>Date: Thu, 16 Jun 1994 15:02:07 -0600 (MDT)
>From: ethan@xxxxxxxxxxxxxxxxxx (Ethan Alpert)
>To: netcdfgroup@xxxxxxxxxxxxxxxx
>Subject: ncvarget and array reordering
In the above message you wrote:
>The whole idea here is, if negative edge lengths were allowed, then user
>defined reorderings of data durring reads and writes would be possible.
>Saving both programing time and CPU time.
I believe that this can be done using generalized hyperslab I/O, though,
admittedly, this wouldn't be as convenient as using negative edge
lengths in regular hyperslab I/O.
--------
Steve Emmerson <steve@xxxxxxxxxxxxxxxx>
