Open Ellerbach opened 10 months ago
Not sure if this proposal was discussed in a meeting I wasn't part of, but I have some questions about the reason why this would be needed. I thought in general we agreed on two models:
The whole reason why we added GpioPin was so that we wouldn't need something like this (which seems to be an approach that uses an emmulation to mix the two scenarios above) since devices can just take individual pins, and we also agreed all devices would get constructors that took Gpio Pins in favor of this. Am I missing something here?
@joperezr yes, we discussed in one of the call. As we have a lot of bindings to adjust, idea is to have a transparent way to use a virtual GPIO controller! So having a virtual controller will solve this issue of having to add all those new constructors and adjust all the internal codes to support GpioPin. Because, it's not just adding ability to have a GpioPin in the constructor, it's about adjusting all the logic in all the bondings to replace GpioController.Write, Read, etc by the GpioPin.Write, Read, etc.
Adding virtual GpioController and virtual GpioPin. This allows to recreate a GpioController from different GpioPin created form potentially very different GpioController.
Microsoft Reviewers: Open in CodeFlow