CircleID

Brett,

I'm surprise no one added comments. Usually everyone has an "opinion" when it comes to spam. But maybe it is a tired subject already.

Nonetheless, I will provide my opinion.

To think out of the box, you must first get everyone in the box.

The basic overall problem with resolving the "spam problem" is that we have not updated the core SMTP process in 20+ years!  We are just adding "kludges" and inconsistent layers which explains why there are multiple methods, yet no universal and common method acceptable by all.

In my opinion, we need a mind set change in the "necessary folks", those who have a strong voice in molding the process since day one.  I might be bold to suggest, they need to retire and pass the torch to the next generation engineers.  The irony is that many of these folks are directly or indirectly associated with the "spam problems" we are trying to address.  So there is a vested interest in not having a 100% strong commitment to solving the problem at the core level and any attempt to do so is militantly resisted.  A preaching of "backward compatibility" is made, yet, these same people are advocating new designs that will required SMTP change.  They don't see it because they are not SMTP vendors.  We are an SMTP vendor. We see the proposals and the SMTP change requirements.  So lets just do it right once and for all.

The industry has been ready and primed for the 2nd coming and revamping of the telecomputing email communications system.  The first revamp during the late 80s/early 90s consolidated all the early proprietary systems to use the internet email system.  If we don't get a widely adopted solution to the problem, I'm afraid we will begin, if not already, to move towards proprietary and/or exclusive, and closed solutions isolated to certain larges groups only.

A "cost system" will most certainly begin a separation concept.  As a long time developer and producer of electronic mail/hosting products and solutions, there is no doubt in my mind it will further separate classifications of email users.

Of course, the question is, is this good?  Is this desirable.  Maybe it would be, maybe it won't. That all depends on how those businesses moving in this direction structure their operations.  I think it goes without saying that many see it as a new business model for profit, with an idealistic and more likely unrealistic perception it will reduce the problem.  How much will be the non-payee be at a disadvantage versus those that pay the extra cost?  Which also begs the rhetorical question if spam is a concern anymore for these businesses?  As long as the direct market pays the extra the cost of getting access to the the user inbox, the idea is that they have a higher priority now. Not lower.

Wat about the legalities?  What if a direct market is not getting his money worth?  What if the USER still says "BLOCK" this mail even if the marketer is paying the fees to the ESP/ISP?  Will the user have the option?  Will they have a choice?  Will the spam be wrapped instead as part of the EMAIL presentation and not actually part of the body? i.e, like GMAIL.

In the end, I believe what most people want is to reduce the malicious abuse at all levels, the system level and at the user level.

This can only happen, in my view, by taking a step back into the balcony and revisiting SMTP.  A great majority of the problem is NOT having the all actors to get "inside the box." By far, by industry measures, at least 80% or more of the "bad actors" are simply not 100% SMTP compliant.  Any SMTP vendor who has approached the problem, like Santronics has, by going back to SMTP basic, has realized an equal amount of the reduction in malicious abuse.

Once you enforce SMTP compliancy, then you can begin to talk about everything else. Whether it is wrapping it with a certification concept, trust layer or cost, it will still have the same problems if the compliancy issue is not addressed.

Hector Santos, CTO
Santronics Software, Inc.
http://www.santronics.com

Clear definitions are for people who want lines where none really exist.

One could envisage a weak approximation to a perfect spam system that still produced the desired outcome for all the emails a particular person received. i.e. that a perfect system can't exist in theory doesn't prevent a practical system attaining perfect results in a larger number of cases. Just as a chess computer can't yet play perfect chess, doesn't stop them beating the worlds best humans.

All that is needed is a classifier system that is recognisably better than the person whose email is being filter. Once you attain that level, I suspect the human will be accepting of the odd error. Although humans are very slow to recognise when the machines are better than them at something, animal pride.

Hector, for a lot of proposed antispam solutions, particularly those that argue that smtp must be replaced, I'd strongly encourage the proposers to read http://www.rhyolite.com/anti-spam/you-might-be.html

I've had my share of words with the EFF on their standpoint on spam filtering (a quick google search http://www.google.com/search?q=suresh+ramasubramanian+eff should give you plenty of material on this question, a lot of which could be repeated in any comment I'd make to Simon's quite interesting article ..)