Hi,<br><br>Don't want to break the current thread about lispy Erlang and realtime support but there is just a thought about the initial question I need to get off my chest.<br>> Does anyone see the benefit of a Lispy version of Erlang on the JVM?
<br><br>I would see two huge benefits if Erlang as it is today would run on the JVM:<br><br>a) Much easier to call Java from Erlang. JInterface works but is not really what I want to sell as the future. Besides, why reinvent the wheel - Java has so much to offer.
<br><br>More important:<br><br>b) Much easier to sell Erlang. The JVM certainly belongs to the big players in big corporate server environments (finance, telecom, ...). Everybody from the decision makers to the admins are used to the fact that many server processes will run in the JVM and that it will work.
<br>On the other side the whole Erlang univers from the code to the Erlang VM is mostly unknown by them and what is unknown is considered obscure. Making it quite difficult to sell. Just think how long it took for Java and especially its VM to be accepted - even given a much broader support from big players. You might remember all the wrong preconceptions.
<br><br>Now consider Erlang in JVM: 'Well, we write this application in Erlang but for you, dear customer there is no difference. You will run it in your proven Java VM. Your Engineer do not need to adopt to a new environment. And no, you do not have to mind upgrade paths. If a new technology evolves we can for sure expect support for it in Java and we can always use Java as fallback if there is no production ready implementation in Erlang. We just use Erlang where it helps us to implement faster/better quality and with less cost than we could with Java.'
<br><br>Yes, I read about Scala. :)<br>And yes, I know that i might dream a bit because it might be very difficult (tail recursion?) to port Erlang to the JVM.<br><br><br>Cheers,<br>Michael