The Active Network
ActiveWin Anonymous | Create a User | Reviews | News | Forums | Advertise | VBA in Excel | Users Online: 0  
 

neowin.net

Amazon.com

  *  

  Microsoft contemplating charging for emails
Time: 05:24 EST/10:24 GMT | News Source: | Posted By: Bill Roach

MICROSOFT IS UNFOLDING something it calls the Penny Black project in which people sending emails might have to pay for the privilege. Microsoft claims in an article on its web sit that the "Penny Black" project wants to reduce spam by making senders pay. In the same article, Microsoft says it is contemplating different ways a sender might pay, including "plain old cash", CPU cycles, memory cycles and Turing tests.

Read Only Comments
Return to News
  Displaying Comments 1 through 5 of 5
  This is an archived static copy of ActiveWin.com.
#1 By 2459 (24.170.151.19) at Saturday, February 15, 2003 08:22:13 AM
This is basically research for a form of certified email:
This idea http://www.research.microsoft.com/research/sv/PennyBlack/cpu.html
http://www.research.microsoft.com/research/sv/PennyBlack/AntiSpam.ppt
is actually pretty good. Users on your recipients list get through without problems, while unsolicited mail is challenged. If the challenge isn't met, the unsolicited mail is bounced back to the sender. The main drawback is extra seconds of CPU time, but it proves a benefit to the recipient where unsolicited bulk emailing is concerned (i.e., makes unsolicited bulk emailing more time consuming/CPU-expensive).

Taken from MSR link:
For completeness: m should include a few extra pieces of information, such as the sender's and recipient's e-mail addresses, and a timestamp. This ensures that in a bulk mailing the function would need to be recomputed for every recipient, and that multiple transmissions to the same recipient will each incur the cost of a computation of f.

f is not amenable to amortization: the cost of computing f on many (even slightly) different messages m1, ..., mk is no less than the sum of the costs of computing f(m1), ..., f(mk). In other words, there is no ``volume discount.''

Specific families of moderately hard functions suitable for this technique are proposed in [DN92], where they are called pricing functions.

Each recipient chooses his own pricing function. In addition, the recipient may maintain a list of correspondents from whom he is willing to receive uncertified mail.

On receipt of a pair (m,y) from a sender not on the list of correspondents, the recipient's mail program checks whether m is timely and if y = f(m) (checking is cheap). If so, m is revealed to the recipient. Otherwise the mail program automatically "bounces" m back to the sender. One possibility is for the automatic reply to contain a program for computing f, say in C# managed code. Alternatively, the reply may contain a link to a web site from which the code can be obtained. Whether or not the sender must become involved manually at this point depends on whether or not the sender's mail program is compliant.



The Inquirer:
But we suspect any attempts by the Vole to introduce Penny Black would lead to a stampede for the nearest exit with many users seeing Penny Red, instead.* µ

* MAYBE NOT because it is such a bad idea, per se, but do we want Microsoft to administer it?


Microsoft Research:
Undoubtedly, there would in due course be an issue about such a service being run by any one organization. This should be fixable in the usual way, by having a Ticket Server Federation.The end-user of a ticket would choose to trust some particular incarnation of the ticket server, but it would be fine if there were other ticket servers too. The (small number of) tickets servers would trust each other because of corporate contractual agreements. A user could get a ticket from any ticket server and send it to any end-user. The end-user sends the UseTicket request to a server whom the user trusts, and that server forwards the request to the appropriate originating ticket server. There are of course some obvious optimizations of this flow, to avoid excessive message traffic.

Gotta love the Inquirer's lack of comprehension skills.

This post was edited by n4cer on Saturday, February 15, 2003 at 08:28.

#2 By 7721 (4.61.240.55) at Saturday, February 15, 2003 05:46:26 PM
This is yet another brilliant idea from Microsoft!! If implemented correctly so that the recipient need do nothing beyond a one-time selection of a pricing function, it would drastically reduce the amount of received junk mail with little cost to a typical user. I wouldn't mind for an outgoing e-mail I wrote to wait in my outbox for 30 seconds or so (I'll already be doing something else anyway), but imagine a spammer sending out 100,000 unsolicited messages! That's over 1 month of processing time! Goodbye, "accredited university diplomas in only 30 days!!!!"

Besides, I'm sure it would be possible (if not the default) to allow uncertified messages through if they're from addresses in the local address book. That way, you will get the mail you want quickly, but organizations that have something valuable to offer you and are willing to pay for it can still reach you without you having to worry about some important newletter being filter out by address book rules.

Hope this will make it into Palladium.

#3 By 135 (208.50.206.187) at Saturday, February 15, 2003 07:08:50 PM
Oh, this is an old idea. Bob Metcalfe was talking about pay for email back in like '96.

But the problem with spam is really just that SMTP is a poorly designed protocol with no authorization or authentication mechanism. Over the years people have tried to jerry-rig these into the protocol, but they aren't universal yet.

#4 By 6859 (12.219.23.70) at Sunday, February 16, 2003 02:26:59 PM
This is a gay story. Whatever....

#5 By 2960 (68.100.157.191) at Monday, February 17, 2003 09:43:07 AM
I think it's bogus too. Besides, why should _I_ have to pay because other people's spam?

Who would the money go to?

TL



 

  *  
  *   *
 
replica watches