ietf
[Top] [All Lists]

Re: filtering active content

2001-07-25 17:50:02
From: Keith Moore <moore(_at_)cs(_dot_)utk(_dot_)edu>


No one says a bulk header would be harmful.

actually, in the past it has been harmful.

- some list expanders used it as a loop-prevention flag - assuming 
  that any message that had precedence: bulk had already been 
  fowarded to the list.

- one mail gateway used precedence for 'priority' and would bounce 
  any message that contained a precedence keyword that it did not
  understand.  IIRC "bulk" was one of those not understood.

the latter mail gateway is probably long gone (and may have been fixed
before it died) but it wouldn't surprise me if some list expanders
still had that behavior.  and the IETF list gets forwarded to lost of
sublists.

Neither of those count as harmful.  They may be unintended, but no
worse.  The first is probably a desirable feature because of the harm
non-technical harm that is very often done by forwarding one mailing
list into another (i.e. people who don't understand what's going on).
Besides, anyone using precedence:bulk for loop prevention probably has
messed up other things.  The second is no worse or much different from
any other obvious bug.  Anyone who's fiddled with MTA's has done worse.


that's what happens when people use nonstandard protocol extensions -
without a widely agreed-on definition of what the extension does, 
people will use the extension to do whatever they like - whether it
conflicts with existing usage or not.

And so the IETF's main list will continue to be used as a vector of
virus for incompetently designed systems and flooded with idiotic
vacation messages, viruses, and amazingly stupid virus warnings.
People here will continue to say bad things about Microsoft for being
slow to fix bugs, and ignore the signs in some the virus and vacation
warnings that imply their sources are running Windows 95.  How is
Microsoft supposed to fix those machines?

Microsoft is a predatory monopoly that abuses whatever it can, including
play fast and loose with standards (or so I understand the U.S. Federal
court system to have determined).  It is so provincial that it makes
the canonical IBM blue suit of 25 years ago look like a Renaissance
man.  However, when it comes to actually doing something like writing
code or configuring a computer instead of instead of cranking out yet
another 1000's documents that no one reads (not to mention implements),
the IETF makes Microsoft look preceptive and responsive.


Vernon Schryver    vjs(_at_)rhyolite(_dot_)com



<Prev in Thread] Current Thread [Next in Thread>