future plans for F2003

[dave_frank]

Now that Intel is WAYYYY-BEHIND on having a compiler that meets current Fortran standard, when might users expect a release that does?

'05 ? '06 ?

[Steven_L_Intel1]

Dave,

Intel Fortran already supports many features of F2003, so I would dispute your assertion. We'll continue to add F2003 features in the future.. That's all I can say at this time.

I will note that it typically took 2-3 years after standardization for major vendors to offer compliant versions of F77, F90 and F95. F2003 is a larger leap from F95 than any of those previous standard revisions.

[keefer]

Dave,

I've read a summary of the proposed standards for F2003. Except for three items, such as allowing a component of a derived type to be declared as an array, most of the specification pertains to making Fortran look more like C++, e.g. inheritance. These issues would seem to me to be of most importance internally to a programming group and not externally to a customer/client. They clearly would make compiled code much harder to optimize. The three new features that would have the greatest impacton an individual programmer, or small group, I believe are already incorporporated into V8.1, or soon will be. Intel may be "wayyyy-behind" the "legal" specification, but, IMHO,they are not way behind in developing a reliable implementation.

Regards,

Keith

[stigandersen]

> Intel Fortran already supports many features of F2003...

Dear Steve,

Could you point me towardssome documentation showing whichF2003 features are currently supported by IVF? (The section on Fortran 2003 Language Features in the Intel Fortran Language Reference appears to list "only" features related to TR15581)

Kind Regards and Thank You in advance,

Stig

[Steven_L_Intel1]

We don't have this specifically documented. In addition to TR15581, there are many smaller things that are standardization of existing practice which Intel Fortran supports, such as VOLATILE. We'll document this in the future.