[Go-essp-tech] Minutes of ESG meeting 2009/08/07

[Pascoe, Stephen (STFC,RAL,SSTD)]

Dear Colleagues,

Bryan, Ag and I have prepared some minutes of the teleconference last
week and put them on our IS-ENES wiki at
http://proj.badc.rl.ac.uk/isenes/wiki/EsgMeetings/Telecon20090807

Comments and points that need further clarification are marked in bold
with a vertical bar comment mark.  Please point out any inaccuracies and
answer anything that you can.  I will be preparing a short document
distilling the issues BADC/IS-ENES are concerned about for our next
teleconference that we hope to arrange for mid September.  Invite to
follow.

Cheers,
Stephen.

---
Stephen Pascoe  +44 (0)1235 445980
British Atmospheric Data Centre
Rutherford Appleton Laboratory

-----Original Message-----
From: Dean N. Williams [mailto:williams13 at llnl.gov] 
Sent: 07 August 2009 12:25
To: Cecelia DeLuca; Lawrence, BN (Bryan); V. Balaji; Michael
Lautenschlager; Don Middleton; Luca Cinquini; Bob Drach; Sylvia Murphy;
Pascoe, CL (Charlotte); Pascoe, S (Stephen); Stephan Kindermann; Gavin M
Bell; Ann Chervenak; Rachana Ananthakrishnan; Nathan Wilhelmi; Nathan
Hook; Bernholdt Dave; Frank Toussaint; Stephens, A (Ag); Philip Bentley
Subject: Today's Agenda in a couple of Hours

Dear Colleagues,

I don't know if I have everyone listed on this e-mail exchange, please
forward to those that I may have missed. Sketch of today's agenda by
Bryan...

1) Discuss scope of this telecon. Expected outcome: a clear
understanding of how data and metadata flow within the ESG (+friends)
consortium. Who is responsible for the data flow, what software needs to
exist, who is responsible for it, and what the timelines for delivery a)
are, and b) need to be.

2) Data Flow. Modelling centres create data, and ingest it into their
ESG data node. How does data flow to the ESG core? How does the
catalogue metadata know about replicants of data at modelling centres?  
What happens for modelling centres who are not ESG nodes or don't have
mates who are ESG data nodes that can take their data?  What does the
current list of expected modlling centres look like?  When and how do we
expect core replicants to move. How will the catalogues know? Who is
keeping track of all that, and how are they updating the rest of us?

3) Metadata Flow. How does the metadata from the questionnaire get to
the ESG team? What about the gridspec files? How do we stitch the three
sources of metadata together (cf output, model metadata, gridspec). Who
does it? Where? What are the common elements of metadata along which we
do the stitching?

4) Vocabularies. What vocabularies are we managing? Where? Are they in
common? If not, what's the route to commonality?

5) Revisit expected outcomes: can we write down a clear sequence of
activities, and a clear data flow, with unambiguous activities resulting
in tools and/or products? Do we have a clear timeline end-to- end, and
clear ownership of every step? Can we have a commitment from ESG to
publish their milestones, deliverables, and progress
*REGULARLY* on the go-essp-tech list?

6) How do we expect to evolve these data and metadata flows?

Best regards,
	Dean
-- 
Scanned by iCritical.