Reply to thread

Message

[QUOTE="donLouis, post: 1700293"]
It is perfectly possible and sensible to use the colloquial sense when
deciding whether it is OK or not from the standard's point of view.  So,
I fail to see your point about OK not being defined by the standard...

If it doesn't invoke undefined behaviour and it doesn't require a 
diagnostic, it is OK from the standard's point of view.  OK?

The original statement was made in the *explicit* context of a C standard.
You had no business to *implicitly* ignore that context in your reply.

By this silly logic, it is not OK to return EXIT_FAILURE, because your
environment will complain.  Yet, a hello world program that doesn't
check the return value of the output call has no a priori reason to 
return 0 any more than EXIT_FAILURE.

This is not a valid argument when deciding whether a certain programming
construct is OK or not.  The OP wasn't asking for personal opinions, but
for facts, and the fact is that the C89 standard explicitly allows it.[/QUOTE]

I think that the problem is that dogma takes precedence over
the standard.
[/QUOTE]

Verification