[MITgcm-devel] sun os

[Ed Hill]

On Wed, 2004-09-29 at 08:27, Alistair Adcroft wrote:
> Martin,
> 
> The error msg would help.
> 
> We did discuss this and it was a minor mistake to fix last time (SUFFIX 
> order) to recover general make compatibility. We *do* have a policy of 
> using the lowest common denominator make and are not adopting gmake as a 
> standard but the implementation just isn't in line with our policy.  :-)

Hi Alistair,

Whether its your "policy" or not, I've seen Makefile/shell failures on
old Alpha machines as recently as last week and its not just one little
SUFFIX ordering problem.  On some systems, there are problems with the
shell scripts invoked by make and on others its the make syntax itself.

And its a *losing* strategy to keep tweaking the Makefile and the
associated shell scripts to get the combination to work on one ancient
machine only to have it break on some other obscure bit of hardware. 
And all the while we keep adding more genmake and Makefile features. 
Yes, we should (and do!) aim for maximum portability but its just silly
to claim that we aim to support every make and shell out there.

On the other hand, its very easy for both us and our users to point them
towards KNOWN GOOD make and shell implementations that are essentially
100% portable.  In the _rare_ cases where their system-default make
and/or shell fails, then we have a solution!

Ed

-- 
Edward H. Hill III, PhD
office:  MIT Dept. of EAPS;  Rm 54-1424;  77 Massachusetts Ave.
             Cambridge, MA 02139-4307
emails:  eh3 at mit.edu                ed at eh3.com
URLs:    http://web.mit.edu/eh3/    http://eh3.com/
phone:   617-253-0098
fax:     617-253-4464