CDR: [Spam wars, continued...]

[Bill Stewart]

At 12:22 PM 11/5/00 -0600, Jim Choate wrote:
>On Sun, 5 Nov 2000, Declan McCullagh wrote:
>
>> Well, let's take this up one level of abstraction. We can stop spam
>> from flooding our inboxes (an economic bad) by:
>> 1. law
>"Congress shall make no law ...".
>> 3. cypherpunkly end-user technology
>I obviously support anything an individual wishes to do with respect to
>making choices, provided they don't involve me without consent. This
>aspect should be pushed strongly.

I agree with Jim that anti-spam laws are bad in principle;
in practice they're usually worse :-)
Some kinds of cypherpunks technology don't involve the law; some do.  
For instance, user-supplied filters can trigger libel laws
("Hey, your filter called me a SPAMMER!  I'll SUE!").
Carrier-end filters have a similar effect (I'm sure Jim would prefer
that any such filters be installed by the user, not the ISP,
even if they do get installed at the ISP's end of the wire.)
Either way, they're reputation servers of some sort.

But another set of cypherpunks technology is 
"the cure for bad speech is more speech" - responding to unwanted
streams of bits ("Buy our SPAM today") with equally unwanted bits
("Ping of Death", cracking, and other attacks.)  These tend to
violate ISP acceptable use policies a and occasionally laws.

>> 2. AUPs with backbone providers/hosting services (industry self-regulation)
>
>I oppose these because I don't think some organization should have control
>of my speech simply because I purchase a service from them. If I buy, for
>example, a 128k ISDN line what the content of that 128k is most assuradely
>isn't my providers interest. It violates the spirit of the 1st.
...
>Economics in general is not the way to set ethical standards.

>4. social contracts (for those of anarchist and libertarian bent)
>
>Considering human psychology, not bloody likely.

But social contracts and economics are two sides of the same coin.
While the DoD-funded ARPANET policies against commercial speech were
censorship, the Usenet prevailing opposition to newsgroup spam was
a social-contract thing.  In many cases it's broken down,
except on moderated newsgroups, but a lot of it is still there.
And ISPs implement spam blocking not because they care
(some do, some don't, usually depending on pricing models),
but because their users keep telling them "Hey, I don't want spam."
And social-contract relationships between ISPs mean that
if you don't want somebody spamming your ISP, you don't spam theirs,
and if somebody keeps spamming you and your users don't want it,
you stop doing business with them.

But spam is one thing, and politically incorrect content is another.
Some ISPs have no-porn policies because they see a market for it,
but many have them because they don't want legal problems if some
prosecutor decides that they're selling obscenity in an election year.
And with others, it's a marketing / branding thing - so if you don't
like that kind of ISP, don't do business with them.

>5. technical standards (ala Open Source)

Open source technical standards doesn't have a lot of relationship 
to the spam problem.  They do mean that you can find and fix 
opportunities for spamming that are being abused, and that
spammers can find and exploit new opportunities.


Back to the 128K problem, with an ISDN line I agree it's nobody's business.
You're getting dedicated bandwidth to a router pool, and whether
you're on full time or not is strictly a pricing thing
(and you'll have a much different opinion depending on whether you're
charged by the minute or jsut flat rate.)
With cable modems, it's a bit different, because the technology is
very asymmetric.  IMHO, the carriers like Excite at Home are clueless and
annoying in their server policies - but it's partly because they don't
have consistent traffic management technology, and partly because
they can get rid of most of the problems by blocking web servers
and now Napster which tend to be resource hogs.
They'd rather be able to shut down users blindly if they have
performance problems rather than having to argue about your web server
using much less bandwidth than your "client" video-conference program,
even though the web server is a much more efficient way to display
your fish tank and coffee pot to the world.  


				Thanks! 
					Bill
Bill Stewart, bill.stewart at pobox.com
PGP Fingerprint D454 E202 CBC8 40BF  3C85 B884 0ABE 4639