Re: Loops (was Re: do { quit; } else { })

bart <bc@freeuk.com> writes:

On 17/04/2025 20:55, Keith Thompson wrote:

bart <bc@freeuk.com> writes:

>

Here's what you wrote upthread:
     What's far more useful is that it naturally works with i64
     so has limits 4 billion times bigger than you get with C's
     default 'int' type.
Here's what you could have written:
     What's far more useful is that it naturally works with
     i64 so has limits 4 billion times bigger than you get with
     32-bit int.
That's clearer, shorter, and correct, and would have avoided this
entire subthread.  I would not have commented on it.

>
But it misses out the important bit. Clearly i64 has 2**32 times the
range of i32; that will be the case in any context.

I don't see what "important bit" my proposed wording would have
missed.  Can you clarify?

What's relevant here is that my language's "don't care" 'int' type
defaults to 64 bits, while C's 'int' type, on pretty much any PC from
this century, will default to 32 bits.

I don't think "default" is the right word here.  I'll explain why
if you ask.

Which means people are more likely to encounter problems involving
wrap, or UB.

Right, which would have been crystal clear if you had written
"32-bit int" rather than falsely asserting that 32 bits is some kind
of "default".  The phrase "32-bit int" makes it clear that you're
discussing implementations on which int is 32 bits wide (which is
very common), without falsely implying that a width of 32 bits is
required or "default".  And it would demonstrate an understanding
of the difference between "C" and "most implementations of C"
(or "most hosted or PC implementations of C).

Whether C's 'int' should have moved to 64 bits on 64-bit machines has
been discussed here in the past.

More correctly, whether C implementations for 64-bit machines should
make int 64 bits has been discussed here in the past.  C has always
permitted int to be 64 bits if chosen by the implementation.

That would have kicked such problems a /long/ way down the road.

Yes, for those implementations, but it would have caused other
problems (which might be an interesting topic for another thread).

--
Keith Thompson (The_Other_Keith) Keith.S.Thompson+u@gmail.com
void Void(void) { Void(); } /* The recursive call of the void */