As the guy who wrote the original rant, I partially agree, in that I view forums PRIMARILY as a place for substantive domain problem discussions/ideas, not a place for the 'Community' to fix problems that I believe should not occur if the products are produced through a well-engineered product development process (which includes providing documentation and examples). I ultimately got the GPIO example to work by stitching together some ideas I found here, where people had similar problems with other chips, and some educated guesses. But I am not prepared to say I did it 'right'. That is no way to run a railroad. As an aside, I naively hoped that ST, with a smaller community of engineers that actually have to make physical things work, was not going to behave like Microsoft - which throws things over the transom and lets users ('insiders') find the problems and fix them, all in the guise of a 'community driven process'. BS.
My use case: I have a neural network engine. It took an intern about 3 weeks to port the runtime code to Windows IOT Core (with all the aggravation of making Windows UWP apps) running on a Raspberry Pi. It took me about 3 days to mostly port it (C code, originally designed for chips from Transputers to SPARC and Intel, all from one code base) - no compile/link errors, and then get it running in the FreeRTOS windows simulator. I expected another day to get FreeRTOS to actually run on the DISCOVERY board. After messing around for two days and getting close (the FreeRTOS docs are pretty decent), I decided to go back to basics and try a 'simple' example. When that also took me down a rabbit hole, I decided it was time to speak up. What gets me most is that I know what I can do when I understand the tools and the environment - it is not rocket science (though it does require one to be somewhat anal) to produce an example that a true neophyte can use. I would do it, but right now with my current knowledge, it would be too much of the blind attempting to lead the blind (which, in effect, is what non working demos and poor documentation is). So I know someone at ST who knows what they are doing, could do a lot by closing the office door and deciding 'today I am going to fix this'.
A comment below talks about prestige and budgets and big companies. Well, with a purported 7400 product and design engineers, and $10B revenue, and healthy profits, surely there is some Product Manager who recognizes what a big win it would be to do this, and then get accolades from the 'community' that buys the chips and builds products with them...
my .02 $CI
