Using the same controller and config on two PCs

[A500+]

Originally Posted by A500+

From memory they're missing but I'll double-check that for you. I'm not at home right now but should be back in front of my desktop machine tomorrow so I'll check then.

I can confirm they're missing on the "wrong PC" when looking in the remap/test window. If I click on 'Test' and press any buttons the controller is recognised but the existing mapping is missing or ignored - it's as if a different controller entirely is plugged in.

I still can't seem to get it to output the full logging information to winuaebootlog.txt. I have no idea what I'm doing wrong. Instead I launched via the command line passing the -log parameter, went through the process of loading the config manually and checking the gameports and input panels then exiting. I then copied the contents of the console window and pasted in into a text file. I've attached that, along with the usual, seemingly incomplete winuaebootlog.txt to this post. I hope it helps!

Originally Posted by A500+

Not to create or add to any confusion, but on a possibly(?) related note, I set up a "fresh" copy of WinUAE and the folder structure I've been using (I'm trying to keep everything self-contained and portable so I can have multiple copies/versions of games and Workbench etc. that I can use anywhere) but now I can't get any of the paths to take at all. I input them in the Paths panel using relative paths but as soon as I go to any other panel then come back to the paths panel they've all reset to default. This happens every single time no matter what I try. 'Use relative paths' is checked, 'Portable mode' is checked, 'WinUAE default (EXE directory)' is selected in the drop-down list, 'Start in Quickstart mode' in unchecked... I don't know if I'm doing something badly wrong or whether this is another bug but it's infuriating. I'm using the beta exe for this little test and I've set it up - or I'm trying to at least - the same way as the others which all work fine (controller mapping issue aside). I'm stumped. It's very possible - probable even - that I'm doing something wrong somewhere but thought I'd mention it at least in case it's relevant somehow to the problem at hand.

Edit: Ignore me, I'm an idiot! There was a spelling mistake in one of the paths. It's working now I've corrected that. My bad! At some point when I get chance I'll map a controller with the new, clean setup and see if it works on both machines. I'll let you know if it does, otherwise I'll remain silent on the matter so as not to add any confusion or unnecessary information.

Sorry for the wall of text!

[Toni Wilen]

winuaebootlog-console.txt DOES have "missing" messages and it also logs "fixup" message where the mapping is corrected to match new device indexes.

Does the controller also work correctly when "parse_custom_post" and following line has "-> ..." line?

EDIT: better use -log command line parameter when testing because it forces everything in console window. winuaebootlog.txt to winuaelog.txt transition might hide them without enabling full logging.

0 0: 0 -> 1 (NO MATCH) (HIDI2C Device)
0 0: 1 -> 0 (Wireless Controller)

This another important bit. First line means "HIDI2C Device" was missing when config was loaded (but existed when config was saved). Device name is still remembered and kept and moved to slot #1.
Second line means "Wireless Controller" is found but now has different slot (1) and was moved to slot 0 (because "non-existing" devices are always moved to end of list, if it has space)

EDIT2: Except for some reason following lines use original joystick ID again..

[A500+]

Originally Posted by Toni Wilen

Does the controller also work correctly when "parse_custom_post" and following line has "-> ..." line?

Sorry, I'm not quite sure what you mean. Are you asking if the controller works when that line appears in the log? If so I don't know but I can go back and check.

I know I said I wasn't going to confuse the issue by posting the results of my new test, but it might offer some more insight perhaps as although it still doesn't work it now exhibits slightly different behaviour. So I set everything up and mapped the controls on the desktop this time (the previously non working machine), synced the files to the laptop and tested the config with the same controller on that machine. The controller still doesn't work but this time if any button on the controller is pressed in-game then the mapped keyboard controls stop working! They work fine before a button on the controller is pressed. That doesn't happen the other way around - with the other config(s) the keyboard controls carry on working across both machines even if buttons are pressed on the non working controller.

For clarity, and in case this has anything to do with it, in both cases I map keyboard controls and DS4 controls to Custom 1 in the game ports panel, as I like to have the option of using both/either.

I can post the log and config from the new test if you like?

[Toni Wilen]

winuae.7z updated. Now fixup is done in a more logical position which is also earlier. Should also appear in the log earlier now.

[A500+]

That seems to have fixed it. I'm literally on my way out the door now so I only had chance to do a quick test with one controller on one game but it looks promising. I'll do some more extensive testing first chance I get tomorrow and let you know how it goes.

Nice work man

[Toni Wilen]

Great. Test also saving config in another PC (use different config file name) and copy it back to other PC and check if it also works.

Hotplug does not yet work correctly (if config is already loaded when configured device is inserted). It will be fixed later today.

[A500+]

Okay I've done some extensive testing and everything seems to be working beautifully. The configs mapped on the laptop now work with either/both controllers on the desktop and vice-versa. I tested with several games and several different configs.

As suggested I created a new config on the desktop and mapped new controls using one of the controllers and after syncing back to the laptop both controllers work fine using that config.

Looks like it's fixed. Sterling work sir! Thank you.

The only thing to note - and I assume this is actually the expected/correct behaviour - is that with a 2 player config using the controller that was originally mapped to player 2 is still mapped to player two even if it is the only controller plugged in. Before (on the "working" machine) it would switch to player 1 if it was the only pad and would only be assigned to player 2 when both pads were connected simultaneously. Like I say though I'm assuming that was to do with the bug and the new behaviour is actually the correct/expected behaviour?

It's not a problem either way as I managed to solve it by mapping both controllers to each port - when either one is plugged in it defaults to player 1 but when both are plugged in they are assigned player 1 and player 2 respectively, thus replicating the "old" behaviour, if that makes sense?

I'll happily test the hotplug fix for you when it's done too if you like.

Thanks again for your continuing efforts to improve WinUAE - they are much appreciated. Looking forward to the next release

[A500+]

Originally Posted by A500+

It's not a problem either way as I managed to solve it by mapping both controllers to each port - when either one is plugged in it defaults to player 1 but when both are plugged in they are assigned player 1 and player 2 respectively, thus replicating the "old" behaviour, if that makes sense?

Actually after more testing that hasn't fixed it. Is there way to map two controllers such that they will occupy player 1 and player 2 slots when they're both plugged in, but either one plugged in on it's own will only occupy player 1 slot? It's not a big problem if not - I can live with it. I'm just curious is all.

[Toni Wilen]

Originally Posted by A500+

The only thing to note - and I assume this is actually the expected/correct behaviour - is that with a 2 player config using the controller that was originally mapped to player 2 is still mapped to player two even if it is the only controller plugged in. Before (on the "working" machine) it would switch to player 1 if it was the only pad and would only be assigned to player 2 when both pads were connected simultaneously. Like I say though I'm assuming that was to do with the bug and the new behaviour is actually the correct/expected behaviour?

I am not sure if I understood. "Player" assignments should not change (selected custom entry is not changed).

Or did you mean selected custom #x entry was mapped to connected pad even if it was supposed to be mapped to other custom #x entry (in other joystick port)?

Originally Posted by A500+

Actually after more testing that hasn't fixed it. Is there way to map two controllers such that they will occupy player 1 and player 2 slots when they're both plugged in, but either one plugged in on it's own will only occupy player 1 slot? It's not a big problem if not - I can live with it. I'm just curious is all.

I don't think this is possible automatically.

btw, winuae.7z updated, now also hotplugging should work correctly.

[A500+]

Originally Posted by Toni Wilen

Or did you mean selected custom #x entry was mapped to connected pad even if it was supposed to be mapped to other custom #x entry (in other joystick port)?

Yes. This is what happened originally, though I'm assuming it was a "happy accident" of the bug. It had the effect of essentially always making a single controller occupy the port assigned to player 1, regardless of whether or not it was that specific controller that was used to create the mapping originally. Only when both controllers were present did it assign each to their own respective port.

Originally Posted by Toni Wilen

I don't think this is possible automatically.

I didn't think so either but thought I'd ask in case I was missing something. Like I said it's not a problem - I just need to remember to carry the player one controller around with me. If not it's no great hardship to change the custom slots around anyway. I was just being lazy

Originally Posted by Toni Wilen

btw, winuae.7z updated, now also hotplugging should work correctly.

Great. I'm away with work at the moment until Wednesday but I have my laptop and both pads with me so I'll get to testing later tonight if I can. I'll let you know if it works. I'll also test on the desktop once I'm back home on Wednesday too.

[A500+]

Apologies for the delay in reporting back - sometimes real life gets in the way of the fun stuff

I can confirm hotplugging works like a charm on both machines. I tested several different games and didn't find any situations where it failed.

It even works on a two player config, so long as the controller originally used to map player 1 is used first (using the "player 2" controller first still starts a two player game even if it is the sole controller present).

Thanks for your efforts to fix these issues. Looks like your hard work has paid off. I'll continue testing and let you know if I encounter any unexpected behaviour but so far everything looks good