bfox at ua.com said: > Perhaps he can't read that header file without "contaminating" his > clean-room implementation? I don't think we can argue that the node layout on the medium, as described in that file, could possibly be considered sufficient to 'contaminate' an alternative implementation. -- dwmw2