[Haskell-cafe] Haskell RPC

Joel Reymont joelr1 at gmail.com
Fri May 26 04:20:25 EDT 2006


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/







More information about the Haskell-Cafe mailing list