The title seems a little out of place given that this is a predominantly .NET Oriented Community. But this was an opportunity to immerse myself in a different environment where people face similar challenges. As a side benefit, it was also a little nudge towards learning Python. The NLTK talk certainly helped this motivation along.
I have skimmed over .NET libraries in the past (i.e. SharpNLP); but there is definitive value in examining an earlier and more mature project. As evidenced by the presenter, Sean Boisen of Logos Bible Software, NLTK is not only a useful academia tool, it also serves practical purposes in his organization. His major push being that analytics (based off of well known standards such as Levenshtein distance, Zipf's Law) on large data sets were much more valuable than AI interpretations. Given the disappointments and progress in the AI field, I can understand his assertions. It turned out to be a light dose of Python as he didn't dig deep into the bowels of NLTK and more complex examples.
The other benefit of attending LinuxFest is in picking up on similar problems/solutions that the Linux community faces. While the technology stacks differ, the problems we face are inherently the same. A great example of this was a network protocols and security talk. Man in the middle and time-based attacks apply regardless of the stack. Another example was on the subject of caching via xcache and memcache. Other highlights included The State of Mono Project and Ted Haeger's presentation on Bungee Labs' Platform as a Service. It contained quite a bit of marketing buzz such as "build.theApp() != theCrap" but did unearth some interesting issues startups in the open source community face around licensing models in the cloud. It was an eye opener as I am not in the startup space nor have I given much thought to this area. Check out his blog post on Sharing Source Code In The Cloud.
This turned out to be an awesome learning experience. It has been a couple years since I had an Ubuntu (hoary hedgehog) server setup and ran on a LAMP stack. While I was definitely exposed to some neat tools on the Linux platform, the most important takeaway was one I sought for prior to the event. That is: Keeping focused on the problems at hand and not getting too entangled into a particular technology stack. These, along with Google TechTalks, Papers and MIT OpenCourseWare, are the types of things that can prevent one from having tunnel vision and becoming complacent.
