On Monday, June 11, 2001, at 10:08 AM, Satya wrote:
> Seriously, you could go with the Usenet convention of rejecting
> messages based on ratio of quoted text to new text, with the problems
> that *that* brings. I know the problems. Don't tell me.
>
don't. Don't. don't even.
As one of the people who came up with that idea, many, many years ago,
it was an absolute, abject failure.
If you want to do something like this, you have to be smart about it.
One is to be sensitive to message length (it IS likely that a one line
response to a ten line message is legitimate, for instance, but a 10
line response to a 100 line inclusion is a lot more likely to be bogus).
The other, though is to be sensitive to whether or not the person edited
the message or not. What I prefer to do these days is look for an
included footer, as an indicator that they just left the entire stupid
message in the reply. If that exists, bounce it. If not, assume they did
edit it enough, and even if you don't agree iwth how they edited it,
don't worry about it and leave it alone.
--
Chuq Von Rospach, Internet Gnome <http://www.chuqui.com>
[<[EMAIL PROTECTED]> = <[EMAIL PROTECTED]> = <[EMAIL PROTECTED]>]
Yes, yes, I've finally finished my home page. Lucky you.
I'm really easy to get along with once you
people learn to worship me.
------------------------------------------------------
Mailman-Users maillist - [EMAIL PROTECTED]
http://mail.python.org/mailman/listinfo/mailman-users
