[IP] more on On the passing of full text on large mailing lists
Begin forwarded message:
From: Rich Kulawiec <rsk@xxxxxxx>
Date: July 8, 2004 1:32:32 PM EDT
To: Peter Wayner <pcw@xxxxxxxxxxx>
Cc: David Farber <dave@xxxxxxxxxx>
Subject: Re: [IP] On the passing of full text on large mailing lists
There are some fundamental errors and misconceptions in this.
I'm only going to respond to those I know something about,
because the rest are outside my expertise (I'm not a copyright
attorney and don't even play one on TV).
* First and foremost, the article is available for no cost. Nada. The
only people who lose money in the process are the creators who are
denied advertising revenue. The list is "stealing" something that's
available for free. That's about as low as you can go.
1. It's not available to everyone. The "set of people who can receive
email" is a larger set than than 'the set of people who can browse the
web".
All kinds of people are on non-IP network connections, behind
HTTP-blocking
firewalls, and so on. Remember: "Internet != web".
2. They're not going to get advertising revenue from me or anyone else
who's using a web browser/web proxy that blocks the ads, or who's using
a text-only browser. The only time I waste my bandwidth and my CPU
fetching
ads is when Privoxy misses one -- and that's rare.
You might argue that I'm somehow mysteriously "cheating" them out of
revenue too. Nope. I'm merely being selective about which URLs
I choose to use my software, my computer, and my connection to fetch.
3. Not everyone is so fortunate as to have a high-speed or flat-rate
connection. Many people have slow connections and pay charges per unit
time.
The ability to download mail (including downloading it in the
background,
or doing so at off-peak times to reduce costs) and then read it while
offline is very important to them.
4. Even people who are lucky enough to have cheap, flat-rate, high-speed
access often find themselves away from their connection. But they can
still read mail and compose replies while offline.
* Second, every computer scientist knows that passing a pointer (URL)
is much more efficient than passing the entire data structure. This
action just wastes bandwidth and clutters the disk of the people who
aren't interested in the topic. When the list is big enough and diverse
enough, it's just two or three steps up from spam.
1. The correct, canonical definition of spam (in the context of SMTP)
is:
unsolicited bulk email. Content and size are irrelevant: always have
been.
Thus, any traffic that you recieve via IP or any other list which is
using proper confirmed/closed-loop opt-in methods of subscription and
submission is not spam BY DEFINITION.
It's not even close.
2. Volume (both in terms of message count and message size) on the IP
list
is tiny. I'm on lists that routinely handle hundreds of messages a day:
have been for decades. They're quite commonplace. They're not a
problem.
3. *IF* we accept that volume is a problem -- and it's not -- but
if we do for the sake of argument, then using the web is even worse:
the nicely compact plain ASCII text content we see on this mailing list
will be bloated with HTML (often bad and broken HTML), Javascript, CSS,
and all kinds of other cruft that adds absolutely nothing in terms of
content -- but routinely increases the data transmission requirements
by an order of mangnitude or more.
Oh, and it's probably stored on disk too: browser cache.
4. Everyone on this list (or any other one, for that matter) should
know how to unsubscribe. They should also know how to operate their
mail client properly and delete messages that they don't wish to save.
And anyone who has to deal with substantial volumes of mail should
consider using a tool which makes classifying/sorting it easy: I happen
to use procmail, but it's far from the only available choice.
5. Locally, refused incoming SMTP traffic (consisting mostly of spam,
viruses, worms, backscatter from broken/abusive anti-spam/anti-virus
programs, etc.) is running at between 93% and 97% of total SMTP traffic.
In other words, even if mailing lists accounted for 100% of total
non-junk
mail traffic seen here, they'd still be less than 10% of what we have to
deal with. And they're not even close to that: a back-of-the-envelope
estimate is that they're perhaps 20-25% of legitimate traffic, at most.
Thus, aggregate volume of all mailing lists combined is a complete
non-issue: it's a trickle compared to the torrent of abuse.
* Third, the NYT compiles ratings of stories.
If they're doing it based on web hits, that's rather foolish of them.
---Rsk
-------------------------------------
You are subscribed as roessler@xxxxxxxxxxxxxxxxxx
To manage your subscription, go to
http://v2.listbox.com/member/?listname=ip
Archives at: http://www.interesting-people.org/archives/interesting-people/