kiibohd / controller

Kiibohd Controller
GNU General Public License v3.0
807 stars 270 forks source link

LCD not restored after key press following Latch #295

Closed dfaught closed 4 years ago

dfaught commented 6 years ago
dfaught commented 5 years ago

Ping.

This is still an issue. In fact, it's worse. Slave side is non-operational using the latest build. Flashed both sides independently, whichever side is the slave is non-responsive.

haata commented 5 years ago

kk, I shall take a look.

dfaught commented 5 years ago

Stepping back through commits, 449d4d9cd5097f129b0ee9b565cab656f8ae282f is the last commit I found where the slave half functions. LCD behavior is consistent with the original bug.

dfaught commented 5 years ago

Bump... still an issue building against HEAD

haata commented 5 years ago

I'm noticing a few issues on Infinity Ergodox with the LCD status. Right now I'm using the lcdFuncMap.kll to handle the state transitions. One possible fix is to move to the internal layer signals. The downside of doing this is that KLL configurations will need to be updated (possible re-written) to work with the configurator (it's the right thing to do though).

dfaught commented 5 years ago

I'm happy to re-write my klls if it will resolve the issue. What do I need to do? I'll also try to get a debug capture of the layer stack tomorrow once I get in to the office. Although the layer behaves properly, just the lcd doesn't return to prior state. Debugging my keyboard beats my real job...

haata commented 5 years ago

I haven't gone through a full set of testing, but this should fix the LCD update issues.

There may still be bugs but these will be more to do with the layer trigger events.

dfaught commented 4 years ago

Confirmed. Built from head this morning. Flashed both sides. LCD is properly restored after post latch key press.

haata commented 4 years ago

Pushed to the configurator now.