At 09:54 PM 4/21/2002 -0700, Craig R Hughes wrote: >Jim Paris wrote: > >JP> Sure, the numbers can be debated (and I'm not really interested in >JP> doing so). But you will at least agree, I hope, that your hashcash >JP> solution will slow down spammers _only_ if they happen to be sending >JP> more than 28,800 per day. (Or 10,000 a day depending on computer >JP> speed, whatever). > >That last part also starts to get at the fact that you are battling >Moore's Law, >which IMO is not smart.
all cryptography is battling Moore's Law. Why do you think that DES is no longer safe to use for general-purpose encryption? Why do you think there is a big flap over whether or not 1024 bit RSA keys are still safe? The trick is not to try and figure out what will be good forever but instead figure out what attacks you are vulnerable against and develop a plan for change when those attacks become somewhat possible. The camram coin technique is vulnerable to Moore's Law as well as hardware based accelerators. This is the classic weakness of proof of work puzzles. We are searching for a puzzle that is harder to implement in hardware as well as being somewhat less susceptible to Moore's Law. We think we've found one and part of the way it makes things difficult is that consumes a megabyte of memory which increases the odds for it being flushed from the cache and forcing the CPU to run at memory bus speeds. Will it work out in practice? It's hard to tell. We will find out more with experimentation. in any case, if one is going to be successful in fighting spam, one needs to look at solutions other than filtering, blacklisting network connections and ports, and legislation especially if one wants to preserve end to end connectivity and the ability for any node on Internet to participate fully. My experimentation with camram is an attempt to provide another part of the solution. ---eric _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
