Print

Print


Sean,

This is great.  Congratulations!

I’ve done some further work on using MASON with Clojure but haven’t abstracted what I’ve done out of the app that it’s in, nor fully documented it, or added all of the features I originally intended.  Is this potentially relevant to the new grant?  If so, I can try to expedite making what I’ve done publicly intelligible (after I finish the current five projects!?! :-).

My goal using Clojure and MASON this time around was not to make the code as fast as possible, but rather to make the code more idiomatic to Clojure.  I treated speed as a secondary goal, for the most part.  I think that many people who’ve been using Clojure for a while by choice like Clojure because various aspects of the design integrate in a way that makes a lot of tasks easy and fun.   When I was trying to make MASON as fast as possible with Clojure, I produced Clojure code that was ugly, in my opinion, and that didn’t take full advantage of the features that make Clojure a pleasure to use.  I now think that if top speed with MASON is the goal, I think you're better off just using pure Java.

For me, the point of writing ABMs in Clojure is that I love programming in Clojure, and as far as I have seen, MASON is the best ABM library for Clojure, even if there are things that have to be thought through to use the two together as comfortably as possible.

This is the application that uses my new MASON-Clojure interop strategies:
https://github.com/mars0i/pasta
There are some notes in the doc directory.


Marshall


Marshall Abrams, Associate Professor
Department of Philosophy, University of Alabama at Birmingham
Email: [log in to unmask]<mailto:[log in to unmask]>; Phone: (205) 996-7483;  Fax: (205) 975-6610
Mail: HB 414A, 900 13th Street South, Birmingham, AL 35294-1260;  Office: HB 418
Website: http://members.logical.net/~marshall

On Jun 7, 2017, at 7:52 AM, Sean Luke <[log in to unmask]<mailto:[log in to unmask]>> wrote:

George Mason University has received a large three-year NSF CRI award to extend and enhance MASON starting this Fall.  The grant is to myself, Prof. Robert Simon, and Prof. Andrew Crooks.  This means that both MASON and ECJ have overlapping CRI awards.  We have a lot of work ahead of us!

Thanks to everyone on the MASON-INTEREST and ECJ-INTEREST lists and to the participants of the 2012 NSF MASON workshop, which proved integral to the proposal effort.

Sean