Lightning Talks

Vagrant

  • Vagrant loves Python
  • Building and distributing VMs
  • Gives isolation, repeatability, and verification
  • Move dev to virtual machines
  • Move production ops scripts to setup environment
  • Vagrant command line, to manage life cycle
  • Designers can use it too
  • http://vagrantup.com

Testing CSS

Pyparsing

  • Time trial using Pypy
  • Search for integers in a string of random alphas and numbers
  • Pypy ~10x faster
  • Verilog parser (~16k lines)
  • Cpython (500 lines/sec)
  • Pypy (1131 lines/sec)

Pandas

  • @wesmckinn
  • Agile Tooling for Small Data
  • First need to small the small data problem before big data
  • DBs, Flat files, time series, mean you may want it
  • indexed data structures for relation data
  • Fast manipulation tool
  • Data alignment
  • Join merge
  • group by
  • Reshaping/pivot
  • In memory and fast
  • Meant for quant finance application backbone
  • ~26k loc
  • In productions since 2008
  • Data Analysis is dominated by thing like SAS
  • Lots of people want to expand in these areas
  • Operations to naturally select portions of data
  • Can plot data
  • Would love collaborators

DSLs

  • Peter Wang (@pwang)
  • Crazy crazy ideas
  • Would like Python to ignore some syntax where we can do whatever the hell we want
  • It might be awesome
  • Calling it extern
  • Just syntactic sugar
  • Hacking import hooks to make it work
  • .pydsl file
  • uses pyparsing under the hood to transform the dsl
  • Aimed at scientists
  • People want it: weave, numexpr
  • Everyone needs it
  • Let’s Python assimilate into existing systems

stackful

  • @erikrose
  • This is a hack
  • Wish things weren’t global
  • Dynamic variables like in Perl
  • Perl has local variables which leaks onto things it calls
  • stackful implemented as with statement
  • Thread safe
  • Implementation is funny
  • No hook in Python for reference
  • Just override every single magic method in Python
  • Should be able to be used