About the bugs

[Michael Welzl < >]

Hi, and happy new year everyone!

I just wanted to share that I found that the remaining bugs in my code seem 
to affect only situations of multiple flows.
It seems very stupid: because I “hacked” the code into the OO structure of 
ns-3 like I did, variables are not tied to a specific node (or stack 
instance), but they are shared… so if you have multiple senders, they share 
state information, which of course is nothing but a stupid bug.

This seems pretty easy to fix, but something’s a bit odd in my tests now…

My big problem is that I’m COMPLETELY tied up with family stuff these days, 
and I have very little hope of finding more than 10 minutes to work on this 
before returning to Rome (on 7th). Sorry…   but if I DO find a little time 
and get to fix them I’ll of course tell you!!!

Anyway, if you apply the fix I recently sent (simply lower-bounding the RTT 
estimate with 1ms), the code should really work fine for a single flow only 
(if you find that it doesn’t, please report that bug to me!).

BTW, obviously, I mean “a single flow of the TEB kind”, of course it 
shouldn’t be a problem to run one TEB flow and 100s of other flows!

Cheers,
Michael