Pavel Kirienko

@facontidavide Thanks! But the issue is still there though :D

An architecturally cleaner solution is to remove the template parameter constraint that the `BaseType` shall be a native integral and allow the user to instantiate `fpm::fixed` with `BaseType` being a...

> weird types such as float or classes A generic type is expected to generalize some behavior for any T irrespective of its identity as long as it meets specific...

Faster telemetry should not be a problem. We should consider making this tunable when migrating to UAVCAN v1.0.

We are aiming at early June. If you could lend us a hand here, it would be appreciated. Also, may you be interested in sharing the info about your project...

> We've got experienced firmware engineers on the team so building and testing any sapog changes are well within our ability What I meant is that if your engineers could...

We use it in [PyUAVCAN](https://github.com/UAVCAN/pyuavcan/) -- a Python implementation of the [UAVCAN](https://uavcan.org) protocol stack.

> Regarding #13, I've given it some thought over the weekend. Where I may be willing to contribute is in (custom) test hardware for the project (Such as AVR, STM32,...

As far as I know, for MCP2518FD there are no AEC-Q/ISO26262 qualified options, which is likely to be a deal breaker for some high-integrity systems. TCAN4550-Q1 is functionally identical to...

Nice. @dagar can we adapt it to support CAN FD (it seems to support only Classic CAN at the moment) and refactor it to remove dependencies on Libuavcan v0? Much...