WiiMote Control

Enhanced IR Support and some bugfixes.

Instead of just one (mostly useless) IR Output, there are now 4 (additional) X/Y/S triplets for each of the tracked IR points. It's important to note the S (not Z!) -- it's the IR point's size. Bigger typically means closer :)

The WiiMote is only capable of tracking 4 IR Points, so please don't ask for more points; As Pierre would say, "It's Not Possible" :)

The output pitch/roll angles are now 0-360.

The patch no longer disconnects when switching from fullscreen to windowed modes.

There is now a disconnect input on the patch, to attempt to reconnect if it's acting strangely.

Creating a patch no longer prevents you from creating more wiimote patches.

Maybe some minor stability improvements. Nothing noteworthy though.

I'm thinking that for the next beta, I'll maybe include multiple patches: * WiiMote + Motion Sensor (no IR) * WiiMote + Motion + IR * WiiMote + IR * WiiMote + Motion + Nunchuk * WiiMote + Motion + IR + Nunchuk * WiiMote + IR + Nunchuk * variations on the above, with the classic controller.

Thoughts?

More modifications. Some for stability.

after reviewing many crash logs, it appeared as though Connect and Disconnect events were the leading causes in crashes, mostly connects.

Connects were crashing when LED/Rumble stuff was updating. At first, I thought this was because the updates were getting sent to a stale (disconnected) Wiimote object. However, this wasn't quite the case. Instead, it was usually crashing because the wiimote connection, while established, was not ready for LED/Rumble updates, so we were sending them too soon (often as soon as we connected). I added a 45-frame timer that delays LED/Rumble updates, and this seems to have resolved some connect-time crashes (as well as the occasional perpetual blinking lights issue that I rarely observed).

I modified a few more pieces of the framework to handle notifications a bit more swiftly, and that looks like it has corrected the previously noted "_nsnote" crash mentioned in the previous beta. Sometimes it will crash when QC exits though, in "_cleanup" -- Still looking into this one.

The Pitch and Roll outputs also output full 360° values now, so you can emulate the WiiMenu rotating cursor. They're not properly ranged though (-270 - +90 or something), I'll try to get to that in the next beta.

A bunch of stability improvements come with this release of the WiiMote patch.

Connect and Disconnect are now fairly stable; There's at least one hard-to-track bug left (you'll see a "NSCFNumber doesn't respond to selector 'start'" exception sometimes, where something is getting released before it should), but many of the outstanding bugs are now handled. Thanks to all the Beta testers for submitting dozens of Crash Logs that helped me isolate many of these issues.

There have been 94759301 requests for better stability with this patch; Hopefully this helps :)

Other fixes: rumble and led state is updated properly on reconnect now, and you can create a patch after you delete the first one (before, creating a single one prevented you from creating more until you quit QC and restarted)

[PS, it's also built for 64-bit cpus now too]

This version supports framework-level timeout handling (where it should be), so you can now reliably wait as long as you want before connecting. Still a bit picky, but at least it's not due to randomly timing out anymore.

Hot on the heals of the previous WiiMote Control is this very new beta. Shortly after releasing yesterday's version, I gutted it and reimplemented the connect/disconnect parts.

Cool New Things:

• Force Feedback support
• LED Control
• Bluetooth timeout error is now handled, and appropriate cleanup/restart action is taken behind the scenes to expand the 15 second window indefinitely [didn't quite work]
• Hopefully more stable