On Thu, Jan 09, 2003 at 07:23:55AM -0500, Ed Tomlinson wrote:
> The htl graph posted on cruft now and then shows that currently
> a htl of 19 gets most sites and that 25 is rarely necessary.  
> 
> A side effect of ARKs should be improved specialisation as nodes
> using dynamic naming services will keep their identities on freenet.
Yeah, but if the network is five times bigger? ARKs should in the not
too distant future lead to many transient nodes becoming permanent ones,
and many new nodes being permanent rather than transient. And the
network will continue to grow, and the simulations suggest that it
scales at roughly O(n^0.28), and we really don't want to have to
increase the HTL right at the last minute as we put out 1.0-pre1 (no,
this is not likely in the near future... but the point is it's better if
we increase it earlier rather than later...). However, maybe you have a
point. It probably makes sense to defer this for 0.5.2 and see what
happens (other routing-related changes will probably go into 0.5.2 -
multiplexed connections probably, probabilistic caching when the
datastore is 90%+ full, etc).
> 
> I agree with gj - think we can leave this alone.
> 
> Ed Tomlinson

-- 
Matthew Toseland
toad at amphibian.dyndns.org/amphibian at users.sourceforge.net
Freenet/Coldstore open source hacker.
Contracted full time by Freenet Project Inc.
http://freenetproject.org/
ICTHUS.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: 
<https://emu.freenetproject.org/pipermail/devl/attachments/20030109/41c7e9ad/attachment.pgp>

Reply via email to