Mon Mar 24 18:12:05 CET 2003
On Mon, 24 Mar 2003, Chris Pressey wrote:
>My concern, when I pointed that out, was more about the
>thousands of plain servers and gen_servers that have been
>written, although mdisp might indeed provide some salvation
>for short-lived processes.
I'm not sure if it's possible, or even relevant to solve the
general case of a server trying to make sure that there are
no more messages before it finally dies. After all, this is
fundamentally a race condition.
The dispatcher contrib had an example adaptation where
gen_server-like processes were dispatched, and went to sleep
automatically after a pre-configured idle period. It was not
the most efficient, but the idea was to allow for a way to
quickly turn gen_server processes into dispatched servers.
There was once discussion about being able to give a process
several registered names, allowing it to act as a proxy for
temporary processes (I'm not sure if I do the idea justice -
but I'm sure someone will correct me otherwise.) That might
allow for a much cleaner adaptation to an mdisp model.
Ulf Wiger, Senior Specialist,
/ / / Architecture & Design of Carrier-Class Software
/ / / Strategic Product & System Management
/ / / Ericsson AB, Connectivity and Control Nodes
More information about the erlang-questions