Hello folks, back from second and final day of PyCon India 2013. The sessions today were better than yesterday one an average, though there were a few disappointments. Proceeded to the venue with Guruprasad, just like yesterday and were just in time for the keynote speech by Kenneth Reitz. Kenneth in the guy behind the requests HTTP module in Python.
Kenneth’s keynote was more tuned towards creating and running free,open source projects, and he was sharing his own experience. He stressed the importance of having a good documentation for any open source project. In fact, he said that the first thing that we should do with any project that we start is creating the README file with some examples of how the API should be used. That way, he says, we end up creating the APIs that we want and not, as it usually happens, use whatever we created.
His keynote went on to say, that we should look at catering to 90% of the use cases instead of trying to satisfy every possible use case. He compared creating software to that of sculptures and art – They are not manufactured or engineered. They are discovered 🙂 He went on to talk about what the general etiquette of the contributors and maintainers should be in an open source eco system.
After the key note, the various sessions started. I attended the sessions on “Predicting Blackswan events”, Data mining, and “Data visualisations with Python and PowerPoint”. I attended a BoF on Testing and an “open-space” talk on Emacs.
The first session on “Predicting Blackswan events” by Mike McKerns was really a top-class talk. It involved a lot of probability and higher mathematics and there were considerable parts of the talk that I did not understand. But the modules( Mystic and Pathos ) that he and his team has built to conquer this relatively new area of mathematics was wonderful. I would need another couple of sessions over his slides and the resources he has shared.
The second talk of the day that I attended was titled something like “Data mining, entity disambiguation, designing data structures of beautiful algorithms”. It looked more like a classroom theoretical lecture than like a session in a popular national conference. It was a disappointment from my point of view and didn’t like it much.
Then came the session from Anand S on Data visualisaton with Python in PowerPoint. He showed how you can do data analysis in Python and represent them as interactive PowerPoint presentations. The talk was really awesome, during which he demonstrated some unbelievably complex charts on PowerPoint generated using Python.
The BoF on Testing was organized by RedHat. It wasn’t a very great session. But there some nice discussions (actually, some heated discussions). There was a short talk on Nitrate – an Free Open Source Test Case Management System(like Quality Center or Test Link for those from the corporates). I guess a lot of corporates should seriously consider ditching those over-priced extremely bulks QC and switching to one of these cooler FOSS solutions. There was a quick demo of “coverage” to cover test coverage for Python.
The final session was the “Open Spaces” talk on Emacs by Noufal Ibrahim. It was by far the most interesting session this PyCon. He spoke on Gnus, Org mode and a lot of other stuff. Org mode demo really made me believe emacs IS magic 🙂 If you are an Emacs user, at a beginner/intermediate level be sure to check out Noufal’s Emacs Movies. It is a nice site where he publishes web casts on some stuff that are cool in Emacs (well, every thing is cool about Emacs) .
In closing, this year’s PyCon was very good. The T-Shirt was nice 🙂 The venue, arrangements, food, and a handful of talks – all these were excellent. Thanks again to the organisers and volunteers who made the event a big success.