On Tue, 31 Aug 2010 21:41:16 -0400, Eric Chadbourne wrote:
> > That would defeat the purpose of Freenet :b. Entirely. Imagine if
> > many people started doing this -- the data wouldn't spread properly,
> > and would fall off the network fast. We sacrificed latency for
> > redundancy (and security) from the beginning.
> 
> i hear what you're saying.  and i don't disagree for normal clients.
> but, let's pretend, that i wanted to upload ~10TB of video.  now.  i
> also want it to be quickly unavailable to others fall out of other
> folks data stores too soon.  so some sort of extended persistence
> would be wanted in the beginning.  what do you suggest?  i suspect
> the answer nothing is currently available and i must roll my own...

Yea, data persistence isn't that great at the moment. But you "hosting"
all the data just isn't how Freenet works. (Someone correct me if I'm
wrong.) Each node is intended to specialize around a tight keyspace --
I believe this is intrinsic to the routing algorithms. I'm not sure how
your "location" (every node has one, which is directly related to the
keys you're supposed to specialize in) would be handled if you hacked
the code to promiscuously handle any key. My guess is it would make the
topology of the network less efficient.


> so, ahh, what does the persistence column exactly mean in freenet?

Good question :b.
_______________________________________________
Support mailing list
Support@freenetproject.org
http://news.gmane.org/gmane.network.freenet.support
Unsubscribe at http://emu.freenetproject.org/cgi-bin/mailman/listinfo/support
Or mailto:support-requ...@freenetproject.org?subject=unsubscribe

Reply via email to