VegX/Vegbank discussion
---------------------------------------
Participants: Peet, Lee, Jones, Schildhauer, Regetz
Agenda
-----------
1. Status of VegX
2. Status of BIEN work
3. Status of VegBank
4. Work on Vegbank import/export of VegX
Discussion
----------------
1. VegX -- moderately stable; overall structure is fine; implemented at NZ Landcare Rsrch; fine tuning now
-- reasonable point for implementation (1.4.1 schema is most recent?)
-- Bob to send manuscript of VegX
-- still may be changes for supporting repeated surveys
2. BIEN: tried to implement a relation framework that imports/exports VegX
-- hoped this would be a couple of years of effort, but student moved on after 2 quarters
-- some pains during creation of OR mapping
-- scarcity of VegX documents, even those from LandCare were invalid
-- LandCare NZ implementation is a subset of the full VegX
-- Need to build a valid test corpus of VegX documents
-- Need clarification on things like geospatial fields in the VegX schema
-- iPlant will contribute $50K for a year to advance the BIEN database (maybe 2 years)
-- Susan and Nick may be also drafting another proposal for iPlant work
-- also working with iPlant on world-clim style data resource
-- our concensus is that focusing implementation on extending VegBank would make sense
-- SVN URL for BIEN work https://code.nceas.ucsb.edu/code/projects/bien/
3. Vegbank:
-- New server is up and running on VM at NCEAS
-- Migrated to the new server virtual machine: vegbankvm.nceas.ucsb.edu = vegbank.org
-- a development virtual machine also exists: vegbankdev.nceas.ucsb.edu/
-- People that I don't know are loading data here and there.
-- Loading is still a bit awkward, some of the post-processing (generating keywords and denormalizations) are not working smoothly.
-- Loading data is still too complex - people can't just load simple data in a simple way on the website.
-- Export of .csv files can download over 10K plots at one time (~1K before)
-- refactoring necessary for downloading? to eliminate memory-bound limitation
-- rebuilding website by ant still a bit problematic - permissions issue - need to restructure .war file
-- Lee funded to maintain VegBank at approx. 10% FTE ongoing
-- and to load data in at 1.5 months
4. VegX import/export
-- VegX and VegBank schemas are compatible at a high-level, need a detailed comparison
-- talk to Brad -- he may have done a more detailed comparison
-- VegX may have explicit fields for items that are in user-defined fields in VegBank
-- Two implementation paths:
-- custom loader
-- external XML-to-XML conversion
-- Would a simple import/export be sufficient to BIENs needs?
-- They probably don't want to see classifications
-- Would want additional geospatial features (which was motivation for PostGIS/GeoDjango architecture)
-- Simple occurrence data may be more simply represented than how co-occurrences are in VegBank
-- Could also use new UIs in VegBank to deal with individual occurrence searches
-- Could some of the restructuring result in features that aren't wanted in VegBank per se
-- Would this be deployed within the current VegBank, or as a new BIEN3 extension and deployment?
-- Branding itself could be an issue....
-- Governance of the resource would need to be worked out too...
-- current governance of VegBank is via ESA veg panel
-- what is the governance of BIEN? Missouri? iPlant?
Action items
------------------
-- TODO: Michael Lee: find definitive most recent version of VegX (on TDWG wiki?)
-- TODO: Michael Lee: will start doing model comparisons and starting on VegX/VegBank XML conversions
-- TODO: Bob and Mark to discuss BIEN desire to extend VegBank to support VegX and BIEN needs
-- involves statement of work for BIEN activities
-- Next meeting: Conference call March 15, 2011 @ 12:00 pm pacific time (3pm eastern)