Al,
On 2004 Oct 6 , at 18.17, Alasdair Allan wrote:
>> It's not clear what to do, though, beyond putting warnings about this
>> in
>> the documentation and README (which I've done, and will commit
>> shortly).
>> It can't be worked round automatically, since setting F77 is a
>> legitimate and documented way of deliberately forcing the choice of
>> compiler. The real problem, here, is that the IRAF startup is
>> behaving
>> badly, in setting up the variable in this way.
>
> Could we check for IRAF being present, check that F77 is set to the
> dumb
> IRAF default, and then ignore it? Or is this a hideously hard thing to
> do?
It's probably messy rather than hard. If you feel that developers
really can't be trusted to read the the documentation or the README,
then the top-level ./bootstrap is the place to put such a test. I do
mention this case pretty prominently in the new README -- do you think
that might be sufficient?
Norman
--
----------------------------------------------------------------------
Norman Gray : Physics & Astronomy, Glasgow University, UK
http://www.astro.gla.ac.uk/users/norman/ : www.starlink.ac.uk
|