We changed the master project to once again run the random observation operator in python 3, using the gridded statevector and the ObsPack GlobalViewplus2 on capegrim. Next to do is to make this project also run without the specific parts but instead with baseclasses, and to program something that allows different templates on different machines.

No commit message

rename *PlatForm to *Platform

renamed class names Ct- to CO2-
changed class name Observation to Observations 
small cleanup changes to template.py

write_members_to_file: is called with _lag_ instead of lag +1, as inside there was a "lag-1" call... that one removed from inside as well

write_sample_info changed to take the whole path as a parameter: so now we explicitly see the path in pipeline, plus we don't need to return it back.
same for write_obs_to_file

managed imports in obs files

removed a need of using samples.DaCycle=DaCycle (now it is only passed in the "initialize" function)

performed actions discussed in email #2
removed unused __str__ functions
a bit of ordering of the pipeline.py

third option, the pyCDF interface for reading and writing. Note that it does not read HDF filessvn move da/ct/io.py da/tools/io_cdf.py!