What you want to do is what I tried to do at the end of a several weeks long sneakernet series. Work fetched by one host [a duo], copied via a USB stick [complete BOINC data dir] and crunch on another [8 core] that was not set up yet to internet. When it was, and did connect [sitting in an airport lounge], the completed / reported tasks were taken on the first connect [CEP2 jobs going to Harvard, but fetched from Toronto], then the WCG server figured out the host was an imposter duplicate on different hardware and WCG did the server aborted for the remaining tasks [the server determined the tasks were never assigned to that host]. So here I managed, one host under one ID fetched jobs and it got logged on the RS pages as reported by a different device (different numeric host ID). Like Vasili had "one ping only", you have one connect chance to fool the servers [it still thinking the old host is reporting]. Message being: Tasks fetched by 1 device must be reported by the same device or they will get binned.

--//--

It was connected until recently. When I try to ping the gateway I get the message "Destination host unreachable." The PC is directly connected to the gateway via Ethernet and can access the configuration software. All other PCs connect via wireless adapters and can talk to each other and access the internet.

RESOLVED - I turned off my fire wall and all WUs uploaded and even downloaded more. Firewall is back on while the PC crunches and I get it fixed.

Thanks everybody for getting me to think.