[concurrency-interest] Are functional languages so much better at concurrency?

Joseph Seigh jseigh_cp00 at xemaps.com
Fri Feb 9 06:26:05 EST 2007


Michael Hicks wrote:

>Check out a recent article in Queue by Sutter and Larus, which  
>addresses the larger issue you are worried about (correctness for  
>concurrent apps), and briefly addresses functional programming in the  
>context of other solutions:
>
>http://research.microsoft.com/~larus/Papers/queue01.pdf
>(the published version is at http://doi.acm.org/10.1145/1095408.1095421)
>
>You might also find STM Haskell to be of interest (a version of the  
>Haskell functional programming language outfitted with software  
>transactional memory, used to support atomicity): http:// 
>research.microsoft.com/~simonpj/papers/stm/stm.pdf
>
>  
>

There's a Java STM implementation on Herlihy's home page here
Maurice Herlihy's Home Page <http://www.cs.brown.edu/people/mph/home.html>
that you can try out.  He has it set up so you can try out other STM
implementations against it.  I haven't gotten to the point on my STM
implementation to see if the api's are compatible enough for that.

STM works well as long as your program is obstruction-free,  a term
Herlihy coined.

--
Joe Seigh


More information about the Concurrency-interest mailing list