alexeames wrote: ↑
Sat May 19, 2018 8:39 am
With a board as simple as the Pro Hat there is little gained by programming the EEPROM as no pre-configuration is necessary, since it's a completely blank canvas.
I would suggest the EEPROM should be programmed with default data which identifies it as a HAT which doesn't configure any GPIO.
That accurately reflects what the HAT is (as delivered), identifies what it is and that it is fitted (when it is), and what configuration it does (none).
I guess it depends on whether one thinks the HAT ID mechanism is purely intended to configure GPIO or also a means to identify the HAT fitted.
I suspect in this case it had not been appreciated that there is something to be gained for some from being able to identify which HAT is fitted even if it requires no GPIO configuration.
The HAT spec is just for convenience and a useful means of providing for the Device Tree mechanism. It has no legal authority, is not enforceable, compliance with it cannot be enforced and non-conformance cannot be punished, compliance is only voluntary.
For the good of everyone it helps if people do stick to the specification but they don't have to.
I haven't seen any evidence that the Foundation's POE HAT meets the HAT spec or does not but, if it doesn't, I think it is a very poor way to lead by example. I believe it should either be made HAT compliant, given some other name, or the HAT spec must be updated.