On Sat, Aug 29, 2009 at 12:03 PM, Matthew
Toseland<toad at amphibian.dyndns.org> wrote:
> On Saturday 29 August 2009 00:59:55 Michael Yip wrote:
>> By your saying then that means the key is closest to the data source
>> node then?
>
> Yes, there appears to have been some confusion. When you request the data, it 
> will come from one node, which we can call the data source, as distinct from 
> the request source. If the data is popular, it will be in many nodes' caches, 
> but hopefully for less popular content it will come from the store of a node 
> whose location is close to key.

In case there is any confusion: there is no particular relationship
between the node where the data was stored (the "data source") and the
person / node that originally inserted the data.

Evan Daniel

>>
>> Evan Daniel wrote:
>> > On Fri, Aug 28, 2009 at 7:03 PM, Michael Yip<mhy831 at cs.bham.ac.uk> 
>> > wrote:
>> >
>> >> Wait, is there a difference between fetching the key and fetching the 
>> >> data?
>> >>
>> >
>> > No, just imprecise terminology. ?You start with the key and fetch the
>> > data it refers to.
>> >
>> > Evan Daniel
>> >
>> >
>> >> Evan Daniel wrote:
>> >>
>> >>> On Fri, Aug 28, 2009 at 6:46 PM, Michael Yip<mhy831 at cs.bham.ac.uk> 
>> >>> wrote:
>> >>>
>> >>>
>> >>>> But then how come the routing is done by choosing the node whose
>> >>>> location offers shortest distance to the key?
>> >>>>
>> >>>>
>> >>> Because that's how you find the key. ?Keys are expected to be stored
>> >>> near their location, so you look for them near their location.
>> >>>
>> >>> Evan Daniel
>> >>>
>> >>>
>> >>>
>> >>>> Martin Scheffler wrote:
>> >>>>
>> >>>>
>> >>>>> Am Freitag, 28. August 2009 08:48:09 schrieb Michael Yip:
>> >>>>>
>> >>>>>
>> >>>>>
>> >>>>>> Just a simple question:
>> >>>>>>
>> >>>>>> Does it mean that the key would be closest to the location of the data
>> >>>>>> source node?
>> >>>>>>
>> >>>>>> ...
>> >>>>>>
>> >>>>>>
>> >>>>>>
>> >>>>> no, the key is determined by the content.
>> >>>>> the routing tries to push the data into the right direction where it 
>> >>>>> can
>> >>>>> be
>> >>>>> found later.
>> >>>>> all the keys that the source node inserts are going different ways.
>> >>>>> if your node is neighbour of the source node you will only see some 
>> >>>>> keys
>> >>>>> from
>> >>>>> the insert which are in your keyspace.
>> >>>>> and even when your "cancer nodes" dominate the routing table of the
>> >>>>> source
>> >>>>> node, you can't know the content until you have the URI to access it.
>> >>>>>
>> >>>>> good byte
>
> _______________________________________________
> Devl mailing list
> Devl at freenetproject.org
> http://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
>

Reply via email to