What's C's biggest mistake?

[Walter Bright]

Marco wrote:
> bearophile Wrote:
> 
>> Walter Bright:
>>> I find the responses to be very curious, particularly the "not in
>>> the spirit of C" ones.<
>> There are people that think of C as something set in stone,
>> something >that has a "necessary" design. Few years of discussions
>> in the D >newsgroups teach that instead C was not born as a single
>> atomic >perfect thing, it's a collection of design choices and
>> design compromises, >the original authors have chosen only part of
>> the possible alternatives.
> 
> For the most part the C design is complete at this point in its
> language lifecycle with only minor tweaks allowed in the core
> language that don't break existing ISO C90 standard code. C99
> introduced VLAs which was a mistake bolting on a new feature.
> 
> Of course - cleanup of the standard libraries should continue.


VLA's are not a mistake because they are a bolted on new feature. They 
are a mistake because they are poorly designed and have very limited 
utility.

On the other hand, I know the array proposal delivers because it's been 
very successful in D for nearly 10 years now. It won't break any 
existing code, either. And it's simple to implement (easier than VLA's).

It's just a win, all around.