[m-dev.] for review: new debugger command set, part 1

[Zoltan Somogyi]

In response to my changes Erwan wrote:
> > step [-NSans] [<num>]
> >         If this command is given at event <cur>, continues execution until
> >         event <cur> + <num>. The default value of <num> is 1.
> 
> In opium at Prolog, this command is called next. So that's the name I chose
> for my opium at mercury. Maybe it would be a good idea to keep the same
> name for commands that are doing the same thing in both debuggers. 
> So, if you really like "step" I can change it.
> 
> > finish [-NSans]
> >         Continues execution until it reaches a final (EXIT or FAIL) port
> >         of the call to which the current event refers. Reports an error
> >         if the current event refers to a final port.
> 
> I thougth we agreed to call that command "skip" ?

Our plan always was to implement an alias facility, so the precise names of
commands do not matter *too* much. Although we agree that being compatible is
nice, there are too many command languages to be compatible with: several
Prolog variants as well as gdb, dbx etc.

The first command above is called "step" rather than next because what it
does is similar to what step does in gdb, and quite different from what next
does in gdb. The second command is called "finish" because that is the name
of the equivalent command in gdb.

> By the way, I've got some problems with the current cvs release of the
> mercury compiler.
> 
> Am I doing something wrong ?
> ...
> It seems that library/builtin.m and the old library/mercury_buitin.m are
> in conflict. 

mercury_builtin.m was divided into builtin.m and private_builtin.m, and the
compiler was modified to refer to these modules instead of mercury_builtin.m.

Fergus, can you please tell Erwan exactly what he need to do?

Zoltan.