On 2014-11-19 at 18:17, Robert J. Hansen wrote: N>> I agree with several other important points you raise, but this one is not a big >> deal. I have a highly customized mail setup. My SpamAssassin downloads rules >> from the internet, but trains its Bayesian filter on only the e-mail I >> personally receive. > > I don't mean to sound like I'm dismissing your experience, because -- > well -- your experience shouldn't be dismissed. (Nobody's should.) > But I do think you might be overlooking something: you already > experience a significant benefit from the aggressive, God's-eye-view > anti-spam efforts of Google, Yahoo!, Microsoft, and more. The things > they do for their users have a ripple effect in making your own > anti-spam fight a little easier. > > A couple of months ago Mike Hearn wrote a brilliant treatise on > end-to-end cryptography and anti-spam technologies, with a long > digression on how anti-spam technologies work at Google. It's worth > every second it takes to read. > > https://moderncrypto.org/mail-archive/messaging/2014/000780.html
He’s mainly explaining how do you fight spam in a centralized way, and then explain how all the centralized techiques are unusable when using crypto. That’s normal, crypto and decentralization comes together. You need to think according other paradigms. It’s like when you live in society. You can either think the autoritarian way “if I were the Great King Controlling Everything what could I do to fix the problem?”, or the social/free way “what should I do so that if everybody did like me the problem would get fixed?”. So that involves way much complex maths (well, actually, *different*: in the centralized world it’s already really complex, but the complexity you need to decentralize is compensated by the local private data you can access and the crypto techniques you become used to), DHTs, meshes, crypto, symmetric communication, political thought, users education, etc. I don’t consider that an issue. Quite the opposite: the result —and we always end finding it— is *beautifull*. It’s like admiring the almost perfectness of the way human body chemical biology works. It’s like admiring a fractal. You just end with something approaching what you observe within organic structures, something more resilient, perennial, big, free, flexible… Also he speaks about using bitcoin, which is not a good point bitcoin not being really secure: you just need more computational power than the half of the network and you can takeover it. Big government can do it. Also bitcoin needs anyway a lot of computational power, worse, it *encourage* it by competition. That’s really catastrophic ecologically. And finally it suffers from the problem of globalizing everything, contrarily to the Internet (and GNUnet) historical architecture where everything is the most local possible (within the Internet only IP attribution and DNS are global, within GNUnet *nothing* is, so you could transparently divide, join and grow GNUnets without any problem). Yet proof-of-work can be effectively used to prevent abuse. GNUnet use it to prevent spamming its global DHT with lot of revok’ certs it will store for a while. It could be made on messages if we didn’t need a certain fastness (merging all asynchronous communication means even microblogging will have the same requirements) and we didn’t already had concepts of mesh, WoT, bayesian filtering, F2F and cryptographic signature.
signature.asc
Description: PGP signature
_______________________________________________ Gnupg-users mailing list [email protected] http://lists.gnupg.org/mailman/listinfo/gnupg-users
