paper on OO performance with gfortran

[Damian Rouson]

> On Aug 11, 2015, at 9:08 AM, Daniels, Marcus G <mdaniels@lanl.gov> wrote:
> 
> For what it is worth, as a user, I would start using it right away.  

Thanks for this email.  User feedback is very useful.  When I submit feature requests to compiler vendors, a common response is that they are not hearing user demand for the features.  We have a chicken-and-egg problem wherein many users are hesitant to adopt features that aren’t broadly supported, but many compiler teams are hesitant to invest resources into implementing support for features that users aren’t requesting.  I’ve heard of several people who are quite pleased with PDTs so it must be worth the effort. 

> It's an appealing way to reduce content coupling in performance-sensitive parts of a code.   Allocate and deallocate in critical paths are not that appealing.  

It would be nice to see some code snippets demonstrating your points.  

Damian