The long and winding road

The other day, I posted something in comp.lang.fortran in response to a post asking for a new feature in the Intel Fortran compiler. I suggested that the best thing to do was to submit an issue to Intel Premier Support asking for the feature since the more customers who ask for a feature, the easier job the Fortran project manager has in justifying it. This prompted a startled reply from someone who thought that I was the Intel Fortran project manager. "Heck no," I replied, "I'm not even the most senior engineer on the project!" Well, really, I'm not on the compiler project itself anymore, but I still sit and work with those who are. Yes, I started my Fortran career at DEC in 1978, but there are others on the team who have been at it longer.

Stan Whitlock, now he IS the Intel Fortran project manager, as well as our Fortran standards committee representative. He joined the DEC FORTRAN-10 (for the PDP-10) engineering team in 1976. Stan later worked on Fortran for the DECsystem 20, VAX APL, and then rejoined the DEC Fortran team when the short-lived MIPS-based DECstation line was introduced. He's led the Fortran team ever since, through the years of Alpha, Digital Visual Fortran and now Intel Fortran. But wait, there's more...

Dave Eklund joined DEC in 1975 doing support for DEC FORTRAN-10, but support also meant bug fixing and eventually development. He stayed with the DEC 36-bit systems and then joined Stan on DEC Fortran, so he's been doing Fortran continuously for 31 years now.

And then there's Rich Grove. Rich started with DEC back in 1971 on PDP-11 Fortran. Rich was later the project leader for VAX-11 FORTRAN-IV-PLUS, to give it it's full name, and he interviewed me for the job I eventually landed on the VAX Fortran project. Rich later became the project leader for DEC's GEM code generator and optimizer, which powered the DEC compilers for MIPS, Alpha and IA-32 (with Digital Visual Fortran.). When Rich joined Intel, he was named an "Intel Fellow", an extremely senior position in the company, with the role of "Compiler Architect". While Rich doesn't work daily on Fortran, he sits just a couple of offices down from us and keeps Fortran in mind as he helps shape the future of Intel compilers. (Edit: Rich retired from Intel in 2007, after a long and illustrious career.  We had a wonderful dinner in honor of Rich, attended by many who had worked with him in the past.  Rich is now enjoying his grandchildren and still drops by from time to time to say hello.)

I should also mention Peter Karam, another Fortran compiler developer who has been with the project since he started in 1980. Peter tried being a manager for a while, but soon found that his heart was in development so that's where he returned and what he does today.

As you can see, there's a core of dedicated engineers who have guided a set of Fortran implementations for more than three and a half decades, starting with DEC, through the couple of years of Compaq, and now Intel. (We missed HP, which bought Compaq shortly after we joined Intel, which was a bit more than five years ago.) Of course, there's more to the project than just these folks, including many developers who have worked on compilers for 20 years or more (and some young whippersnappers, too.)

I once told a group of customers about our long Fortran heritage and that we were "a bunch of old farts". One of the group replied, "That's good - Fortran needs old farts."

P.S. Doctor Fortran now has his own URL. You can find the doctor at http://www.intel.com/software/drfortran I'll be posting more regularly in the future.
Para obtener información más completa sobre las optimizaciones del compilador, consulte nuestro Aviso de optimización.