Hello? Is this thing on?

[Vinay Gupta]

Hi, just wondering if anybody's got the software set up and how they are finding it so far?

V>

[Tony Arcieri]

It's a bit too early for that I'm afraid ;) I will add a project status to the README

On Wed, Aug 1, 2012 at 6:48 PM, Vinay Gupta <hexa...@gmail.com> wrote:

Hi, just wondering if anybody's got the software set up and how they are finding it so far?

V>

--
You received this message because you are subscribed to the Google Groups "cryptosphere" group.
To post to this group, send email to crypto...@googlegroups.com.
To unsubscribe from this group, send email to cryptosphere...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msg/cryptosphere/-/8mGvOBEihikJ.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
Tony Arcieri

[Vinay Gupta]

Outline looks good. Have you considered using shamir's secret sharing (probably the http://manpages.ubuntu.com/manpages/precise/man7/gfshare.7.html version?) as a way of addressing some of the abuse problems? Like mebbe yer True Name is split among your Peers, and (because all traffic is signed?) uploading Evil can result in collusion of your peers to out you, something like that?

We need these protocols to have courts. It's the only way to keep the CP and other unreasonable stuff off them. But then, what's the Law? Well, one Law per network, perhaps, akin to nations...

Such are my thoughts.

http://guptaoption.com/cheapid

Vinay

[Tony Arcieri]

On Thu, Aug 2, 2012 at 9:56 PM, Vinay Gupta <hexa...@gmail.com> wrote:

Outline looks good. Have you considered using shamir's secret sharing (probably the http://manpages.ubuntu.com/manpages/precise/man7/gfshare.7.html version?) as a way of addressing some of the abuse problems? Like mebbe yer True Name is split among your Peers, and (because all traffic is signed?) uploading Evil can result in collusion of your peers to out you, something like that?

In general I think a capability-based security model is a great way to handle keeping everyone in their own sandbox. It'd be possible to use Shamir + secure multiparty computation to provide a  multiparty signature system that needs consensus among a group. That's definitely not in the MVP, though.

 

--
Tony Arcieri

[Johnson]

what are your plans for combating child porn, other than IP-banning? i definitely don't want to host child porn, it's bad enough that so much of it goes through Tor.

[Tony Arcieri]

On Sat, Sep 29, 2012 at 8:45 PM, Johnson <ben...@gmail.com> wrote:

what are your plans for combating child porn, other than IP-banning? i definitely don't want to host child porn, it's bad enough that so much of it goes through Tor.

It's possible to have a blocklist of IDs for offending content. However, that same list of IDs can serve as a map to offending content. However, it would, at the very least, prevent hosts honoring the content blocklist from inadvertently storing that content (of course any trivial change to a file could be used to evade the blocklist)

--
Tony Arcieri

[Carlos Agarie]

Hey Tony, how's the project going? I'm studying cryptography by myself and would love to help when I get "good enough".

Also, maybe you're too occupied to be able to work on it for now, but it'd be great if you put some of the tasks that need to be completed.

[Tony Arcieri]

One of the downsides of choosing to implement a project like this in Ruby is that there aren't exactly a lot of Ruby bindings to modern cryptographic algorithms. Even ECDSA isn't usable on JRuby.

That said, I've been working on some bindings for the algorithms I want to use. The first one is Ed25519, an elliptic curve-based digital signature algorithm:

https://github.com/tarcieri/red25519

I've also been working on a similar but unrelated project which is a capability-based credential manager:

https://github.com/livingsocial/keyspace

I figure if I can't get that right I have no business trying to write the Cryptosphere ;)

There's been a lot of cross-pollination between the two projects. I've been reimplementing the various cryptographic underpinnings over and over, refining them each time.

The next step is ripping out the existing signature algorithms (the existing code in the Cryptosphere uses RSA in a similar manner to Tahoe. Keyspace uses ECDSA) and replacing it with Ed25519.

--
You received this message because you are subscribed to the Google Groups "cryptosphere" group.
To post to this group, send email to crypto...@googlegroups.com.
To unsubscribe from this group, send email to cryptosphere...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msg/cryptosphere/-/8N0W-njlMV0J.

For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
Tony Arcieri