Il giorno ven, 23/03/2012 alle 09.41 +1000, James A. Donald ha scritto:
[cut]
From time to time you check that stuff that is stored on someone else's
system is available by random sampling.
When stuff stored on someone else's system becomes inaccessible, you
note him as unreliable, and thus decline to issue IOUs or services to
him in future, though since IOUs are transferable, you are still up for
any IOUs you have issued to him in the past.
On 3/25/2012 8:25 PM, danimoth wrote:
I could accept x bytes, so someone is granting me IOUs for that. I throw
all things on the trash, and meanwhile (before someone mark me as
unreliable, which needs some time) I could spend all IOUs I earned
before, which will be valid (regardless my unreliability) in the future
for the other peers. A cycle could be closed, if time to find me is too
high (or time to close a cycle too low). So, someone other will pay my
lies.
Let us call a key that has a reputation for delivering, a nym key, and
an IOU for such a key a nym IOU, that is to say, a nym with reputation.
A nym without reputation, we will call a no nym, anonym.
For the system to work, there needs to be lots of nym keys and nym IOUs.
Problem is that all keys and IOUs starts out as no nym IOUs.
If you have a no nym key, then when you provide storage, will probably
get no nym IOUs, which will frequently turn out to be worthless - your
client will have provided real storage for worthless promises of
storage. But in so doing, eventually develops a nym, and so will see
demand for its storage, enabling it to demand name IOUs
In creating reputation, will frequently provide real store for worthless
store, and in discovering reputation, will frequently store pseudo
random data on other people's systems, since storing valuable data on
nym storage is probably a bad idea.
_______________________________________________
p2p-hackers mailing list
p2p-hackers@lists.zooko.com
http://lists.zooko.com/mailman/listinfo/p2p-hackers
