Re: [Tsvwg] Re: HighSpeed TCP for Large Congestion Windows

Reiner -

>Such applications would have to move *terra* bytes *back-to-back* to start 
>having that problem. I'm not convinced that such applications are running 
>on the *public* Internet today or in the foreseeable future. Please, 
>convince me if you think otherwise.
...
>As it stands, it seems that the ID describes a solution for a problem that 
>nobody has today or will could have tomorrow. Also, my understanding is 
>that IETF WGs only deal with *real world* problems, and that the more 
>academic problems are dealt with in the research community.

My understanding had been that the IETF was the right place to
bring proposed changes to TCP congestion control, whether the
changes were motivated by the needs of what is currently a small
subset of the community with underutilyzed high-bandwidth pipes at
their disposal, or whether the changes were motivated by pressing
concerns that affect the vast majority of users.  But I assume 
that the tsvwg co-chairs can clarify this for us.

My own understanding is that the HighSpeed draft describes a solution
for a problem that a small subset of people have today, and that
is not a very pressing problem at any rate (given the alternate
solution of opening N parallel TCP connections).  However, my own
view would be that this issue is clearly approriate to bring to the
IETF.  If people judge otherwise, that is fine by me.

My own *personal* view is that HighSpeed TCP would be an appropriate
work item for tsvwg, and further, that it would make perfect sense
for the HighSpeed TCP draft, and the related Limited Slow-Start
draft, to become Experimental RFCs, in the fullness of time, and
then to sink or swim on the basis of further experimental evaluation.
But if the IESG, the ADs, or the rough consensus of the working
group judge otherwise, again, that is certainly fine by me.
I am certainly not trying to argue that this is one of the more
pressing or earth-shattering issues on the table, just that it
strikes me as a appropriate issue to bring to the IETF, as the
standards body responsible for the standardization of TCP.

- Sally
http://www.icir.org/floyd/

_______________________________________________
tsvwg mailing list
tsvwg@ietf.org
https://www1.ietf.org/mailman/listinfo/tsvwg