On Tue, 3 Dec 1996 10:00:16 -0800 Dave Crocker <dcrocker@brandenburg.com>
said:
> When someone says "Many X do Y", a response of "That's not true,
>because there is one X that clearly doesn't do Y" is simply
>non-responsive. It entirely misses the point of the original assertion.
Dave, I run the largest mailing list site in the world (3,037,510
deliveries yesterday) and I am intimately familiar with aggregation
patterns for traditional mailing list traffic. My knowledge is based on
factual data, not speculation. I did not give the AOL.COM example just
for the sake of finding a little contrived example that doesn't fit your
theory, but because it is highly relevant to the discussion. A handful of
hostnames account for a large fraction of the Internet population, and
yes these people do subscribe to mailing lists. The fact that massive
aggregation to these sites is possible not only means that major savings
in bandwidth and turnaround time are possible, but that these sites would
be direly impacted if aggregation did not take place. As a responsible
Internet vendor, we strive to do what is best for the community, even
when this means more programming time to achieve the same result.
> To give a specific example, from my original statement:
>Enterprise-wide network mailing lists are often entirely internal to the
>enterprise and may well show NONE of the aggregation that we have been
>discussing.
It may very well be that enterprise-wide network mailing lists have this
property, however this is of no consequence because enterprises have tiny
number of mailboxes compared to the Internet and much smaller mailing
lists. Not many companies have 10k employees to start with, let alone 10k
who use e-mail regularly. A typical enterprise list has some 100
subscribers, which is a piece of cake even for the freebies. Finally, in
most cases mail systems like cc:Mail or MS Mail are used, and the
bottleneck is by far the mail gateway. In this context, aggregation is
usually irrelevant, whether possible or not.
In the interest of not turning this into a flame war, let me just clarify
that this is all based on data that L-Soft has (both from its own mailing
list service and, for the second issue, from information provided by
customers) and that I am not going to engage in a flame war about
theories when I can just look up the answer in our files.
Eric
|
|
