On Sat, Jun 14, 2003 at 12:10:31AM -0500, Tom Kaitchuck wrote: > On Friday 13 June 2003 10:22 pm, fish wrote: > > I just don't think that they should be >1meg, > > Uh, just to clarify you mean >1meg right? > > I agree with you. It doesn't seem to me that there are many nodes accepting > blocks >1meg anyway. If you are combinning it into a container only to break > it up into chunks I don't see the gain. > > However this could be implemented at the application level by combining all > the SMALL images (aside from the sites thumbnail) into one file. But leave > the HTML alone this way each page is still independently referenceable and > unnecessary redundancy is eliminated. > > A trick that might be nice would be to to try to get all the small images to > combine to a nice even size. That way you pad less. > > Also, your comment about size and low bandwith users got me thinking: Is their > some way to have nodes bias the size of the data they store biased on their > available banwidth. And if so how could routing, load distribution, and > utilization of low bandwith nodes be improved biased on this?
No, that information is not available to routing. If large chunks are a problem we should impose a limit - we do not want only a few nodes to cache a really large chunk. > > Anyway that is the start of a whole new thread, so I'll post it to tech@ once > I think it through a little more. -- Matthew J Toseland - [EMAIL PROTECTED] Freenet Project Official Codemonkey - http://freenetproject.org/ GPG key lost in last few weeks, new key on keyservers ICTHUS - Nothing is impossible. Our Boss says so.
pgp00000.pgp
Description: PGP signature
