| Line 2 was replaced by line 2 |
| - __Participants__: Jones, Gallagher, Altintas, Graham, Schildhauer, Potter |
| + __Participants__: Jones, Gallagher, Altintas, Graham, Schildhauer, Cornillon, Potter |
| At line 12 added 2 lines. |
| + * Ilkay: should have someone from ROADNet -- Vernon? -- for the workshop |
| + * __TODO__: Mark and Matt to upload edited documents to wiki |
| Line 29 was replaced by lines 31-47 |
| - ** |
| + ** data models seem compatible -- looked at the Ptolemy model and there's a reasonable mapping between the types |
| + *** fixed point type probably maps to a float |
| + ** sequence is an ongoing problem |
| + ** have to map grids, but will lose implied semantic information |
| + *** grid is collection of arrays, main array with other arrays that specify, e.g., coordinates in lat and lon |
| + *** need to try to maintain the implied relationship between these 3 arrays to ease processing |
| + ** hasn't figured out how to implement an actor |
| + *** __TODO__: Nathan work with Ilkay on actor |
| + ** also need to map Common Data Model (CDM) and EML |
| + *** [http://knb.ecoinformatics.org/software/eml/] |
| + *** unifies HDF5, NetCDF4, OPeNDAP |
| + *** much of the CDM has to deal with quirks of HDF5 focused on HPC access to data |
| + *** __TODO__: Nathan develop a more formal mapping between the major data models |
| + ** Peter: suggests Nathan, Peter, James propose a schedule for data model mapping description and implementation in Kepler |
| + |
| + ! Next meeting: May 2, 2007 10:00 am pacific time |
| + |
