Regarding the "TheOtherGCCBasedFortranCompiler" Wiki Page

[Paul Thomas]

Brooks and Tobi,

I really appreciate this exchange going on.  Like Brooks, I have felt very uncomfortable about this Wiki page; not in small part because of the issues of professionalism of the gcc effort that he raises. I had a stab at diminishing the emotional content of the page a few months ago.

 I would like to see the first two and last two paragraphs retained with a link to a new page containing the middle section.  The reason for this is that a potential user of either compiler is not likely to be looking for a history or legal lesson.  Instead, I believe that we should try to give a brief summary of the advantages/disadvantages of both:

1] g95 is more standard savvy and does one or two f95 things that gfortran should but doesn't (eg. aliassing of components of derived type arrays - let me say again that we must tackle this one!).
2] gfortran benefits from being a component of the gcc suite.  It therefore has a pool of developers/maintainers and gains from all the backend improvements. Also, I am encouraged to the belief that gfortran will eventually overtake g95 in the facilities that it has to offer, as the incorporation of Cray pointers and gomp show.
3] gfortran's runtime performance is better,largely as a consequence of 2], I think. I withheld the g95 performance figures from my report on the Wiki.  I think that I will add them this weekend. 

Best regards

Paul T