Gabor Fichtinger <gabor@xxxxxxxxxxxxxxxxxxxxxx> wrote:
> On the other hand, it is really unfortunate that the user cannot
>know how the NetCDF Data Model is implemented in terms of
>XDR-elements.
If you are interested in understanding how storage is implemented, then
read Section 9.1 of the netCDF User's Guide. The level of detail it
contains should be sufficient for understanding the consequences of our
particular netCDF implementation.
> Interface 2 [the physical storage format] is known only for Unidata:
Hardly, since the source to our implemenation of netCDF is freely-
available. It's not like were trying to hide anything.
>... This the corner stone of the storage space and
>access-time efficiency.
Yup. We look at it as a "quality of implementation" issue.
> Does already exist a specification of Interface 2, other than a
>commented C-source I received from Unidata some time ago?
Nope. And, given our limited resources and backlog of work, one isn't
likely to appear anytime soon. Sorry.
Steve Emmerson <steve@xxxxxxxxxxxxxxxx>
