[log in to unmask] writes:
> If this is an oversight...
> is there a mechanism to correct this before potential harm?
If it were necessary, that would be the kind of thing the
interpretation process is for. In this case, I doubt there is any
substantial question of the intent. It is explicit in the f77
standard, and some might argue that the words in 4.3.1.3 of the f90
standard are also adequate, although I find them a little vague for my
taste. Any such interpretation would be more a matter of formality
than of really settling a point of ambiguity. So I regard it as more
a matter of curiosity and as something that might be worded better in
the future than a matter critical to fix. Others might, of course,
differ.
There have been interpretations that clarified subtle but important
issues that multiple compilers got "wrong". Those compilers were
eventually fixed to conform to the passed interpretations, although it
sometimes took a while to make it through the vendor release cycle.
I really can't imagine a vendor getting this one wrong. If they did,
they'd get feedback from their customers a lot faster than through the
standard interpretation process, though I have no doubt about what the
answer that came out of the interpretation process would be.
--
Richard Maine
[log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|