Per Hurtig
Per Hurtig
The problem is related to the resolver. When avoiding to use the resolver (i.e., by using an IP-address directly), the problem goes away. For example (towards bsd10.fh-muenster.de): `./client_http_run_once -n 85...
Is this an issue related to the implementation or just a policy issue? Does Firefox favour IPv6 over IPv4?
When trying to recreate this, Firefox crashed right after startup. I can see in the debug information printed by NEAT that it opens a number of connections towards several mozilla...