On Wed, Jan 14, 2004 at 11:43:29PM +0100, [EMAIL PROTECTED] wrote: > >> then wrong noderefs can only slip into the rt when the local rt already h= > >as these bad references (e.g. when switching von unstable->stable or the ot= > >her way around without purging the rt) or when importing new=20 > >> seednodes which contain these bad refs. > >>=20 > >> would it be useful if the local routing table would be checked on node st= > >artup if every ref is suited for the build? if there is such a check when a= > >dding refs while the node is running, then this check logic would be=20 > >> reusable at startup time, wouldn't it? > > > >It is. > > errr.. then i just don't get it.... > > when the node starts up, the rt gets cleaned, so no bad refs are left. > then, while the node is running only good refs are added to the rt. > > then why are connections from incompatible (e.g. stable/unstable) nodes in the OCM? > because they contacted us? > then why aren't they rejected for using an incompatible protocol?
Their queries are rejected. Their connections are not rejected because we don't exchange that info at connection time. We don't close the connections because they'd only retry. > > sorry for bothering you and flooding the list, but i imagine i've seen some > confusion on the list too..... well.. dev that is.... -- Matthew J Toseland - [EMAIL PROTECTED] Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so.
signature.asc
Description: Digital signature
_______________________________________________ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support
