First, I want to thank you all for all the helpful responses.
On second (or third) thought, perhaps the problem of "local looping"
offline PCs is going to have to be resolved anyway because the debugger
needs a socket connection to talk to the jvm running the application being
debugged. If this is the case, we should stick to the socket connection
between Emacs and the debugger as well, i.e., we end up with something like
this:
Emacs/JDE <-socket-> debugger <-socket-> application
for debugger communications and
Emacs/JDE <-standard I/O-> application
for user command-line interaction with the application under test.
I had tried the hosts file solution and it did not work for me. However,
maybe this is because I had
127.0.0.1 homepc
in my hosts file, not
127.0.0.1 homepc localhost
as the NT/Emacs FAQ advises. When I did this and invoked (using the
BeanShell, what a godsend it is turning out to be!)
java.net.InetAddress.getLocalHost(), I got an exception "homepc is
unknown". I will try the FAQ solution at home tonight. If any of you have
gotten this solution to work on a Windows 95 system, I'd like to hear from you.
Regards,
Paul
