[provenance-challenge] PC4: moving forward with a common scenario

[Paul Groth]

Hi All,

Thank you all for participating in the scenario poll, we got a 
tremendous response (Poll results are here: 
http://www.doodle.com/s6d27hxffutztrka
). Again, thanks to all of those who submitted a scenario. We are going 
to have a discussion at the scoping workshop next thursday at IPAW 
around the scenario we'd like to use. However, in order to best use our 
time, it would be great if we could hone in on a scenario now. All of 
the proposed scenarios have interesting points that we could include in 
the final pc4 scenario. So what I thought I'd do is make a suggestion 
here about a possible combination scenario and we can have an email 
discussion about that.

----------------
Proposed Combination

The Provenance of Published Crystal Images (ProvenanceOfCrystalImages) 
got by far the largest number of votes so I suggest using that as the 
framework to organize the scenario.

ProvenanceOfCrystalImages combines both authomated steps and those 
performed by humans, it also emphasizes long term preservation. To me 
this incorporates the notions behind UnderstandingScientistIntent as 
well in particular why a human performed a particular part of the 
experiment in a certain way.

I suggest we combine the output with a wiki like pdbwiki.org and be able 
to track the edits to a wiki page as suggested by the 
WikipediaRevisionHistory scenario. Another possibility would be to 
connect it directly to wikipedia but I don't see how that can happen 
(suggestions).

It seems that the suggestions from ModelingObjectsMultipleGranularities 
can also be included by splitting up the crystallography process across 
two distributed groups. This also incorporates the PaperSubmission 
scenario, either the outcome of the crysatallography should be published 
as a paper or the wiki page should be treated as the publication. 
However, I think actually tracing the outcome of a pdf would be more 
compelling.

We can incorporate the ComposingMultipleWorkflows scenario by ensuring 
that different parts of the automated portion of crystallography process 
are conducted by different workflow systems.

Finally, TrustAndPrivacyManagementInCloudPlatforms can be incorporated 
by executing parts of the process in a cloud or an off site service. 
This would incorporate the trust issue in that the scientists would like 
to have confidence that these external services are accurate.

-----------------
I've tried to incorporate the top rated scenarios here.  Personally, I 
would like to ensure that what do has parts that are of interest to 
people outside e-Science. Also, we need to ensure that by combining 
everything this is doable.

Thoughts? Comments? Questions?

Thanks!
Paul