Notes on Kepler

pennington.pdf: presentation on SEEK and Kepler

My take on this presentation is that there are Bio folks who really wish they had a system that lets them have technology-enable science rather than what they have now which is technologic systems for scientists. They want collaboration. They want access to heterogenous data. They want it all under one application and umbrella. The presentation describes a variety of systems. But, it seems to only devote one or two slides to each, so the slides are somewhat low-bandwidth. But, the slides provide a good place to find jumping-off points. They mention a few different systems. One of my next steps will be seeing what each of these systems can do. -- PatrickStein - 21 Mar 2005

More information on the Kepler project is available at http://kepler.ecoinformatics.org/

It would be particularly worthwhile to look at the publications and presentations. Follow links on the above page.

A short synopsis from the main Web page is:

Scientists in a variety of disciplines (e.g., biology, ecology, astronomy) need access to scientific data and flexible means for executing complex analyses on those data. Such analyses can be captured as 'scientific workflows' in which the flow of data from one analytical step to another is captured in a formal workflow language. The Kepler project's overall goal is to produce an open-source scientific workflow system that allows scientists to design scientific workflows and execute them efficiently using emerging Grid-based approaches to distributed computation.

Kepler is currently based on the Ptolemy II system for heterogeneous, concurrent modeling and design. Ptolemy II was developed by the members of the Ptolemy project at UC Berkeley. Although not originally intended for scientific workflows, it provides a mature platform for building and executing workflows, and supports multiple models of computation.

Information on Ptolmey II: http://ptolemy.eecs.berkeley.edu/ptolemyII/

-- HarveyRhody - 23 Mar 2005

It's really unclear how much Kepler actually does. It looks to me like Kepler might just be some fringe on the edges of the PtolemyProject.

When you download and install Kepler, it installs an executable that appears to just be a simple web browser. It shows a page of links to sample Workflows and such. When you click on them, it brings up the source code for them. If you burrow down into the directories and try to open one of its Workspace files, it spews a ton of XML errors.

If you open up some of the HTML files that came with Kepler, they instruct you to open up Virgil from the Ptolemy directory. But, the Ptolemy directory that Kepler installs for itself doesn't have any of that stuff in it. Hmmm....

I'm thinking that Kepler may just be a collection of "Actors" on top of Ptolemy. I think they may have added integration with Web Services (WsdL) and such for their aim of Grid Computing.

-- PatrickStein - 24 Mar 2005

And, for a slightly more detailed analysis of what Kepler is/isn't, see WhatCanKeplerDo.

-- PatrickStein - 05 May 2005