RSS
Twitter
Microsoft Research Offers Peek At The Future
from the useful-or-not? dept
Whether you're a fan of Microsoft or not, they do appear to be doing some interesting research in their labs, including a potential spam stopper. They're testing out a system that would force any email exchange to "solve a cryptographic puzzle" for any sender not in the recipients addressbook (or whitelist, I assume), putting some strain on the CPU. For spammers sending out millions of messages, it would drastically slow down their ability to send out the messages. At least that's the theory. There isn't much in the way of details, so if anyone knows any more, I'd like to hear about it. The other research talked about isn't as exciting, and includes things we've already spoken about, like their system to record everything that happens in your life. I still wonder, if you record your entire life, when will you have time to replay it?
Reader Comments
(Flattened / Threaded)
No Subject Given
I still wonder, if you record your entire life, when will you have time to replay it?
I can't help but think of Spaceballs...
"What the hell am I looking at? When does this happen in my life?"
"Now. You're looking at now, sir. Everything that happens now, is happening now."
"What happened to then?"
"We passed then."
"When?"
"Just now. We're at now, now."
...
(reply to this comment) (link to this comment)
"not much in the way of details"
In other words, the usual meaningless hype.
Why waste space on this?
(reply to this comment) (link to this comment)
Similar countermethod discussed for "post office a
In Bruce Schneier's most recent (and as always excellent) Crypto-Gram, a "post office attack" was described which had what appears to be a similar countermethod as the Microsoft anti-spam method mentioned. Here is the issue (it is the lead article): http://www.counterpane.com/crypto-gram-0304.html If you follow the link for the "scripted attacks" paper at the end of the first article (URL pasted below) you will find that in section 5.4 it walks you through one possible "client puzzle". These wouldn't work nearly as well for email since the sending server and the receiving server almost never communicate directly. But I assume any proposed solution would require that the receiving server be able to open a direct communicaton with the sending server once an email was received in order to initiate the puzzle. http://www.avirubin.com/scripted.attacks.pdf Seems like you'd have to work out how to eliminate IP spoofing in email for it to work (not likely) or at least have servers keep an index of what they sent (not easy) ... possibly explaining why it is still in research.
(reply to this comment) (link to this comment)
recording everything would be a boon
You don't record everything for the sake of playing it all back. You record everything so that next week, when you ask "whom did I run into on the way to lunch last thursday?", you can look back and find out.
BTW the solve-a-puzzle spam idea isn't new.
(reply to this comment) (link to this comment)
Add Your Comment