It seems that concurrent (non-STW) GC for non-functional languages (randomly accessed mutable data) is a myth. Yes, Java claims to have an efficient GC (on pap

It seems that concurrent (non-STW) GC for non-functional languages (randomly accessed mutable data) is a myth. Yes, Java claims to have an efficient GC (on paper). In reality, everyone know how it works (it bloats, and GC pauses are increasing dramatically). On the contrary, Haskell has less sophisticated GC but much smaller and more predictable GC timings. btw, some really smart people are suggesting that it is the code organization, not GC algorithm is a crucial factor. http://home.pipeline.com/~hbaker1/LinearLisp.html http://home.pipeline.com/~hbaker1/LinearLisp.html