Remember the first thinkweek paper, which I co-authored, I mentioned some time ago? Well, we got feedback from Bill Gates and it was positive :-) Bill only comments on a very small percentage of the thinkweek papers he receives. That email was a happy work moment :-)

The subject of the paper? Well... some other time :-)

Cherie sent me this... Funny...

Just bought my 10-day pass Edge card for Whistler/Blackcomb. So, at least 10 days of skiing there. Can't wait to try my new skis, bindings, and boots. I am planning to do lots of skiing this year. Of course I am also going to the local mountains but Whistler/Blackcomb is going to be my regular destination. California and skiing with Chris is on the table too and will definitely use the opportunity to visit the Alps during one of my trips to Europe.

Ian told me that the snow is already falling in the Cascades. Soooooo excited for the new season :-)

My group is interacting with scientists from all disciplines. We are trying to help them in their research endeavors through the use/application of technology. It is therefore necessary for us to keep informed about all technologies available; it is only then we can offer informed advise and talk about the potential benefits. That's why I love my job... not only I get to interact with extremely clever people and think about the future of "Research in the Cloud" but I also get to experiment with new technologies (so that I know my stuff before I open my mouth :-)

Catharine van Ingen has been doing a lot of cool stuff with scientists. One of her many great engagements involves the Fluxnet. She is trying to help Environmental Scientists with their data management issues and also assist them in making sense of it. She was the one who put me in touch with Deb Agarwal, who is also working with the same scientists.

Deb wanted to play with PopFly as an investigation on how easy it can be for scientists to create a mashup of the Fluxnet towers over Virtual Earth using high-level abstractions, like the PopFly 'blocks'. It wasn't without problems but we have to remember that PopFly is only in beta. I think the result shows how trivial it can be to combine/process/analyze/visualize data, if we get the abstractions right. Very cool stuff.

I needed a use case as an excuse to teach myself how to build Silverlight 1.1 (alpha) applications. I took a copy of the data Deb was using for her PopFly mashup and started playing. My approach is nowhere near as easy for scientists since I had to write .NET code. However, one could convert my code to a PopFly block in order to get the smooth interface.

I wanted a Silverlight-based navigation control for Virtual Earth. I started from the one that comes as part of the samples on silverlight.net and I had something running. Then, through an internal distribution list, I found out about the excellent work that Peter Blois did in implementing a Silverlight-based interface to Virtual Earth. I contacted him and he pointed me to his blog entry, where he made the code available for everyone to download. Using his code as the basis, I added my own visualization code for the Fluxnet data based on suggestions by Deb.

You have to know the names of the towers in order to search for one. The transition animation is just beautiful (thank you Peter). Unfortunately, as of few minutes ago when I last checked, the Virtual Earth search functionality for locations (the top text box) is not working (it might be a temporary Virtual Earth problem... ah... the beauty of mashups on the Web). You can use the Control Panel from the left to change the way the data is visualized.

Enjoy!

Feedback/problems/ideas more than welcome.

Hugo Hiden, NEReSC's technical director, has started a blog, called "eScience Ramblings", where he records his experiences with using Microsoft's technologies while building solutions for eScience. I am really looking forward to reading more. So far it's been great. I am going to be sending answers/suggestions to Hugo whenever I can but I am hoping other Microsoft and techie folks will also help him, whenever he encounters a problem or there are alternative/better solutions.

Hugo, have you looked at the Entity Framework as an alternative to Hibernate? The integration with LINQ is great. It goes without saying, however, that you should use the tools that make you more comfortable. I am only bringing the Entity Framework up in case you haven't come across it.

:-) "Work stealing", "loop decomposition" at runtime to take advantage of the available hardware parallelism, "data parallelism", "declarative programming for parallel computing"... It's soooo cool to see these concepts really making it to mainstream computing. Paul and I worked on these ideas some years ago*, building on the previous great works in the space. I even had a .NET implementation of the NIP runtime that hasn't been touched since 2002. Time to implement NIP.DSM using Parallel FX perhaps? :-)

I need more hours in a day or fork myself. Nah... forking is sooooo thread-oriented... I need to "evaluate in parallel" :-)

* Run-Time Support for Parallel Object-Oriented Computing: The NIP Lazy Task Creation Technique and the NIP Object-based Software Distributed Shared Memory
Savas Parastatidis, PhD Thesis, School of Computing Science, University of Newcastle upon Tyne, UK, 2000

Wow!

I've been hanging on my failing iPod (it's almost 3 years old now) for a while now, waiting for the new Zunes to arrive. Apparently they have been getting good comments out there, especially because the team opted to bring the new functionality to the v1 owners as well.

As soon as the Zunes are available to buy, I'll get the 80GB and the 8GB ones. Can't wait.

Isn't the Web great? Saying 'hi' from Global Research Library 2020, as part of a demonstration on what the Web is all about and what it can do for the future of research.