[Go-essp-tech] Minutes: 1/21/10 ESG Model Metadata Demonstration

[Sylvia Murphy]

On 1/21/10 the Curator team conducted its latest demonstration of the developmental state of the model metadata display within ESG-CET gateway. Thanks to everyone who participated.

Present:

Julien Chastang
Luca Cinquini
Gerry Devine
Hans Ranthun
Stephan Kinderman
Balic Neboisa
Charlotte Pascoe
Allyn Treshansky
Sergey Nikonov
Cecelia Deluca
Don Middleton
Frank Toussaint
Heinrich (from Hamburg)


The following items were covered:

* An update on the model metadata developmental timeline
* The always-on component navigation system
* The display of inputs
* The display of experiment information including requirements
* The display of conformances
* The links between datasets and simulation metadata pages
* A review of basic and technical properties
* Two options for the advanced search interface (presented by Luca)


PREVIOUS ACTION ITEMS COMPLETED:

1) Remove version from the component tree.
2) Add discipline to the component tree.  
3) Split technical properties into two accordions, one for general information and one for configuration information specific to the run. 
4) Find out how METAFOR is planning to handle output variables.
5) Get a sample dataset from GFDL so we can demonstrate linking metadata to datasets within the CMIP5 example. 
6) Find out if and how the current go-essp-tech discussion on versioning affects us (e.g. will we need to version the model metadata).
7) Find out if the current go-essp-tech discussion on replication affect us.  
8) Remove the hasExperiment from the example's subcomponents to more accurately reflect what we anticipate in the questionnaire.
9) Verify with Luca which OWL files are loaded into the main portal and make sure there are none in there that should not be.
10) Talk with Luca and others about the search text on the main gateway interface.  Is there a way we can suggest that the advanced search also includes the ability to browse metadata about the models and experiments?
11) Add a description of the date format to the definitions of X_StartDate & X_EndDate.
12) Change the simulation search on the search interface so that simulations are only associated with full models - sub-components of simulations won't be returned as search results (Luca).

CARRY OVERS:

1) Add definitions to the targets on the search interface somehow.
2) Connect gridspec files to grid tree.

NEW ACTION ITEMS:

1) Open by default at least one accordion per tab.
2) Make the variable name under inputs stand out more (e.g. bold it, put the other information below it etc).
3) Determine which facets we should use to search on models (Sylvia will send out an email). 
Discussions:

1) The primary discussion of the day revolved around the two search interfaces that Luca demonstrated.  The first interface was the current search interface which is based upon MSpace.  The second interface was a prototype based upon more typical e-commerce faceted searches (e.g. Amazon.com <http://Amazon.com> and HomeDepot.com <http://HomeDepot.com>). Sergey liked the current interface better.  He thought it was cleaner and more professional looking.   Gerry, Charlotte, Allyn, and Hans liked the second interface better.  They thought it was more intuitive.  Cecelia commented that there is a lot of white space in the middle of the page in the second interface and wanted to know what might go there.  In commercial sites, there are usually pictures of products.  Luca indicated it could be possible to put instructions there or to list datasets based on a non-narrow search.  Luca commented that he liked the second interface better because it would be easier to maintain and modify.  He did point out that the adaptation of this type of commercial search has its drawbacks for ESG, namely the number of variables (~400) that would potentially have to be displayed.   A "read more ..." strategy was suggested to handle this.  Someone suggested that in the second interface, the term "Search Categories" should be replaced by "Search Restraints" or something to that effect.  Luca indicated that the second interface needed to be shown to the ESG community before any decision was made on which interface to choose.  

2) Cecelia is concerned that we need a way to differentiate values that are only applicable for simulations. These values will inherently be displayed on the simulation pages and not the model component pages but they may not be readily identifiable as simulation specific since they will be mixed in with values that will be displayed upon both types of pages.  Cecelia asked the METAFORIANS on the call whether the questionnaire output would differentiate attributes that are simulation specific.  Charlotte indicated she would go back and review the questionnaire output to determine if this was the case.  Cecelia and Sylvia are going to talk about possible strategies off-line.

3) Don updated the group on the ESG release schedule.  The release that was scheduled for next Friday (1/29) has been moved to the middle of February (2/15).  This change will effect Curator's developmental timeline and demonstration schedule.  Since we will now have at least two more weeks of development time before the next ESG release, Curator will try and get a preliminary XML upload capability into the system before the release.  Another demonstration may be required between now and then.  

***********************************
Sylvia Murphy
sylvia.murphy at noaa.gov
303-497-7753



-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/go-essp-tech/attachments/20100127/382d1f58/attachment-0001.html