After discussoin at the July 23, 2001 meeting about the current wish list
we decided that we could begin ranking these projects by first grouping them
into sensible groups and then ranking them in each group. Then we ranked
the relative importance of the top level of each group to get a first look
at how things will fall out. The groups are listed in their order and
the topics are listed under each group. Note this is not a true order
as it is only ranking the top in each group...but it gives us a sort of bubble
sort ranking to start with.
Also we tried to take into acount what would be done on the road to another
goal here. For example it was demed that unless you can extract data
in science usable form from the archive, extracting information automatically
from that data would be worthless.
For the next meeting we should finish the sort, moving things up and down
from this order to get a final relative order for these projects.
Top level - finding stuff on the web
- Search portal with info maps
- “my” portal style customization
- Agent like searches/data gatherers
- Ability to flow results to an unrelated query
- Single source access for data and catalogs
- Cross Mission queries (using several missions in one big query)
- Visualization without downloading a file
- Instrument independent quantities in database
- Seek out and/or develop more extensive metadata such as nonlinear WCS
- External catalog/data ingest and distribution standards
- Data in science usable form (improved reductions)
- Data extraction services
Some general comments about how these
are implemented were raised
Lastly it was mentioned that there is a clear destinction between data and
meta data in that data relates to the science while the meta data relates
to data descriptions. However they all fall under the umbrella of information
and that is really what we are talking about distributing here.
- Should not limit ourselves to working in a web browser
- Web has more dynamic support than pure applications
- possibility of a web plugin with links to external applications (e.g.
VTT or StarView)
- Plugin requires sysadmin support on the client side