B
Ben Bacarisse
Bartc said:Hans Schneider wrote:
I think until recently lccwin didn't allow type specifiers in any order,
even though the C standard rather unwisely didn't prohibit this.
There was a problem (now reported as fixed) that involved the new C99
type "long long" but that version, quite correctly, accepts "signed
int short". I am skeptical about the claim, simply because lcc-win32
is based on a well-respected C parser from lcc. I suppose that it is
possible that there is an even older version of lcc-win32 that did
reject the declaration but I doubt it. I think the OP is being
provocative.
Presumably
for historic reasons.
I can't think of any historical reason to reject that syntax and, as I
have said, I would be surprised if (plain) lcc rejected it.
Now lccwin seems to go along with other C compilers rather than take a
stand. So you can now write every declaration in at least N!
different ways.
I think it would have been very unwise to have done anything else. Of
course, a flag "-Wodd-type-specifier-order" might be useful.