[J3] For review: Official US position on features for F202X
longb at cray.com
Fri Mar 8 12:30:03 EST 2019
> On Mar 8, 2019, at 11:00 AM, Vipul Parekh via J3 <j3 at mailman.j3-fortran.org> wrote:
> On Wed, Mar 6, 2019 at 6:15 PM Steve Lionel via J3
> <j3 at mailman.j3-fortran.org> wrote:
>> .. Please look it over
>> and let me know of any omissions or errors. ..
> I suggest item US16 under VI. Declaration and Syntax aids for Generic
> Programming be revised to state, "Useful to ensure .. that two or more
> dummy arguments have the same type." That is, I do not see anything
> in the proposal that restricts the use case to "two dummy arguments”
The descriptions are not intended to be complete. The cited paper is the official source for information. Indeed, one of the use cases is to declare a local temp variable to have the same characteristics as a dummy argument.
> Also, why is US23 "BITS data type" under "Incomplete or Undecided
> Features"? Keeping in mind M218 paper 19-159 and its references,
> particularly 07-007r2,pdf, why would "BITS data type" not be viewed as
> anything other than a complete feature, at least for the purposes of
> J3 list for consideration by WG5? Also from J3 perspective, why would
> this be "undecided" at this stage given the feature made it all the
> way to 07-007r2 draft toward Fortran 2008 standard with past support?
> If the response is simply "process" that this feature hasn't again
> gone through the process of a repeat paper submission(s) with use
> cases, formal specifications, then formal requirements, then syntax
> and review and votes and what not, what does J3 have in mind to fast
> track Fortran feature items which have already been developed
> previously such as the "BITS data type"? What has changed with bits
> in the world of computer science and the Fortran standard since 2008
> that suggests reinventing this wheel?
There should be no uncertainty about the details of the BITS type feature, as this had been already specified in a F2008 draft. It is left to produce a detailed edits document relative to F2018 for reintroducing the feature. (And possibly adding how it would interact with the new Enums, assuming we decide to do those). It was mainly put in the “incomplete” section because J3 / US had not made a final decision to include it as a feature proposal. It’s more the “undecided” than the “incomplete” aspect of the section heading. The important decision comes at the WG5 level. But, I agree with you that it is quite different from the Generic and Exception cases in that the specifics of the feature are known.
> Vipul Parekh
Bill Long longb at cray.com
Principal Engineer, Fortran Technical Support & voice: 651-605-9024
Bioinformatics Software Development fax: 651-605-9143
Cray Inc./ 2131 Lindau Lane/ Suite 1000/ Bloomington, MN 55425
More information about the J3