On-Board Slave Changelog
There is a problem with the bit-bang
slave bus code interacting badly with versions 3.1 and 3.2 of the
On-Board Slave code. It appears to work with all other versions, though
version 3.7 is best at this time.
- Re-introduced the high bit that determines full report vs. a mere
lighting command. Made it work this time. A simple lighting command
will not cause the switches to be reported.
This minimizes an interaction problem with C2Pro firmware versions 3.00
and 4.00 that use the new bit-bang slave communications code.
- Now uses a shorter timeout once a message has started (as opposed
to the "nobody is talking on the slave bus" timeout, which is still
about 130 ms). This will speed recovery from message timeouts and free the bus sooner.
- Ignores odd-numbered steps when calculating acceleration, but ONLY when calculating acceleration. This makes the acceleration detection much smoother and more accurate, hopefully improving the user experience.
- Disabled the high bit detection added in 3.2, it doesn't work.
- Disables the version and date reporting sequence which seems to
hurt the hardware controller so that we can use the newer,
bounce-resistant encoder code.
- Added the high bit to distinguish between lighting commands and status report + lighting commands.
- Version/date sequences now use a custom format.
Version 3.1, 2009 Apr 03
- Improved the way the RIC inputs are debounced and reported, borrowed from the PFC 430 code.
- Translated the 'C' code of the original on-board slave.
- Added version/date reporting.