http://www.dslreports.com/forum/r20468945 en Tue, 08 Dec 2009 21:25:57 EDT Tue, 08 Dec 2009 21:25:57 EDT http://www.dslreports.com/forum/remark,20469070 sded : Thanks a lot; that helps clear up confusion in several places that use different terminology.]]> http://www.dslreports.com/forum/remark,20469070 Mon, 12 May 2008 10:33:03 EDT http://www.dslreports.com/forum/remark,20469032 dave : Neither of those processes run user-mode images, and generally speaking, the "name" displayed for a process is just the name of the image file it's running. So display tools have to make up names: "System" and "System Process" are not good made-up names.

Process Explorer uses "System Idle Process" for pid 0 and "System" for pid 4, which is much better.

Operating systems have to be running some process; the idle process is what gets run when there's nothing else ready to run (there is in fact one thread per processor in the idle process).

The system process is a process that acts as a container for threads created by kernel-mode components - memory management components, drivers, etc.

I don't believe that the idle process (pid 0) ever has any connections. Rather, I think that certain TCP control blocks have their 'ownership' field zeroed out on process termination, meaning no current owner, and TCPview misinterprets that as 'owned by pid 0'.]]> http://www.dslreports.com/forum/remark,20469032 Mon, 12 May 2008 10:23:02 EDT http://www.dslreports.com/forum/remark,20468945 sded : They both appear in tcpview, System Process at PID0 showing a lot of connections (kernel mode threads?), System at PID4 showing mostly netbios interaction.]]> http://www.dslreports.com/forum/remark,20468945 Mon, 12 May 2008 10:03:42 EDT