> - I see two concerns: translation/modification of data and input-output
> of data. Which is the main one for InfoMover (to me, the name implies
> the second one, and a pipeline the first one)

on second thought, I think the thing to do is to settle IO stuff first
and then add in modification functionality later (should be easy in an
event-like pipeline anyway =). Maybe also the right thing to do for your
book...

...another thing interesting in a COP system is how to handle multiple
sources and/or multiple sinks. Will that be part of the future set?

- Leo



--
To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>

Reply via email to