2 May 2017, 21:37
@Substring thanks a lot! I will try these points and give a feedback then.
Nintendo Adapter Box
-
@lackyluuk
not it s another
your id vendor and id product are differents, you need to add my informations and try if you didn t do it.oh not raphnet need a kernel patch too
-
@acris yes I have tried your info, but no luck
-
i think kernel patch is missing in recalbox 4.1
-
Ni idea if it requires a patch to the kernel, i don't think so as it works fine in es, right ?
-
@Substring yes, in ES it does what it should. But no response at all in a game. This only occurs with the 4NES4SNES adapter. The one for N64 works fine.
I have connected these two adapters with a small usb hub. -
@lackyluuk ok, this is what you should do :
- test pads with evtest, i think they should work
- same goes with sdl2-jstest
=> if both of them work, then it's a retroarch problem - out of curiosity, try to remap the pads in retroarch so that we can see if retroarch can see them afterall
Give a try to https://github.com/recalbox/recalbox-os/issues/1244#issuecomment-2981808465 because it has an updated retroarch version, i know i had to rewrite 1 patch regarding inputs, so maybe that would do.
Still, if it works in ES, it means it works with SDL and linux kernel, so that is a retroarch problem. You could eventually try some non retroarch emulators with these pads (mupen scummvm, advancemame, reicast)
-
@Substring thanks a lot! I will try these points and give a feedback then.
-
@Substring I have found some patches for this problem I think. They are from raphnet itself. Do you think this could work?
http://www.raphnet-tech.com/support/retropie/
(And we should let him know about recalbox, haha ;))
-
@lackyluuk 4.4 kernel doesn't seem to require these patches. Anyway please test what I suggested. The quriks are not necessary (or were they already applied when you submitted your support archive ?).
So, still, i believe it's a RA problem.
-
@Substring Yes of course, I will test it tonight when I am home.
The quirks from @acris were not implemented when I created the support archives.Yes, the quirks from the second point are only for pre-4.2 kernel. But do you think that the first point "Issue with initial values" (the 3 patches) is also covered through the updated kernel?
http://www.raphnet-tech.com/support/retropie/usbhid_iostart.diff
http://www.raphnet-tech.com/support/retropie/usbhid_start_before_connect.diff
http://www.raphnet-tech.com/support/retropie/jsdev_initial_value.diff -
@lackyluuk your controller does work with ES, so i'd say yes. I'll have a look at the kernel code though
-
Results:
-
evtest works super fine
-
sdl2-jstest works fine as well
-
Config in retroarch not successful
-
TADAA: I have tested Mupen and the SNES pad worked even in-game... so definitely a retroarch problem
-
Installed the unofficial new version with retroarch 1.5 --> no success
Maybe let them know about this 4NES4SNES v1.5 adapter?
..pheww..
-
-
@lackyluuk even switching to the input mode to sdl (or sdl2 i'm not sure) did not help ?
-
@Substring I have tried 'udev' and 'sdl2' with the previous version. I should try it tonight with the new verzion again. This went out of my mind..I'll let you know
-
@lackyluuk I think it won't help ... i should eventually ask retroarch (the problem is that we slightly patch RA regarding inputs).
Please send me a reminder next week regarding this issue, this week is a busy IRL recalbox week (we're all going to HFS Summer 3 and will be having a conference there to talk about Recalbox).
-
@Substring alright. I will try it anyway, just curious
Ok, great! Aw, it would be nice to go there as well and meet the recalbox team in real
But it's a bit too far -.-
