30 Modkit controller right now (based on NES Classic Controller), and due that I can’t bind any configurations on my FCEUMM-core. Any help for this?