You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The string protocol is nice and easy to debug, but there's enough variability in parsing time to make it a problem, at least when dealing with an array of Pi 1's (Pi 2's parsing time is almost constantly 0.2ms, Pi 1's varies between 0.4ms and 1.2ms which is enough to warrant refactoring the protocol).
The concept of the protocol can probably remain the same (a separated protocol module with class decorators used to handle parsing and construction of messages), but messages will be strictly binary. This will probably restrict data-types for parameters to something simple to handle (32-bit integer and 64-bit float only ... perhaps a one-byte for a bool). This might be a good opportunity to integrate #33 as well (make part of the header a variable length list of units?).
The text was updated successfully, but these errors were encountered:
The string protocol is nice and easy to debug, but there's enough variability in parsing time to make it a problem, at least when dealing with an array of Pi 1's (Pi 2's parsing time is almost constantly 0.2ms, Pi 1's varies between 0.4ms and 1.2ms which is enough to warrant refactoring the protocol).
The concept of the protocol can probably remain the same (a separated protocol module with class decorators used to handle parsing and construction of messages), but messages will be strictly binary. This will probably restrict data-types for parameters to something simple to handle (32-bit integer and 64-bit float only ... perhaps a one-byte for a bool). This might be a good opportunity to integrate #33 as well (make part of the header a variable length list of units?).
The text was updated successfully, but these errors were encountered: