You are reading my playbook.
If you took a look at the SII Annual Report you can see that I have had some good ideas. In the very beginning of my computer carear I worked at machine language and assembly levels.
It appears that .NET Core has been a major investment for Microsoft. It is starting to be in wide use. Yet interest in the important capability of interfacing of multiple OSs is not highly appreciated at lease at the interface level.
i.e. There does not be a specification that discusses and describes the minimum interface required from the OS.
It is possible to dig through GitHub but that is just a load of methods.
I’m sure Microsoft has a specification of the minimum methods.
They just didn’t start out one day, spend millions, and hope for a result.
As far as the transcompiler to C#, it would also be capable of a translation to C++. The purpose of that might be to build the bit twiddling portions of the kernel. Yet I want to focus on C#. At my company SII we had an internal conflict between the OPL coders and the Apple advocates. That type of “religious” battle is not productive. That is why I want a very limited use of C++ where absolutely necessary. Probably, at the interrupt level and CPU type, although I intend to focus on the ARM architecture.
I have attached an example of the OPL code for RING’s OPL Compiler.
If you look at the “hat” names. this^is^his^name , they were easily and perfectly mapped to Camel Case for C#. thisIsHisName That worked perfectly! And readable.
Lots of other operations have been completed including translating OPL Intrinsics such as OPL’s $SIN to C# Math.Sin.
Arithmetic expressions were also accurately translated.