The final day of the International Lisp Conference 2009 had two invited speakers and lots of lightning talks. First a bit about lightning talks:
John Stracke talked about his Lisp implementation on Python which he named Adder. The summary of description is at http://www.thibault.org/adder/ In his words, “Adder is a Lisp-1 which compiles to Python bytecode. It aims to integrate seamlessly into Python: every Adder function is a Python function, every Adder list is a Python list, etc.”
Richard Greenblatt lightning talk was at the intersection of biology and computing and again his talk could be tought of as connecting in some ways to Sussman’s Evolvability and Robust Design presentation.
Vladimir Sedach talked about why he considered Lisp very powerful by summarizing what he has written in ‘I think I finally understand what Alan Kay is saying about Lisp‘
The final lightning talk was by Sussman (in addition his lightning talk about the switch to Python from Scheme, his invited talk on Evolvability and Robust Design and his contribution to The (Abridged) Art of the Propagator which clearly marked him as one of the most energetic and productive attendees of the conference (I wish I can have half of his energy when I reach his age :)). Sussman provided some insights on how the complex software systems can be analyzed using paradigms from physics, such as thinking about in computation in terms of state spaces and state changes.
The ‘big’ talks of the final day came from various topics. Howard Shrobe from MIT Computer Science and Artificial Intelligence Laboatory gave a talk titled ‘Towards a Secure Programming Language: An Access Control System for Common Lisp’ where he provided a very detailed description of a secure system he was developing and he claimed that if those ideas were adopted by UNIX community as well as Intel, Microsoft, etc. we would have fundamentally more secure systems (he gave anecdotal examples in which he tried to do some changes to his code but the secure system he built made things very difficult to change).
Robert P. Goldman from SIFT, LLC and Steven A. Harp from Adventium Labs had talked about ‘Model-based Intrusion Assessment in Common Lisp’. It was stressed that one of the first things system administrators do is to turn off their intrusion detection systems because of the high frequency of false positives. So the talk provided details about how to use Lisp in order to build a Bayesian system that reduced the rates of false positives by using probabilistic techniques and belief networks.
Alex S Fukunaga from Global Edge Institute, Tokyo Institute of Technology talked about his ‘Parallel, Lisp-Based Genetic Programming System for Discovering Satisfiability Solvers’. It was interesting to see Lisp put to use to solve such hard optimization problems (and being compared to WalkSAT) as well as seeing even completely randomized algorithms could reach to good results in some cases.
Tiago Maduro Dias, Technical University of Lisbon and António Menezes Leitão from INESC-ID/Technical University of Lisbon described the Jnil system in the talk titled ‘Jnil: From Java to Common Lisp‘. And this of course reminded me of CL for Java project. Lisp to Java interoperability (and vice versa) seems to be a very hot topic nowadays. Add The Moby Scheme Compiler for Smartphones to the list and you have your favorite language running on mobile platforms 😉
And now a bit about the lightning talks in the afternoon session:
– Nick Levine talked about updating programs: ‘Patching made easy – Lisp wins again’. Here is his beautiful code: http://enlivend.livejournal.com/8308.html
Zach Beane, the graphics utilitymeister talked about his AutoMotivator coded in Common Lisp. He said how he used CL as a glue language to command ImageMagick and nginx web server to serve images created by Common Lisp.
– Ryan Culpepper talked about ‘Debugging hygienic macros’.
The big talks of the afternoon session came from the invited spekaers: Krishnamurti talked how to use Scheme to program mobile devices (it is all about using your phone nowadays) and how this helped middle school students learn algebra (they seemed to be really motivated). The details are in The Moby Scheme Compiler for Smartphones.
Olin Shivers talked about ‘The Anatomy of a Loop: A Story of Scope and Control‘ in which he described how and why he was getting sick of rigid looping constructs and how he managed to build a language to build loop constructs of your own! That was pretty advanced stuff starting with the highest level constructs and moving to intricate details of low level compiler optimizations.