In this topic that may someday transmogrify into a blog post.. I explain how to run a very simple workflow that a) retrieves and html page, b) extracts the image links inside of it, c) finds the url for a comic image, d) returns a Dilbert comic for your viewing pleasure.
1) Download and install the Taverna workflow client
2) Go to the file menu on the top and select "open workflow"
3) Browse to, select, and open FetchDailyDilbertComic.xml (included in the download)
You should now see the attached image of the workflow
4) Now go to the file menu and select "run workflow"
5) Click on the result list and your dilbert should be displayed.
The workflow engine (Taverna) read the XML file and understood it as a set of instructions for gathering the web page, parsing it, and rendering the desired image.
This technology isn't just for Dilbert. Essentially any bioinformatics analysis could be represented as a workflow. This means that I can write a workflow, send you the xml and you can run it, inspect it, and manipulate it to your hearts content. This is a big step forward from developers either a) producing perl scripts never meant to work anyone else or b) having to create full applications to share their work.
Behold, the new era of bioinfomatics is upon us.
I'd love to hear your experiences with Taverna or other workflow engines. Even better, post the xml file for your workflow so I can try it out!