Hi all,
Hmm, there are a few choices here, all with pros and cons. One could
argue that "netcdf" is the file format and anything to do with CF
could be a parameter (e.g. "netcdf;conventions=cf1.1"). This cleanly
separates file format from convention. However, "geotiff" is regarded
as a file format in its own right, even though it could be regarded as
TIFF + geo conventions. This would argue for "cf-netcdf" as the MIME
type. Thirdly, since CF is closely tied to NetCDF it could be argued
that CF implies NetCDF, meaning that "cf" in itself could be a valid
file type, but this isn't typical community usage.
I think this is the distinction for me, there is no such thing as a .cf file
format whereas you can have a ".xhtml" file.
This is true. But there's no reason why there should not be a .cf
file type (or equivalent). In fact I think this might be rather
helpful - but that's probably another discussion. Many systems
(especially web servers and browsers) map file extensions directly to
MIME types and helper applications.
Jon
On Wed, Jul 16, 2008 at 9:22 AM, Dominic Lowe <d.lowe@xxxxxxxx> wrote:
Hi Ethan,
You state "that you can use to write XHTML", whereas I note you don't say "can
use to write CF" ;-)
I think this is the distinction for me, there is no such thing as a .cf file
format whereas you can have a ".xhtml" file.
I fully understand your reasoning though - there are clear parallels between
XHTML and CF. There are also differences and it just depends which
perspective you take. At the end of the day I think either (+ or -) would be
fine as long as our clients/dispatchers can handle it. But my preference is
for "-" :-)