[Go-essp-tech] Minutes: 1/11/10 Metadata Pipeline Telecon

[Sylvia Murphy]

Hi Everyone,

GO-ESSP-TECH had a productive metadata pipeline path telecon
on 1/11/10.  Below is a summary of the outcomes of this telecon.

Note, since things have progressed between the call and these
minutes, I have put any updates in brackets.  

QUESTIONNAIRE PATH

1) Bryan and Luca still need to have an offline conversation
to discuss the pros and cons of OAI version ATOM as the means
of transmitting the questionnaire XML files to ESG.  ESG's
system is automatically set up to work with OAI while METAFOR's
system is automatically set up to work with ATOM.  (Target 11
January). [**NEEDS RESOLUTION**.  No decision has been made as far as I
know.  Bryan reported that he has written the code to provide
the files via ATOM.]

2) Bryan and Ruport provided numbers on how many attributes
were in the questionnaire, which ends up being some hundreds
depending on how you count.  This was sufficient for work estimates.

3) Sylvia will go to the questionnaire and select the Hadley Center
example and export it.  Bryan indicated that this is the most complete
example they have to date.  The development of the XML upload software
will begin with this example. (Target 13 January). [This was completed.
The questionnaire continues to evolve, however, and the format for the
Hadley instance is already out of date.  Rupert provided ESG with a
different, more syntactically correct example, which is still sparse.
See next item for work done on parsing the XML.]

4) The group acknowledged that it will be challenging to try
and write an XML upload package that includes all the questionnaire
attributes by 12 February, the cut off for ESG's Release Candidate 1.
METAFOR will provide increasingly complete XML examples leading up
to the Feb 1 scheduled release of the questionnaire.  We expect
further work on the XML upload task to continue past March.  
[Julien has started working on the XML upload.  To date, he can
parse the components out of Hadley Center sample XML file.  Next
this tree will need to be converted to OWL. A focus on properties
follows.  Rupert has made this task easier for ESG by
providing a RDF version of the scientific properties, which will be
imported directly into the ontology.]

5) METAFOR will provide an even more complete sample XML file by the
time the questionnaire is released (Target ~1 February). [The target for this
may now be during the METAFOR year 2 meeting on 9 February.]

6) The group agreed that the display of output variables in the ESG model
metadata display was not necessary for the March release.  Having the link
to the data collections will be sufficient.  This issue may be revisited
at a later date.   

7) The outstanding items for the ESG display are progressing on time and
so far Curator is receiving appropriate assistance from METAFOR.  Much
still depends on us getting better XML output from the questionnaire.

Items addressed include the display of initial conditions/boundary conditions,
conformances, and scientific properties.  Curator is targeting 12
February to have all of these items in the display.  [ESG has adjusted
its release schedule since this telco.  The release scheduled for the
end of January has now been moved to the middle of February with releases
coming every 2 weeks after that. Curator is pressing forward to get as
much capability into each release as possible.]


GRIDSPEC PATH

1) Balaji reported that a series of telecons have been occurring with
respect to the gridspec software and its status.  Balaji agreed to
provide go-essp-tech with a summary of the findings to date.**NEEDS
RESOLUTION**

2) Balaji and Karl agreed that they are the individuals responsible for
ensuring the gridspec software is functioning and that gridspec
harvesting software is in place.  

netCDF PATH

1) Luca indicated that in order to connect model output to the
simulation that produced it, the model names in the files and the
model names in the questionnaire need to follow the same syntax,
preferably DRS.  Bryan indicated that the model names in the
questionnaire are not a controlled vocabulary and that differences
between the questionnaire and the dataset may have to be handled
manually after the questionnaire output is produced.  An offline
conversation needs to occur to discuss the workflow for this task.  
**NEEDS RESOLUTION**

SEARCH PATH

1) The group agreed that appearance and functionality of the gateways search interface 
needs to be adjusted for CMIP5.  It was agreed to make this the topic of another telcon.  
Curator requested the by notified/invited to this telecon.  [Significant changes have occurred 
along this path since the telco. Two possible search options were demonstrated during the 
ESG Model Metadata Demonstration on 1/21/10.  The search interface is being significantly 
modified by ESG based upon the participates feedback.]
***********************************
Sylvia Murphy
sylvia.murphy at noaa.gov
303-497-7753