Last Modified: August 27, 2003 4:54 PM
Interoperability with existing metadata descriptions. Data providers have to serve too many different ones.
Relationship between THREDDS standard quantities and existing ones (e.g., GCMD, others?)
Communication mechanisms for data and metadata: should they be the same?
Multiple "redundant" servers for certain products:
- Are they needed?
- Are products on different servers the same?
Multiple data locations: Can we merge? Product Identity?
Real-time vs archived data: how to achieve seamless access?
Databases:
- Do we need a central one for all of THREDDS?
- How to deal with existing data provider sites using databases?
Metadata additions:
- Time-to-live entry?
- Lineage entry for datasets that have been augmented, modified, transformed?
Subsetting at metadata level or at data access layer?
Metadata standards crosswalks
Real-time vs archived
URL permanence -- server permanance
THREDDS role vis a vis data models and protocols
Database-based data provider sites
Granularity from client viewpoint (wants to use words of discipline to ask for data) is quite different from data provider viewpoint (browsing and drill down of hierarchies is fine)
Data provider model vs data user model. Need discovery beyond hierarchical browsing.How general a solution should THREDDS target?
Embed data (and tool) links in documents (e.g., educational modules and scientific papers)
Should THREDDS be involved in (the ugly, mess problem everyone avoids) standardizing metadata?
If so, can we start with georeferencing and standard quantities?Security
GDS vs DODS/THREDDS aggregation
HDF-EOS datasets difficult to visualize with usual DODS clients?
Education Extensions to FGDC/ISO TC19115
Is a central THREDDS DBMS needed?
NSDL Source Forge, CVS, WIKI usage
Issues relating to time delays staging data from "backend" archival mass store systems.