NOTE: This is an archive of my old blog. Go to http://gonium.net for my current website.

Conferencing & Presentations

Posted by md on February 15, 2006

DSCF0399.JPG

I think it is time to wrap up my experiences regarding presentations. Now, I investigted how I do presentations for almost 5 months, I have given a lot of presentations and used the opportunities to experiment a little. Now, being at a conference and seeing a lot of presentations, it almost hurts.

The picture above shows the average computer science conference powerpoint deck. Actually, the talk was really interesting and good, but the powerpoint slides were just useless. Please, get away from using bullet lists and putting a lot of text on slides. Of course, you can break the rule if it helps your talk. Think of giving a talk, and the slides are just the background that may help you – or not.

Use pictures. Not only diagrams, but real pictures. Get them from Flickr or iStockPhoto. Tell a story. Even if you are just explaining technical details, you can use an example to illustrate it. And connect your story to pictures. If you talk about the EGEE project, you can use very impressive pictures of particle accelerators, etc. Avoid templates. After the first conference day, no one can stand them any more. This is especially true for GGF slides, e.g. the IP policy slide.

My fear was that my presentation doesn’t look reputable. But this isn’t true – you must make sure you know what you’re talking about, and foster discussions whenever possible. You audience will appreciate your efforts.

Unicore rocks

Posted by md on February 09, 2006

During the lecture today, I also demonstrated how to submit jobs to the Unicore Testbed. It is very easy to install the client and get running.

If you want to play around with the Persistance of Vision Raytracer, here are my files: The Unicore job definition and the möbius input file.

Please note the Unicore job definition is a hack: it doesn’t exploit many of Unicores features. Of course, you can use any povray input file, and change the povray parameters to generate bigger images. But please keep in mind: this is a demonstration testbed, so please do not hack the system, or generate a lot of traffic.

Users don’t know anything… 2

Posted by md on February 09, 2006

I promised that there will be a presentation movie of todays lecture – but I didn’t manage to use my recording software properly. So there is no movie of the lecture available. But I will try to create separate movies for the chapters of the lecture – who wants to listen to more than a hour of lecture anyway?

Stay tuned…