[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: TcpAgent::bug_fix_



>Does anyone know if bug_fix_ is fixing bugs that were in the ns-code
>implementation of TCP Tahoe/Reno,  or is it a fix to problems that
>are inherent with the protocols?  i.e., if one wants to run the
>protocol as it "usually" runs, should bug_fix_ be on or off?

Bug_fix_ is a fix to problems in actual implementations.
See the 1994 technical report on "TCP and Successive Fast Retransmits",
at "ftp://ftp.ee.lbl.gov/papers/fastretrans.ps", for more
information.  My guess would be that most current Tahoe and Reno
TCP implementations don't implement bug_fix_.   

Bug_fix_ is also discussed in the internet-draft
"draft-ietf-tcpimpl-newreno-00.txt", available from
"ftp://ftp.ee.lbl.gov/papers/draft-ietf-tcpimpl-newreno-00.txt".
This draft is being submitted as an experimental RFC.

- Sally
--------------------------------
http://www-nrg.ee.lbl.gov/floyd/
--------------------------------