[Mailman-Users] Allowing users to join without specifying pas swords
Chuq Von Rospach
chuqui at plaidworks.com
Fri Jun 15 21:36:45 CEST 2001
On Friday, June 15, 2001, at 12:26 PM, alex wetmore wrote:
> Computers are cheap. Bandwidth isn't necessarily cheap. I run with a
> very high chunking factor because my MTA properly handles it
But -- if this went away, what would it do to your bandwidth? Are you
paying by packet? If not, the only PRACTICAL difference is that if the
pipe is full, delivery slows down.
What is the real cost is? That messages show up five minutes later than
they would if we don't do this?
Before we play "this is a witch! burn her!" games (grin), we ought to
figure out what the real impact is. If people want, I can post some
stuff (no hard numbers, unfortunately) I've seen based on tweaking this
stuff on my big machines. But the funny thing is, the biggest difference
I see isn't network, it's disk I/O. and most of that is because of
sendmail's design for dealing with this stuff...
I'm not saying Alex is wrong -- he's got some good points. but we need
to quantify what those points are and what the impact is, so we can
decide just how to move forward on this. If the impact is, oh, a 5%
degradation in network throughput, frankly, it's not (IMHO) worth
special casing it. If it's 30%, we need to special case it AND document
it so people know how to tune their systems. And if it's 60%, maybe we
shouldn't do it..
but we need to figure out what the impact is, and not guess or make
Chuq Von Rospach, Internet Gnome <http://www.chuqui.com>
[<chuqui at plaidworks.com> = <me at chuqui.com> = <chuq at apple.com>]
Yes, yes, I've finally finished my home page. Lucky you.
Love is the process of my leading you gently back to yourself.
- Saint Exupery
More information about the Mailman-Users