
A number of future additions are being thought about, in particular:


Objectification - One way to think about a process is to consider it
as a thread of execution through a set of objects. In an OO based
environment the objects instantiated in the process are well defined,
(non-OO processes could use a mapping). component
based apps natuarally fit in this model. Why think this way? Well if
a process is a hierarchy of objects then one can navigate into the
process and extract values for visualization or for triggering
actions. This would enable a kind of meta-programming with the processes.
For example one could monitor the actions of a web server, and when
a certain internal state is reached, spawn a process to perform some
desired action.



Data Driven externalizable methods - It is clear that one can classify
processes, i.e. many processes have man pages while others have info
files, and yet others have related howto's. Wouldn't it be nice to be
able to get to that documentation directly from a process monitor.
This is just one example of what a good externalized mechanism could
provide.


Logging, triggers and actions(internal or externalized).

Another thing that is useful is to be able to see what processes
ran recently, perhaps with some stats such as how much time/space
used etc. I have been thinking a collapsable panel at the bottom
could provide this function. A log could be kept in memory and
stored to disk if desired. Also as processes start/exit/... a
specific action could be performed. i.e. ring the alarm if user
yyy logs in on terminal zzz. 



