On Feb 22, 2005, at 1:46 PM, Kentaro A. Kurahone wrote:

> I was sort of eyeing LRP (http://www.cs.rice.edu/CS/Systems/LRP/)
> as something to look into.  After other code I have in progress
> stabilises though.

I looked at LRP a number of years ago (back in my NASA days).  I came 
to the conclusion that is has some fundamental flaws, but I now can't 
recall off the top of my head what they were.

One sticky scenario is "What about packets that are meant to be 
forwarded?  You have to do some processing on those to even determine 
that in the first place."

It seems to me that LRP could also be problematic for high-bandwidth 
links where you are using very large TCP windows.  If you are always 
deferring TCP processing until the application runs, you might be hard 
pressed to keep the pipe full.

-- thorpej