[freenet-dev] Bloom filters and premix was Re: Possibly interesting paper on Freenet

[Matthew Toseland]

On Thursday 22 November 2007 14:54, Michael Rogers wrote:
> Matthew Toseland wrote:
> > OTOH if it's not 
> > cellular you can maybe triangulate the requestor.
> 
> How about requesting the entire splitfile through a single premix
> tunnel? That prevents triangulation without requiring cells (or rather,
> the triangulation only reveals the far end of the tunnel).

It's also slow, and unreliable, to the point that we'd have to use multiple 
tunnels, which would defeat the object of using a single tunnel. Also, there 
may be an obvious connection between multiple splitfiles, or multiple 
non-split files (e.g. frost posts from the same ID).
> 
> The nodes participating in the tunnel, including the requester, would
> pass the tunneled data along without caching or storing it. So it would
> be impossible to identify tunnel participants (including the requester)
> from their Bloom filters.

Indeed, that's why premix makes Bloom filters safe.
> 
> Cheers,
> Michael
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://emu.freenetproject.org/pipermail/devl/attachments/20071122/7bbd61f6/attachment.pgp