3 ms·
It's not, they did, and I don't think they were. Quote from article: > But the problem goes beyond trusting Mega’s servers, says cryptographer Moxie Marlinspik
by haldean 14y ago
It's not, they did, and I don't think they were. Quote from article:
> But the problem goes beyond trusting Mega’s servers, says cryptographer Moxie Marlinspike, who recently left a position as head of product security for Twitter. Browser-based cryptography also means that anyone who breaks the SSL encryption that protects the data sent by the server could tamper with the code just as easily.
> “The security of Javascript-based cryptography is reducible to the security of the channel in which the Javascript is transmitted (every time you visit the page), which in this case is an SSL connection,” Marlinspike writes to me in an email. “If you can break SSL, you can break MEGA.”
Source: http://www.forbes.com/sites/andygreenberg/2013/01/21/researchers-warn-megas-new-encrypted-cloud-cant-keep-its-megasecurity-promises/ http://www.forbes.com/sites/andygreenberg/2013/01/21/researc...
- AnthonyMouse 14y agoMeh. He's not complaining about Mega, he's complaining about SSL. He's always complaining about it, and he's mostly right. There are legitimate problems with it. In particular, too many questionable entities are registered as certificate authorities in common browsers, so you periodically see these imbeciles issuing wildcard certificates for * .google.com or * .gov which totally compromises the ability of browser-based SSL to prevent man in the middle attacks against anyone who can socially engineer or coerce their way into one of those certificates from any one of the inept certificate authorities. And actually it's not a bad point for a service like this, in the sense that browser-based SSL will never protect you against a government who is trying to read your secrets, because the government can always coerce a local certificate authority into giving it a certificate for whatever site it wants to impersonate. The only way to prevent that is to remove all the common certificate authorities, which doesn't actually improve security at all (it really makes it worse under normal usage patterns), but it reduces your inclination to have a false sense of security. So it's a valid point: If you need really good security, especially against governments, javascript-based password decryption is a fail, not least because it depends on SSL. But those people may still be able to use the service, just not the web interface to it, and I kind of expect they'll be a minority of the users in any event.