On 07 May 2011, at 13:37, Igor Stasenko wrote:
So, where those customers to sponsor development of VM(s) they need so much? In general? They are with the 'cool kids': http://luajit.org/sponsors.html Nothing multicore-ish thought....
That's the main problem, IMO. Most smalltalkers i met (oh.. why.. not just smalltalkers), fear parallelism like fire. Refer to it as something too complex they can't understand. And i think that's why "Parallelism to masses" approach is largely failing , not just in smalltalk, but in whole software industry.
One problem we usually face is: good use cases Computational expensive problems are usually not a big issue, since they are very often just data-parallel. And getting that right by construction is not an intrinsically harder problem than recursion. And then we have task parallelism. Well, just use another app next to another next to another... Your favorite OS, or app server does that for you. Nothing hard here either. It becomes only tricky once you have an application that is kind of slowish, and you are migrating from a 2GHz processor to a 1.7GHz dual-core. Then, you most likely have to rethink the whole thing. And I don't see any magic that will help us with that. Whether that is hard or not, I don't know. The only safe thing to say: it depends on your application. We discussed a similar thing recently. And one of the insights from all the Erlang experience out there is, that the tools alone are not enough, you have to reshape your problem. And I guess that is the hard part. A language can only guide you in a certain direction, we still need to understand the problem enough to be able to gain something with parallelism. Parallelism per se is not the issue, concurrency is the devil... Best regards Stefan -- Stefan Marr Software Languages Lab Vrije Universiteit Brussel Pleinlaan 2 / B-1050 Brussels / Belgium http://soft.vub.ac.be/~smarr Phone: +32 2 629 2974 Fax: +32 2 629 3525