I hesitate to call these "minutes", but here are my "notes" from Monday's meeting. Please
let me know if you feel I missed anything or misrepresented anything.
Thanks!
Becky
Tom Achtor - Goal of this meeting is to talk about the Mc-V bridge and the
long term vision of the project - the best path to achieve the best system
for scientists, the MUG, and students.
DaveP - Mc-V bridge to Mc-X
New version - write a Mc-X server that receives McIDAS commands and returns
text and image
Old version - read existing Mc-X windows and import them into the IDV
Not throwing away Gail's previous work - still reading pixel array and converting
it to a data object. Not memory mapped files. Still need
SFU, but when using a web interface, there will no longer be a need for an
X server or tcl/tk. TomA called it "elegant and simple".
DeeW - the opportunities for outreach are tremendous. Online demos
of McIDAS, web version of McLite, easy packaging for outside users (SATEPS,
KathyS, etc.)
DaveP - The new bridge would be able to do everything you can currently do
and more, but with fewer requirements.
TomW - The first step is to write the protocol to initially talk to a web browser,
but also make it expandable.
TomW - With this new long term version of the Mc-V project, we've revised our
goal from being to recapitulate Mc-X to now supporting the development requested
by the PIs of the various projects.
Becky - Any objections to this method of the bridge instead of the old? There
were none.
BobK - liked the way the project is heading. The future is in the distributed
commuting environment. He sees Mc-V as an island with many bridges
- bridges to data and bridges to visualization. There's the bridge
to the past, which is the Mc-X functionality and important for current Mc-X
users. There's the bridge to SQL databases. There's a bridge
to the hyperspectral data via Hydra. There's a bridge to the future in
cluster computing. He'd like to see the development environment bring
new applications to the system, to do things like running applications on a
cluster, and to see the project really using the network interfaces as the
glue.
TomW - Hydra is the mainstream for development.
BobK - I see Mc-V as the system that can talk to the different pieces
BobK - would like to SEE the path that this is taking. Big proponent
for targeted demos.
TomR - The path is staggered. The long-term target changes, so
the path changes.
BobK - wants to see GOES-R work in this environment
TomA - With this plan, it should be easy to convince Mc-X users to eventually
use Mc-V. With Unidata and professors like Ackerman teaching the IDV,
it should be easy to convince students to use Mc-V. Scientists are
harder, and you have to show that this package is better than what they
currently have.
BobK - There are other packages that can use hyperspectral data, but visualizing hyperspectral
data is now difficult.