Nice! I didn’t actually try to do the build with that setting, since I saw some other things were missing, and I didn’t think it would work on the PicoCalc without modification. I spent a lot of time looking over the code but I guess I should have just tried to build it with that setting.
That said, only changing the compilation option and using “WEBRP2350” instead of “PICORP2350” finally allowed me to see the keybaord error others have reported. Doing a WebMite build this way will cause “I2C Keyboard not found, OPTION KEYBOARD disabled.”
The good news is, that might help track down what’s causing the keyboard problem. As far as the build goes,
“WEBRP2350” has most of the components of “PICORP2350”, but it’s missing:
- GUI.c
- -DGUICONTROLS
- pico_multicore
And it has a larger PICO_HEAP_SIZE and PICO_STACK_SIZE set (it’s 0x2000 for both with just the Pico2, and 0x4000 for both with WebMite).
It has a few extra things, but those all appear to be related to running the webserver in the background.
There are also plenty of changes in the code itself. For instance, I’m not sure why, but overclocking is locked down and on WebMite builds you can’t go above 252000, while in the regular PicoMite build for the Pico2 it allows going up to 396000. (I’ve only personally set it to 380000 to play it safe, but the code allows that little bit higher.)
Since multicore was there for the Pico build and it showed the keyboard error, I’m wondering if it’s a question of processor speed, or something else. I was running WebMite at it’s default limit of 252000, but I still got the error. And I’m assuming people using guu’s build may not have even changed the value from the default (which I think is 150000). So it’s probably not the speed after all. I’m not sure what’s causing that error.
I’ll upload a firmware binary release on the github for the WebMite version if people want to try it, but be aware that you’ll get the keyboard error eventually, at least until that gets sorted out.