Version 4 (modified by pmiller, 15 years ago) (diff)


NDG2 Plymouth 11-12 July 2006 - Peter's notes

Board Meeting

  • Finance - overall exactly on track with budget.
  • Highlight report - mostly on schedule at post-Alpha review. WG group were instrumental in Alpha system integration testing.
  • Next three months: need detailed definition of Beta products. Start playing with Vocab server? Discovery back end is broken, Matt Prichard to work on this. Population!
  • CSML future: Andrew presentation (CSML API and v2). CSML will be used by MetOffice? and other projects, e.g. FP6 MOTIIVE. Read Exeter communique on OGC etc. (metoffice/infomatics?) Development of 'feature types'. GML stable version 3.2. 'Operational affordance' - can only extend one feature type but can inherit properties from multiple...? CSML will deliver core set of base types, derived feature types, basic services (api), schemas, tooling (parser/scanner).
  • So what should DPs do towards CSML? NOCS will be adapting scanner to deliver their HDF data rather than CF-netCDF. BODC is using QXF-read method from Matlab/Java?. Implement CSML v1 for now, then modify for v2. Swath very soon. Parser lets us play with CSML records prior to DX working - use that to help us write our scanner. If going from data->CSML, must write scanner in Python as this encodes the objects and attributes. If going from metadata->CSML, just need to plug in SQL extraction to Python scanner.
  • MOLES future: MOLES is for generating various discovery formats, and will not be driven by external standards e.g. SensorML. x-links to include other docs (XHTML, SensorML, ...) verbatim within MOLES records, e.g for BODC calibration info. Could simplify/thin MOLES... Bryan proposes generating CSML data objects first, then flesh out MOLES from those (then DIF).

All-Hands Meeting

Peter actions

  • Re-read PML sample CSML.