OSX Volume+/Volume- behave erratically

[DanMcLaughlin]

Hi, I'm using the firmware as generated by the version on the TrulyErgonomic site on a couple of 209 keyboards. I (and others) have found that the Vol+/Vol- keys behave erratically. The operating system apparently keeps receiving the keycode, so the volume maxes out or goes to zero. I haven't yet verified if this is the case on Windows. It happens erratically too, sometimes it will work, other times it goes to the end immediately, and it also will do it strangely sometimes.

Dan

[yurikhan]

I have received reports of such behavior. Now that you confirm it, I can investigate. If it won’t reproduce on my machine, will you be able to compile and run a test program that listens for HID events and prints them out?

[DanMcLaughlin]

Hi Yuri,
Sure no problem. I've tried it with two keyboards on two computers. On one is consistently has the issue, on the other isn't as bad, but if I mess around with it enough it will reoccur.

Dan

[yurikhan]

I have reproduced the problem.

When pressing a modifier+key combo such as Fn+F12, you can have two event sequences:

• press Fn, press F12, release F12, release Fn;
• press Fn, press F12, release Fn, release F12.

We are taught that key combinations are pressed the first way, but in quick presses the fingers tend to execute the second sequence and in most cases software copes with it.

The firmware sees that physical F12 was pressed while Fn was down and generates a logical Vol+ press event. But, in the latter case, the release of physical F12 happens while Fn is already up, so it is logical F12 that gets released, while logical Vol+ stays down until somehow cleared.

This affects all media (more precisely, media, power, application launch and application control) keys which are placed in the Fn layer, if the OS autorepeats them. (If it doesn’t, the key is still stuck but the effect of the key only occurs once.)

I will think of a way to handle this release sequence. Meanwhile, the workaround is to hold the Fn key a bit longer.

[DanMcLaughlin]

Excellent work Yuri, I'll do some experiments and see if I can reproduce this idea.

[DanMcLaughlin]

Hi Yuri, Is this change pushed to the TrulyErgo website configurator?

Dan

[yurikhan]

Probably not yet; I don’t have push privileges there. But it is live here on Github Pages.

If you have a bookmark with a saved layout, you can just replace the part before the #. (You may need to compose the link in a text editor and paste that into an empty browser tab, not just copy-and-paste the hash over the loaded configurator.)

As a sanity check, the fixed firmware .hex file should be 19485 bytes long.

[DanMcLaughlin]

Finally got around to updating my firmware from the TECK site - works great! Good job on the fix.