Unidata Users Committee
17 September 2012
|Tom Whittaker (Outgoing Chair), Univ of Wisconsin
Kevin Tyle (Incoming Chair), Univ at Albany
Michael Baldwin, Purdue Univ
Martin Baxter, Central Michigan Univ
Anne Case-Hanks, Univ of Louisiana Monroe
Jennifer Collins, Univ of South Florida
Bart Geerts, Univ of Wyoming
Steven Lazarus, Florida Inst of Technology
Patrick Market, (Outgoing Member)Univ of Missouri-Columbia
Sam Ng, Metropolitan State Univ of Denver
Russ Schumacher, Colorado State Univ
Usercom Student Representative
Stefan Cecelski, Univ of Maryland
Policy Comittee Guest Participants
Lynn McMurdie, Univ of Washington
Michael Piasecki, City College of New York
| Peter Griffith, NASA
Scott Jacobs, NCEP
(For Rebecca Cosgrove and Michelle Mainelli)
Rich Signell, USGS
Agency Representatives Not Attending
Rebecca Cosgrove, NCEP
Bernard Grant, NSF
Michelle Mainelli, NCEP
Leroy Spayd, NOAA
| Vanda Grubisic, NCAR/EOL
Mohan Ramamurthy, UCP/Unidata
Tim Spangler, UCP/COMET
Steve Worley, NCAR/CISL
New committee members introduced themselves briefly, and commented on the morning new member orientation session.
Comments on the orientation were favorable.
Brief discussion of the staff status reports:
There is coordination between ACADIS and DataOne Data Management Plan to NSF for NSF proposals.
NSF is mentioning ACADIS data management plan as an example of how to construct a plan. Rosetta is part of this, and a demo is scheduled later during the meetings..
It was suggested that UPC make product-specific pages using Twitter and other media. There should be a tweet for every bug fix.
Action 1: Linda will work with Ginger and Tina to make a reservation for Center Green for the next summer UserComm workshop. The dates mentioned would be mid-July to late-July of 2015. Plan on 100 people again.
There was a question of how current enrollment numbers compare to previous years, and what the brekdown of registrants from universities/industry/government, etc.
Tom Whittaker attended a meeting where CF as an international standard especially focusing on standard names was discussed. It is important to continue getting the word out at meetings.
A question of making GEMPAK available via Github was raised. Will this happen?
Another question related to the idea of binaries being made. This has not been done for any platforms.
The reason binaries went away was the number of problems with library mismatches, for which the solution was to have users build from source.
OpenMotif was the biggest problem. Even now have to have GEMPAK users build OpenMotif to get GEMPAK to work.
Standard repositories have packages for NetCDF, etc.
Many of these packages are developed/supported by third parties
There are some community members who do this already.
Is a vm distribution an option?
Usercom need to ascertain whether there is a real need for this in the community
Is it worth trying to recruit users to contribute builds?
Gempak 6.7 is out...
The question of why the decision was made made to keep supporting GEMPAK
Scott Jacobs responded by saying that NCEP has several other projects that require GEMPAK functionality, at least in the batch mode. GUI programs will be kept going for a while, but NAWIPS programs will be replaced by AWIPS II GUIs
Another question of maintenance of GRIB tables, to which Scott replied that table maintenance is problemantic, and they are working on possible solutions
By request, the IDVsteering committee is spinning up again
Usercom should be the entity that sends the message out to invite participation in steering committee.
Steering committee members provide suggestions, server as alpha/beta tester for suggested features, conference calls/webex showing discussing functionality.
Action 2: Call for the IDV Steering Committee to go out from Usercom for participants, starting with list from IDVgroup (Julian Chastang)
Steve Lazarus is planning on setting up a regional workshop at Florida Institute of Technology. He would like to see UPC create a YouTube video of how to set up the IDD.
Tom Yoksas said the Africa initiative he's been collaborating on with NCAR will be wrapping up this fall, ending with technology transfer
John Caron is going to a google conference "geo for good", mostly to network
Tom Whittaker said that he's involved in a project with Eumetsat to install McV at Eumetcast stations
McV and IDV developers continue to have monthly telecons and share resources. Site visit is planned for this fall.
NetCDF windows: a lot of folks don't seem to be aware of work. Cygwin supports NetCDF
Many folks don't seem to be aware of the work done Cygwin supporting NetCDF
Unidata is providing binaries for windows -- 12 separate builds with all permutations
The question of how to publicize
There is something called DirectWeb Remoting,
Usecase: ncview in the browser would be awesome. Some clients are a pain to build/configure
Question from Rich Signell-Jira is great, but now I don't know what to do when I have something that might be a bug. Am I supposed to post it to Jira as a bug report or enhancement? People in general don't know where to direct questions.
Comment from Tom Whittaker-There should only be one place to look to track/search for issues
Kevin Tyle indicated- It isn't clear when/if someone from Unidata will answer mailing list questions
Russ Rew asked if we should be posting to the mailing list when issues are moved to Jira?
At the last UserComm meeting there was discussion about grib naming. John Caron said there had been discussions between TDS/IDV groups taking place. Rich Signell inquired about progress on the Opuls project. The funding has been problematic, according to Caron.
Dennis Heimbigner indicted that the UPC canan keep committee up to date on spec revisions
A question about when/why the upgrade to Java 7?
Eathan Davis siad the end of life for Java 1.6 is "soon"
Lots of good stuff in Java 7 file support.
Both NetCDF Java and TDS will update to Java 7, both to be called version 5
According to Sean Arms, there have been problems with Java3D and Java 7
John Caron indicated that they are cautiously optimistic about community support for Java3D
Scott Jacobs made a presentation for Becky Cosgrove (slides).
Move to new NCEP building is complete
RAP has been implemented to replace RUC, with corresponding CONDUIT changes
SREF changes are reflected in CONDUIT
Waiting for hardware refresh to increase LDM queue from 6GB to 20GB, at which
point additional model output can be added
Began testing live feed 10 September 2012 with Unidata (Tom Y)
Testing planned to be complete in mid-October, when they will transition to new servers
Question: RAP 13km or 32km first? Need both? (NOAAPort has 13km but only surface fields, not 3D)
Action 3: Recommend 13km RAP be added, not 32km as previously planned
Question: NAEFS raw or bias-corrected? Rich Signell says: raw grids, SC agrees.
Action 4: Recommend NAEFS raw grids be added
For Hurricane models, there are multiple resolutions available
Action 5: Russ Schumachermaker recommended asking colleagues which resolution do they want/can handle it?
Addition of Hurricane data ON HOLD until we get more information from the community
Note that Pull is available for all of this data, but not automated like IDD delivery
Question: GFS and Global RTOFS: What does the community need, more specifically?
No model upgrades planned until fall 2013 after new supercomputers are in place
Existing products from existing models *can* be added to CONDUIT
HRRR: Becky asks Usercomm to send official request to her boss (Ben Kyger) requesting access
Action 6: Draft note to Ben Kyger
requesting access to HRRR.
Users Workshop Report
Steven Lazarus provided a brief update on progress on BAMS article, to be submitted November 2012 if possible.
Future: When collecting registration information, ask how they heard about the workshop
Action 7 : UPC to create followup survey approx 6 months after workshop:
how have your data usage habits changed?
use of USB stick?
Nomination solicitation for 2013 award to go out "soon"
Please use the nomination form
Blue Sky Session
Tom Whittaker described this session and why it is needed. The Usercomm should discuss items/issues of importance to them as they plan for the future for their classes/labs. Unidata has global footprint. UPC should look
seriously at "cloud" type computing, allowing use of
minimally powerful devices.
Jennifer Collins said that there are not many Unidata users at her
institution. Regional workshops would be good for getting
other disciplines involved.
Mike Baldwin made the point that students are running Mac's not RedHat. They use their own PC's from home.
Marty Baxter is in favor of tablet-style computing, need new interfaces for IDV
that use touch gestures
(Jeff/Sean and Mike Schmidt are working on a Kiosk mode coming soon)
Marty said he would be willing to serve on the IDV steering committee!
Steve Lazarus said there's an issue with using IDV in the classroom -- could be
less clunky, a simplified version of the UI
Stefan Cecelski - would like to have an option just to look at
the data -- quick action when you don't need to modify the
Anne Case Hanks- Likes the idea of a lite
version of IDV on mobile device. Could be images, but want
to be able to composite data sets, display on globe, etc.
Sam Ng would like a Mobile way to get data to students. Switching to
IDV as a start.
Rich Signell said it would be nice to have IDV functionality
broken out as components that could be reassembled into
different workflows. Python might be a good glue for the
logic represented by IDV workflows. (Mike Baldwin added, when you work in a
web services environment, the choice of a development
environment becomes less important)
Kevin Tyle - Emphasis on collaboration in IDV --
Russ Schumacher - Provide "Nate Silver style" probabilistic predictions built
from ensemble weather
Tom Whittaker - How do we get more value from our
data via novel visualizaitions
Marty Baxter - enhance Unidata
outreach efforts to communicate what is already available
within atmos community and beyond. TED talk style
demonstrations of what you can do and how to do it. Unidata
can help with curriculum development, merging computer
science concepts with the meteorological work. Tools to make
it easier to evaluate forecasts?
Bart Geerts said IDV good for
querying 4-D structure, I like to use it for vertical
structure. These transects are not easy to combine, overlay.
SkewTs don't provide enough detail, would like to see that
improved. Would also like to see notes within the interface
Kevin Tyle said that we need more structure to the contact effort?
Steven Lazarus: We have some history from previous iterations
Tom Whittaker: : Tool that Jen put together is useful; is it too hard to keep up to date?
Kevin Tyle said It is important to think about what we want to ask folks
- how has IT support landscape changed?
Tom Whittaker said that When Gary Lackmann was chair, each contact was a mini survey. Is that too much?
Anne Case Hanks - Usercom can decide what to focus on prior to each contact episode
Linda Miller and Doug Dirks
Community Services - Unidata
University Corporation for Atmospheric Research
P.O. Box 3000
Boulder, CO 80307-3000
303 497-8646 fax: 303 497-8690