[provenance-challenge] Re: Provenance challenge

[Yogesh Simmhan]

Hi Sean,

I think the brief answer is "whichever" :) You have access to the code and you are welcome to instrument it to collect provenance or even disregard it and roll your own while sticking to the "activity" defintions that make up the workflow. The challenge does not prescribe how you run the workflow (or even that you run the workflow at all, opting instead to simulate it) to collect provenance. What is key is the ability of your provenance system to run the challenge queries and export/import provenance in the Open Provenance Model for reuse. The intention is to test the model's ability to act as a means for sharing and using provenance from heterogeneous provenance systems.

Best,
--Yogesh

________________________________________________________
Yogesh Simmhan
Post Doc Researcher/eScience Group/Microsoft Research
EMail: yoges at microsoft.com   WWW: research.microsoft.com/~yoges
________________________________________
From: Sean Riddle [swriddle at gmail.com]
Sent: Monday, March 23, 2009 10:50 AM
To: Yogesh Simmhan
Subject: Provenance challenge

Dr. Simmhan:

I'm currently working on the 3rd provenance challenge out of UC Davis,
and I had a few questions about it. First, we've been discussing back
and forth here the purpose of the code given on the wiki in C# and
Java to implement the logical steps of the challenge. One side thinks
that the code is supposed to be unchanged, treated as a black box,
while provenance and workflow system concerns are wrapped around it,
and the other side thinks that it should be changed extensively, so as
to maintain the provenance of certain components like the internal
database. I initially thought it was the former, but the last few
questions from the suggested queries page on the wiki (http://twiki.ipaw.info/bin/view/Challenge/ProvenanceQuestionsPc3
) have me thinking that maybe it is the latter. Of course, the correct
answer could be 'whichever way we feel like doing it,' but I'm curious
as to your original intentions.

- Sean Riddle
Software developer
Genome Center, UC Davis