Bedwyr

Yeah, I know the workup. This will be awhile as it's not in my interest to throw myself at the problem, dedicating time until the event happens. It'll have to be piecemeal as I divide it between regular life tasks. I'll also have to set a time limit on removing everything and testing. So what I'll do is keep a count of game iterations dedicated to replicating the problem. I'll turn all that stuff off and relegate Retroarch activations to the SNES through BB, opening BB directly and disabling Kodi auto-startup. If nothing happens after, I don't know, 10-15 iterations of at least 15 minutes of gameplay, I'll call it good and add things back in. That work for you?

Logitech keboard and mouse (K830/MX Anywhere, SN30 modkit from 8bitdo, XB1 controller, and Steam controller (but just in Steam). I also use Controller Companion, all to make a couch-based environment work. However it's set to have a disabling exception for BB and LB (which I checked) and I also shut it down to see if it would clear the issue. It didn't. BB is also running through its Kodi Launcher, but I think if that were the problem then it would have happened on launch and not when exiting the game. Possibly its autohotkey functionality is borking on entering or exiting Retroarch.

Just ran into it again. - Ended all non Windows tasks in batches alt-tabbing back into BB to test. No change; d-pad works, top button (X on SNES) loads box art. Nothing else. - Video snap of the game (Final Fantasy II US) plays after alt-tabbing out and back in. - Turning on and switching to my XB1 controller: no change. D-pad and top button (Y) works. - Keyboard, no change. - Cannot Alt-F4 - Can Ctrl+Shift+Esc - Windows key responds correctly - Mouse does not appear in BB. I've attached today's log. Debug 2018-10-05 10-18-49 PM.log

I figured as much. Any steps you want me to try or test when it happens again? I can patiently plug away at it. Also: no rocket launcher.

I'm afraid I had to shut the program down, but I don't think that was the problem. I was alt-tabbing in and out as well as using the mouse. No change in behavior. One thought is whether it could have something to do with the BB Kodi launcher. I don't *think* so, but it's possible. Another is that it might have to do with functionality after an update. As in it takes the combination of the update and leaving a game to trigger the behavior. I'm kind of shooting in the dark because I can't replicate what triggers the problem and its fairly rare (every 8-10 runs?). Also, Retroarch is always the emulator I'm using when this happens but that's just because I prefer to use it for most/all platforms.

Hi guys. Random post. I reported a bug/partial-crash here: https://bitbucket.org/jasondavidcarr/launchbox/issues/4557/big-box-encounters-occasional-random I'm leaving BB loaded for the moment and changing power settings to not put the PC to sleep. If anyone wants me to take any investigatory or troubleshooting steps, now's the time.

Hey, I just got back from a trip and haven't checked the Retroarch repository yet to see if someone else beat me, but I did get the modkit working correctly. The autoconfigs I have are finally correct and use the right Bluetooth IDs. If you want to look at the discussion thread to see my learning process, look here: https://forums.libretro.com/t/trying-to-get-a-second-controller-to-work-without-breaking-the-first/17433 It's a bit long, but I think the system is better run in Android mode because for some reason, the N30 and SN30 (SNES modkit) report the exact same PID of 02E0 in hex. Which is stupid and I can't figure out why. I never got a final response so when I get a chance and if it hasn't been done already, the advice will be to assume Android mode and use the proper VID and PID for that. It'll confuse the hell out of people starting up in Windows mode, but at least it avoids the PID conflict. Hope that helps.

Just a short note for you guys for the record that my best experience so far has been the Parellel core for Retroarch. The Vulkan-based emulator is zippy enough that I've had few-to-no slowdowns and far better compatibility (so far). I have not tested it with major edge cases like SW Pod Race (may the developers' bacon burn) or Pilotwings 64, but a notable decline in visual artifacts and mis-displays. Nothing is going to fix the texture quality. It is and will remain a problem for texture heavy games like Dr. Mario 64. But at least it's not worse and sometimes better, especially if I don't use filtering like bilinear. Mupen Plus is still a very good emulator, but I still find the Python frontend to be unreliable and the command-line of the core emulator cumbersome. The Retroarch core also displays more visual artifacts in games like Ogre Battle 64. Same with Project 64 2.x, but on a different mixture of games. Keep in mind that I'm running on a Coffee Lake i7 and a fairly hefty GTX 1060, so YMMV, especially on performance limited systems like the Raspberry Pi.

Just a PSA for posterity. Y'all, if you hooked up your Wiimotes to your PC via the Mayflash Dolphinbar, make sure the firmware is upgraded or the IR LEDs stay on perpetually and burn out. You'll be able to make the Bluetooth connection just fine and do basic gyro motion control, but without the light references, the Wiimote can't track position. Download here I believe: http://www.mayflash.com/support/download/

This weekend if I get some time, I'll try out one of the autoconfig files someone submitted as part of a Github issue and see what happens.

I want to report this and see if there needs to be a bug report opened up on Bitbucket. I was playing a SNES Retroarch game and alt-tabbed out to consult a website guide on Firefox. Tabbed back into the game later closed it. Window focus went back to the website guide. I alt-tabbed again to BB and noticed that it had partially frozen. I could get sound effects of menu navigation but no cursor movement or change in page view. I ctrl-shift-esc'ed and force closed BB. The Windows task bar has now disappeared (I assume that BB suspends explorer while it's in full screen?) and I need to do a reboot. More than likely this won't be repeatable; it'll just be a transient crash or error, but I wanted to outline the events while they're still fresh in my mind in case this is useful. Now to go reboot from cmd.

Yep. If I start in Android mode and try to pair, Windows 10 simply does not pick it up at all. I'll wait until a formal autoconfig is approved for Retroarch and try again then.

It doesn't really work like that. I can't start it in Android mode and successfully pair with Windows. It has to turn on in Windows mode.

FYI I'm trying to help one of the Retroarch developers. Apparently someone submitted the autoconfig file but got the pid/vid wrong. Link here: https://forums.libretro.com/t/trying-to-get-a-second-controller-to-work-without-breaking-the-first/17433

Well that was interesting. So now I understand what Retroarch was doing. The controller is auto-detecting as RB8-64 under xinput. I have no idea why, though. The button numbering on user 1 in the autoconfig file reads the same as yours except for select which... does not exist on the n64 controller. Curioser and curioser. Is there a way to force which device the autodetect discovers? I'm wondering if forcing it to read as 8Bitdo_Pro_NES30_BT will help at all.

You mean try your layout in the autoconfig file? Sure, I can try that.

So going through all the buttons: B 0 A 1 Start 11 Select 10 Rpad +0 Lpad -0 Upad -1 Dpad +1 Now if I do a button swap (hold select + down) B 1 A 0 This means that with the current connection I actually can't use the DiY kit with Retroarch because the button input that supposedly maps to Retroarch's B button doesn't actually exist. I suppose I'll have to contact 8bitdo support unless someone has some insight or can correct some misunderstanding I have.

Thanks for responding. Yeah, I'm trying it again when I'm not dead tired and I still can't get any change proper change. Moreover for some reason my keyboard f1 key won't go into the menu. Would it help if I reported the input button number when I try and set a key? (like 0-11 kind of thing)

Well, no luck this evening unfortunately. The controller options are really limited it turns out. If I press Sel + down I can swab A and B and X and Y but I cannot switch them independently. This is too bad because A is correct while B is set to "turbo A" (mario jumps with A and jumps repeatedly with B, no running) I can't seem to figure out how to get an input profile to stick only to the one the device; not sure what configuration files keep the controller assignments. Changing the button assignments on the 8bitdo also changes my XB1 layout and I have to change it back. I've tried changing the buttons assignments several times and cannot seem to figure out what the B button is mapped to. Yeah, this is definitely not one of those moments where the Retroarch controller schemes "just work".

http://download.8bitdo.com/Manual/Mod-Kit/Mod-Kit-for-NES-Original&Classic-controller.pdf So if I'm reading this right, I may have paired in some other mode because I only paired using the start button? And what I should have done is pair by turning on via start+up?

Update: I got the Mupen core working more to my tastes on Retroarch. I think I'll stick with it for now as the most user-friendly and personal setup friendly option. Now it's just a decision whether to do an Ocarina playthrough on the big screen or the remastered 3DS version...

Yo all. I just got a few kits I ordered to convert my NES and SNES classic controllers into 8bitdo Bluetooth controllers (seen here: https://shop.8bitdo.com/products/mod-kit-for-nes-classic-controller). It looks like things connect alright; however the button auto-layout is incorrect. I need to go to work shortly, but I'll spend some time now or on the weekend dealing with button configuration. My objectives: - The XB1 must remain my primary controller and configuring the new one cannot displace the old configuration. - The XB1 uses a savestate scheme where R-stick + (key) control Retroarch functionality. Example: R-stick + start = Retroarch menu, R-stick + select = exit Retroarch, R-stick + dpad-down = save state, R-stick + dpad-up = load state. That scheme has to stay (I worked hard to get other emulators to play nice with the scheme). - I will need to get the bluetooth controller layout correct. Right now I think B is mapped to A and A is mapped to A-turbo. Select doesn't work. - It will try to get a second custom savestate scheme to work, controlled by select similar to Retropie. If anyone sees any roadblocks or "ummm"s, let me know.

It's fine, fernandoq03. We'll suss out what you're saying, no problem.

It'd have to be the codecs then. CPU utilization when I do this is rarely above 30-50%.

Only external factor I can think of is that I'm running a Handbrake ripping process right now. Maybe the decoding process is interrupting playback somehow???