[RFA/RFC] Report the original location specification for a breakpoint.

[Nick Roberts]

 > > Isn't this going to affect more than MI?  I thought the text from
 > > ui_out_field_string was displayed for console too, just without the
 > > label.
 > 
 > I don't this so, but I'll check. We can always make this field output
 > just for MI.

It's used by "info breakpoints".
 
 > > This definitely needs documentation to go in.  I can't say this
 > > enough.  Undocumented fields in the MI output might as well not exist.

There are already many other fields that aren't documented, e.g.,
pending, what, cond, ignore etc.  In fact the MI output of -break-insert
is very variable.

 > Yes, of course.
 > 
 > > Beyond those, I don't see a problem with it.  Though wouldn't we like
 > > to display this in the CLI sometimes too?
 > 
 > I have no idea. I'd much rather have somebody else decide what CLI should
 > contain.
 > 
 > - Volodya
 > 
 > 
 > 

-- 
Nick                                           http://www.inet.net.nz/~nickrob