> > What is planned for NG routing is to normalize the data size.   In other words
> > we will ask how long will it take to get data of a constant size from a node,
> > and update the routing info with the normalized results.
> 
> Well, not a strictly fixed size. We calculate the average size from the
> datastore and use that, at run time. It should not drift quickly over
> time, so the measure is reusable.

I suspect Ed was getting to that part ;-)  My only concern here is if nodes are 
sharing their RTEs (as I think they should so that nodes are "born" with a 
reasonable network knowledge) that problems could occur if there are significant 
node-to-node differences in average data size in the datastore.  We might want 
to check to see if there are large variations today.

Ian.

-- 
Ian Clarke                                                  [EMAIL PROTECTED]
Coordinator, The Freenet Project              http://freenetproject.org/
Founder, Locutus                                        http://locut.us/
Personal Homepage                                   http://locut.us/ian/

Attachment: pgp00000.pgp
Description: PGP signature

Reply via email to