Re: data access functions
Samuel A. Fineberg (fineberg@nas.nasa.gov)
Tue, 11 Mar 1997 15:25:16 -0800
In message Rajeev Thakur writes
>
>> Date: Tue, 11 Mar 1997 11:20:13 -0800 (PST)
>> From: "Ian E. Stockdale" <ies@nas.nasa.gov>
>>
>> In view of the changes to the I/O chapter over the last
>> several meetings, it seems like it might be useful to look at the
>> evolved text to see how the newer changes interact with the details
>> which were already present. There may be some constraints which were
>> determined based on requirements which no longer exist. Similarly,
>> some choices may have been made to preserve functionality which has
>> since vanished. Small modifications to improve the standard without
>> making any fundamental changes might thus be possible.
>
>I agree.
>
>> As an example, I'd suggest that that the data access routines
>> be required to manipulate data in multiples of whole filetypes. This
>> would make these routines easier to implement. This would also
>> simplify the user interface, and does not appear to remove any useful
>> functionality.
>
>I think this would be a big change though. For one, it would make the
>concept of etype redundant, because of which we would have to revisit
>many definitions.
Actually, it would not make etype redundant. Etypes are still needed to
ensure compatibility between the buftype and filetype. I.e., you would
still need to ensure that the filetype and buftype are constructed of
the same etype. Now, there may be other problems, but all of my MPI-2
I/O based codes read and write complete filetypes.
Sam
-----
Samuel A. Fineberg Address: Numerical Aerodynamic Simulation
MRJ Technology Solutions NASA Ames Research Center M/S 258-6
Supercomputer Applications Segment Moffett Field, CA 94035-1000
e-mail: fineberg@nas.nasa.gov Phone: (415)604-4319(voice) (415)966-8669(FAX)
See my home page at URL: "http://www.dcs.uky.edu/~fineberg/"