Logged In: YES 
user_id=158581

Sorry for not responding earlier. As getting this was taking
a bit of time moving forward, in the interrum I had been
making much faster progress by splitting this off into it's
own little project under the bash debugger project I have on
sourceforge (http://bashdb.sourceforge.net/pydb). 

Being able to make public releases earlier and oftener has
offorded the oppertunity to move much further along much
quicker than via the patch method (where we still haven't
gotten past restarting the debugger).

Now back to your comments/questions. In pydb, there are two
kinds of "restart". "run" is like gdb's run. Debugger state
(breakpoints, watchpoints, display expressions, debugger
settings for list size, etc.) are preserved. Same as in gdb.

Gdb does not to my knowledge have a command called
"restart". However I've taken your suggestion of not having
a "restart" be the same thing as "run". "restart" is a
re-exec of the debugger. No state is save. As a debugger
writer, I personally find that one useful ;-)

Having the R be an alias for "run" is helpful since it's
short and matches what's used Perl where people use the
debugger more frequently. In fact there's a whole book
written on the Perl debugger. (And I also use 'R' in the GNU
Make and bash debuggers.)

As for "ru", well, since I've added command completion, I
feel this issue is less important.