On Fri, Apr 30, 2010 at 9:28 PM, Aran Donohue <aran.donohue@gmail.com> wrote:

Thanks for the excellent links, that's exactly what I wanted. It's interesting that they've chosen not to base the new work on libevent.

The reason was mostly performance concerns due to libev(ent) using callbacks to signal events. Callbacks from C into Haskell can be inefficient. From the FFI addendum:

"Optionally, an import declaration can specify, after the calling convention, the safety level that should be used when invoking an external entity. A safe call is less efficient, but guarantees to leave the Haskell system in a state that allows callbacks from the external code."

As an aside, I really don't think that the case study should be given any more linkjuice as a response to GHC/Haskell IO concurrency questions. While it's a wonderful tutorial on the programming technique side, it's a decade old and was written at a time when serving 4000 requests was a reasonable benchmark. These days modern web servers are moving more and more toward handling tens of thousands of concurrent held-open connections---a different metric and a different scale.

The event library, linked by Don, handles tens of thousands of idle connections without problems (see the idle connection generator [1] I created). Bryan wrote a simple HTTP server [2] that handles 20,000 requests per second on one core.

1. http://github.com/tibbe/event/blob/master/benchmarks/DeadConn.hs
2. http://github.com/tibbe/event/blob/master/benchmarks/StaticHttp.hs

Cheers,

Johan