From Linda, 8/3/04:

Most of the linac steering feedbacks will be at 10 Hz, due to slow correctors.
I expect this linac system will be cascaded. There will be a 120-Hz loop
at the linac-to-undulator with fast correctors. Probably it will be uncascaded,
but it should be simulated as a system.

However, we still will have 120-hz distributed/cascaded feedback, for the 
energy and bunchlength.

Regarding the question of whether we need to mix IOCs with SLC micros in
fast feedback:  I think not.
Our only known possible example was FB31/AB01, energy feedback for 
straight-ahead experiments. Patrick believes we can give LCLS new controls 
over the phase shifters, so that FB31/AB01 energy feedback can be left intact, 
using SLC micros.