[Info-vax] Rust as a HS language, was: Re: Quiet?

[chris]

On 04/09/22 13:22, Don Baccus wrote:
> On Saturday, April 9, 2022 at 5:11:23 AM UTC-7, Bill Gunshannon wrote:
>> On 4/9/22 00:31, Dave Froble wrote:
>>>
>>>
>>> I doubt that many, if any, choose assembler because they like it.  At
>>> times it is the best choice.  That appears to have decreased.
>>>
>> I always like assembler. Have done some recently and expect to more
>> in the near future. I know the assembler for at least 10 processors.
>> But then, I am a dinosaur.
>>
>> bill
>
> My point is that since Chris seems to reject language design decisions that restrict his "freedom" to screw up, why isn't he still writing assembly code?

Perhaps I can answer that myself.

It's essential to have complete transparency and audit trail
from source code right down to the bare metal where necessary.
Any language using dynamic memory allocation makes that far more
difficult to show fully deterministic behaviour. That is the
reason I don't do much work in C++, for example, though some of
ideas in C++ are quite elegant and quite usable in C. Fully
deterministic behaviour is often mandatory for real time systems
and can make any code more robust.

Still happy programming in assembler where necessary, but can be hard
work building even moderately sized systems using it. Also, it makes
maintenance difficult, since few now are really fluent in even
a single machine assembler. Since maintenance can account for
the majority of ongoing development costs,  try to write systems to
be maintainable, even though it means added work. Finally, it's
far easier to express abstract concepts in a higher level language
and C is just high level enough to do that, without added features
that I don't need. As I said, the swiss army knife of programming...

Chris