| Line 5 was replaced by lines 5-8 |
| - * Actor repository |
| + |
| + ! Actor repository |
| + |
| + * Actor repository (Chad organizes, Laura and Shawn participate) |
| Line 7 was replaced by lines 10-87 |
| - *** Should there be a separate package manager for installing 'actor packages' |
| + *** Remote actor names in italics? or some other cue? |
| + *** Probably should be a separate package manager for installing 'actor packages' |
| + *** differentiating remote and local should make sense for both actors and data |
| + * Remote actor display in Kepler library (David) |
| + ** Get documentation to display |
| + |
| + ! UI redesign |
| + |
| + * Icons |
| + ** Need policy guide for how to extend the icon set in a visually consistent way (Laura) |
| + |
| + * Dialogs |
| + ** Workflow, Actor, Director dialogs (highest priority; developer TBD) |
| + |
| + * Semantic annotations (Shawn) |
| + ** take out ontologies that are there now, add better ones |
| + ** make a GUI for adding and removing ontologies from Kepler |
| + ** keep the annotation dialogs in Kepler if possible |
| + ** need semantic annotation of data, different than for actors |
| + |
| + * Query Builder |
| + ** "Display" label -- discuss and decide how to change (Jing) |
| + |
| + * Actor configuration dialog |
| + ** Need more space for identifier (LSIDs are long) |
| + ** Double-click == open actor == look inside |
| + ** Right-click then Configure to configure |
| + *** Consider possibility of the IDE motif with a selectable property palette |
| + ** Parameters -- should be scrollable list |
| + ** Question about relative priority of UI tasks versus other tasks |
| + *** Dialogs are not a show-stopper for Kepler, but still are very important from a user-acceptance perspective |
| + |
| + !!!EcoGrid Breakout |
| + |
| + ! Action items for release (__January 15, 2007 target__) |
| + * Fix memory problems on SRB get interface (Lucas) |
| + * Move GET + PUT method to use HTTP binding to avoid memory problems (Jing & Lucas) |
| + |
| + ! Action items for release (__February 15, 2007 target__) |
| + * Finish LSID support for Metacat (fix bugs) (Jing) |
| + * Add LSID support to SRB, use LSIDs for SRB identifiers (Lucas) |
| + * Separate Query/Get into separate WSDL documents (David) |
| + * Authentication: turn on SSL for ecogrid auth methods (Jing & Lucas) |
| + * Refactor for naming consistency (server and client sides) (Jing and Lucas and Dave) |
| + ** Update namespace to 1.0.0 (currently uses beta1) |
| + * Documentation (David) |
| + ** server side |
| + ** client side |
| + *** build a client package with full documentation, bundle all client stub jars |
| + * Rename to: EarthGrid |
| + * Add portal to SEEK website, rename throughout to EarthGrid (David) |
| + |
| + ! Action items for version after release |
| + * Add delete method to its own WSDL (logical delete) |
| + * Add replication and 3rd party transfer methods (postpone) |
| + * Authentication |
| + ** research how to pass GSI certs over HTTP (maybe in headers) (postpone) |
| + * DiGIR performance issues (waiting on new hire, or Dave will finish) |
| + ** Harvesting to central server is now underway |
| + ** Need to implement EcoGrid interfaces to the central cache |
| + ** support Wasabi (query + get interfaces, RDF to XML translation) |
| + ** support Tapir (but maybe too complicated, so postpone) |
| + * monitoring framework |
| + ** monitoring script that generates simple web page with status update (Jing) |
| + ** more complete user-oriented web report interface for all EcoGrid providers (postpone) |
| + *** see the examples for DiGIR at [http://bigdig.ecoforge.net] |
| + |
| + !!! Kepler: Distributed Execution |
| + |
| + !Proposal discussion |
| + * Use a subclass of RunCompositeActor |
| + * central broker/controller handles communication with different nodes |
| + * use annotation to flag stateful versus stateless actors |
| + * deal with passing data by reference |
| + ** probably modify the Token base class to support passing by reference, or maybe the ObjectToken class to limit the impact to the type lattice |
| + ** actors would have to create tokens that are of a given type with a reference instead of the data |
| + |
| + * security: need it but could be layered on in a second step, but would mean that the system would have to be run in a trusted environment |
