lkml.org 
[lkml]   [1998]   [Sep]   [11]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: my broken TCP is faster on broken networks
Albert D. Cahalan <acahalan@cs.uml.edu> wrote:
> Real users do not ever tolerate 2 minute delays. Don't be silly.
> They hit the "Stop" button after 2 _seconds_ and start a new
> connection to avoid the stall. They kill the window and telnet again.
> In general, humans know to kill and restart a slow network connection.

Some do, some don't. Even when talking about HTTP/1.0, which is
not designed for very good performance, there's a lot of activity
where the user doesn't care about it.

But the internet isn't a user interface -- even when you put a user
interface on it. Do you really want to break all the automated transfers
which currently work (mail transfer, mirrors, etc. etc.)? Sure, they
don't occupy a lot of bandwidth currently. Guess why?

> New connections means _more_ network traffic, not less.

Sure. Another point is that in many cases users will give up and go
away when things get too slow. Neither of these is immediately relevant
to the exponential aspect of the tcp backoff algorithm. [Though they
are relevant in the larger sense, and I have seen some discussion of
arranging so that tcp timing information lasts from one connection to
the next, at least on a per-host basis.]

--
Raul

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/faq.html

\
 
 \ /
  Last update: 2005-03-22 13:44    [W:0.159 / U:0.308 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site