[CF-metadata] Future development of CF

From: Jonathan Gregory <jonathan.gregory>
Date: Tue, 11 Mar 2003 19:40:40 +0000

Dear Steve

Thanks for your email on the future development of CF. I agree, this is an
important issue. I think we have to do two things:

* Use whatever personal contacts we have to encourage people to consider using
CF, and to express their needs for development via this email list. I believe
that we are all trying to do this. It concerns me, like you, that there are
only 20 subscribers currently. I agree with John that we should readvertise CF
regularly on the netcdfgroup. Probably every new release could be advertised,
and the first non-beta release, coming soon, will be a good occasion.

* Use this email list to discuss how to proceed as an open forum. The archive
then serves as a record of how decisions were made - a concern you expressed in
an earlier message.

The growing interest is encouraging. To the list you give, I would add that
several climate centres are interested, the Working Group on Coupled Models
(a WMO committee) supports our efforts, and IGBP will consider whether CF
would be appropriate for their proposed Earth System Atlas.

I agree that we should do whatever we can to avoid the standard fragmenting into
local conventions. Some such agreements may be unavoidable, but in the first
place we must consider whether general solutions could be adopted in the
standard. Of course, this takes time. It is harder to think about whether a
particular idea has greater applicability and to frame some appropriate
solution than it is to work out a simple addition that will deal with your own
specific problem. I will make that point on the PRISM mailing list.

I don't think we should put in place any more procedures until we need them,
because bureaucracy slows things down. We have made a lot of progress so far
by focussing on what we actually need. But now CF is being adopted in practice,
we obviously have to make sure decisions are properly debated.

One reason why we might need more procedures is if decisions cannot be made by
a discussion of everyone fast enough, because people cannot make enough time to
think about it. In fact this has always been a problem, as you know. It becomes
more serious as the standard includes things which require more expertise to
understand. People will then either have to make time to learn about the new
issues, or accept decisions made by consensus. We could introduce web documents
recording "requests/proposals for development". To some extent this would
duplicate the email list, but I suppose that such documents would try to remain
short, and would evolve to summarise the results of discussions, until there
was agreement on what would go into the standard.

Cheers

Jonathan
Received on Tue Mar 11 2003 - 12:40:40 GMT