Big modules vs small modules in Erts

[Inswitch Solutions - Erlang Evaluation]

I have a 3000 lines FSM module spawned by more than 200 processes.
Apart from memory issues and talking about Erts performance, is a big module implementation (with one state data allocation) better than several smaller modules (with several state data allocations) for the same FSM?

I mean:
-module(fsm1)
s1(Event, StateData) -> s2
 
s2(Event, StateData) ->
...
3000 lines

vs.

- module(fsm1)
s1(Event, StateData) ->
wait for fsm2:s2
...
200 lines
 
- module(fsm2)
s2(Event, StateData) ->
...
200 lines

...


thanks,
Eduardo Figoli
INSwitch Solution

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://erlang.org/pipermail/erlang-questions/attachments/20040821/4b722333/attachment.htm>