Thoughts on PyCon PL 2015

I had a chance to participate in 2015’s edition and below are my opinions about both — organisation of the whole event and featured talks. I enjoyed it so big thanks to people devoting their time to make it happen.

OctoDocker

Good idea how to use containers for making sure that pull request really works by setting up patched project inside isolated environment which is easily accessible. Pity it hasn’t been open sourced already as talking about new tool which can’t be used soon after the talk doesn’t make much sense. GitLab has already built-in continuous integration support so would be perfect for me to have OctoDocker integrated in the same way but I doubt it’s possible now.

Apache Mesos for Python applications

I’m not an expert when it goes to this tool so I learned a bit. Lots of demos and quite good balance between theory and practice. Topic is quite fresh and interesting so would like to hear more about it in the future. Speaker went beyond basic tutorials you can find on the Web so definitely it was worth to see it.

What’s Eating Python performance

Rather basic things were introduced like to avoid premature optimization or measuring carefully. Anyway I feel it’s kind of topic which should be refreshed from time to time to not forget state of art of tackling performance issues.

Python & PyPy performance (not) for dummies

It was highlighted that PyPy is something you should consider before even thinking about changing anything in your stack like programming language in extreme cases. Hard to discuss with it. Additionally vmprof was introduced but most of demos didn’t worked because of poor Wi-Fi. Generally speakers should be prepared for that as it’s a known issue while conferences or meetups. I was a bit disappointed as it was rather really basic and didn’t introduced anything new for me.

Import Deep Dive

Solid technical introduction to how import statement works under the hood. It wasn’t overloaded with too many internal details so it was easy to grasp it within 30 minutes.

Inside the Hat: Python @ Walt Disney Animation Studios

Presentation was interesting because Disney is doing cool things but it wasn’t filled up with many technical details. I only learned they use GitLab internally and Python for doing some parts of their pipeline and as a glue for other tools.

Generally speaking most of the talks I had a chance to participate in were rather disappointing. You can learn more by f.ex. reading introduction to Docker or tutorial on http://docs.locust.io/en/latest/. I would rather hear more about people’s personal experience while working with those tools rather than reproducing manuals from the Web.

As of the organization of the whole events I feel it was fine. I’ve missed paper agendas all over the place. It’s always safer as Wi-Fi din’t worked (issue on most of the conferences) plus the mobile version of the conference’s page wasn’t really readable.