On May 25, 2006, at 8:34 PM, Robert Dockins wrote:
If you want to deliver source code to be executed elsewhere, you can use hs-plugins or the GHC API (in GHC HEAD branch).
hs-plugins is too heavy since it runs ghc. I don't need to deliver any type of source code, just a function call and its arguments.
Beyond that, I'd say there are a few too many free variables in the problem description. What would be the design goals and non-goals for such an RPC mechanism? What problems prompted the original lisp implementation? What about fault tolerance, reliability, security? etc.
The design goals is just to have a simple RPC mechanism with synchronous (request/reply) and asynchronous (send a message) facilities. The problem I'm trying to address is running a whole bunch of software servers for scalability. Reliability and fault tolerance are taken care of by having redundant servers and security is not addressed at all. It seems that I can do something similar in Haskell by having heterogenous lists of arguments and making each RPC call into a type class but then I would have to implement the functions myself. I wish Glasgow Distributed Haskell (GdH) was more active and visible! Does anybody know its current state? Shouldn't it be an extension to GHC just like Concurrent Haskell is? Joel -- http://wagerlabs.com/