[photran-dev] Development directions and thoughts from Supercomputing

[Jeffrey Overbey <subscribe-photran-dev@xxxxxxxxxxxx>]

Here are some of my thoughts and post-Supercomputing.  I'd really like 
to hear what others think...

1. I continue to be excited about some of the features our undergrads at 
UIUC are working on (content assist, find all references, error 
corrections, etc.).  The Fortran BOF agreed unanimously that we would 
never defeat emacs, but for the programmers that might be converted to 
Eclipsism, I believe these are the types of features that will do it.

2. Markus's or Mike's C preprocessor from CDT should be integrated into 
Photran soon, and we should worry later about supporting multiple 
configurations in the AST/VPG.  Preprocessor support is critical for 
large projects and should be integrated before we release (and hype) 
search features.

3. Several people are looking forward to OFP integration and a 
solidified AST.  Wyatt would like full location information added to our 
CModel elements rather than just the identifier location.  Beth would 
like a CDT IScope/IBinding interface on the AST, and being able to reuse 
her CDT visitor would be even better.  These are both easy enough to 
add.  I can take care of this and would like to target it for Summer 08.

4. There is lots of interest in C/C++/Fortran cross-language search.  I 
hope one of our new contributors can spend some time on this.  The 
program database we are using is very refactoring-specific and looks 
nothing like the PDOM, but building a PDOM-based index off of it should 
be fairly straightforward.  In a year or two, it will make more sense 
why I want to keep our existing representation around as well...

5. I would like to call our first release Photran 2008 (2009?), if the 
Eclipse priesthood will allow this.  I really wish we were "Photran 2008 
on CDT 2008 on Eclipse 2008," since even I can't remember that I'm 
running "Photran 4.0b3 on CDT 4.0.1 on Eclipse 3.3.1.1 on OS X 10.4."

Jeff