some relevant threads on comp.lang.functional

[Ulf Wiger (AL/EAB)]

This morning, the following question came up on comp.lang.functional:

http://groups.google.se/group/comp.lang.functional/browse_thread/thread/7d7038d2e52b92bf/d94a67e5d0729a35?hl=sv#d94a67e5d0729a35

"I've been mulling over Armstrong and Virding's paper, 
"One Pass Real-Time Generational Mark-Sweep Garbage 
Collection" (In Proc. Int. Workshop on Memory Management, 
1995, see www.sics.se/~joe/pubs/historyGC.ps <http://www.google.com/url?sa=D&q=http://www.sics.se/~joe/pubs/historyGC.ps>) 
Does anyone know whether this was implemented and, 
if so, how well it performed? [...]"



Also, I've been drawn into a debate on the thread "What ML language do you recommend?"
http://groups.google.se/group/comp.lang.functional/browse_thread/thread/6f1029e6176c57a7/d042b21ace010929?hl=sv#d042b21ace010929

The part that concerns me has to do with Concurrent ML vs Erlang, concurrency models, etc.
I quoted Roger Price's experience with Concurrent ML, for example, and the most recent posts discuss the book on Concurrent ML (which I haven't read), and whether synchronous communication is inherently more powerful than asynchronous communication.

Feel free to chip in. (:

/Uffe