At 9/3/03 12:14 PM, George Kirikos wrote:

>That's my point, where I disagree with Robert.
> 
>A lot of the complaints these days are AUTO-GENERATED (e.g. spam
>filters either at the client end, automatically complaining based on
>the "From" field), or from corporate email server filtering (I must
>have received dozens of bounces/complaints about the virus stuff --
>"You're sending out virus, oh, you're a bad person."). I couldn't, as a
>business, go or recommend a service that would simply shut me down,
>based on that phony complaint.

Ummmm. Although I'm sure it's happened, I've never actually seen anyone's 
e-mail being shut down because of some sort of completely automated 
bounce-type report (such as a virus scanner) that complained to the wrong 
place.

If an ISP shuts you down based on that, you should certainly avoid them 
-- for competence reasons, not policy reasons. No decent ISP would 
consider that an actionable complaint.

Similarly, when someone has their e-mail turned off due to a malicious 
complaint (like your friend), the ISP should be able to promptly restore 
access as soon as a short investigation reveals it's false. If the ISP 
doesn't promptly resolve it, again, there is a competence problem, not a 
policy problem.


>Similar to a credit card rating system (where newbie customes start of
>with a low credit rating and limit, like $500, whereas customers with a
>long track record can have credit limits of $20,000+), perhaps an
>"abuse rating" is needed, to get rid of this "shoot first" dilemma. A
>brand new customer signing up from Malaysia or Russia with a hotmail
>account would not be given the same leeway that a 3 year old customer,
>whose address/telephone number are known to be real (i.e. spoken on the
>phone with them), and who has spent thousands of dollars in services
>over those years, etc. i.e. instead of a pure "black" or "white" list,
>one could be given a "score", and treated accordingly.

Yep. Some customers are most trustworthy than others, and in my opinion, 
such people should get more time to resolve spam complaints. You're 
absolutely right.

However, my opinion (or your opinion, or Tucows's opinion) of the 
trustworthiness of a customer is unfortunately irrelevant, because it 
holds zero weight with blacklist operators. They have no patience for 
anyone who says they need several days to investigate complaints, because 
that's exactly what spammers claim. An Internet-wide trustworthiness 
rating system would solve this, but all such schemes (other than gross 
systems like blacklists) have so far failed.

In addition, I've been amazed at how customers I would consider 
"trustworthy" end up spamming without realizing it. A common problem is 
buying mailing lists that they were assured were "opt-in"; almost all of 
these are fraudulent. The "trustworthy" customer doesn't even realize 
they're doing something wrong, but they are quite definitely spamming and 
should have been stopped right away.

Because of both spammers who lie, and honest admins who are surprised to 
find after several days of prodding a trustworthy customer for info about 
a mailing list that the  customer bought an "opt-in" list that was 
advertised through spam in a moment of incomprehensible stupidity, 
blacklist operators aren't going to give admins a while to investigate 
complaints, as it would make the blacklist worthless (in their opinion).

So a mail server admin has a choice to make: you can operate things with 
a zero tolerance policy and potentially inconvenience a single customer 
for a short while in the event of malicious false complaints. Or you can 
leave "trustworthy" accounts active during investigations (which can take 
days or weeks, because in my experience even trustworthy people tend to 
be slow to respond and vague when you ask them to document where they got 
a mailing list from) and resign yourself to being blacklisted sometimes, 
inconveniencing many customers for an indefinite period.

I'm sure you can still find many ISPs that have a "leave some accounts 
active while investigating" policy, and I understand why that's 
attractive -- that's the way things SHOULD work, damn it! But sadly, I'm 
quite convinced this would give you less reliable mail overall.

I think I can even prove it. Why have most ISPs have adopted zero 
tolerance policies? It's obviously not because they like pissing off 
their customers by cutting off their mail in response to complaints -- 
therefore they must be doing it because the previous, more tolerant 
policies created even more pissed off customers with mail problems. This 
is certainly the case here, by an order of magnitude.

Sigh. Our shared opinion about how unfair and wrong this all is doesn't 
change the reality a bit.

--
Robert L Mathews, Tiger Technologies

"A professional in an ape mask is still a professional." -Marge Simpson

Reply via email to