Re: Proposed patch for external debugger access

[John Swensen]

John W. Eaton wrote:
> On 30-Oct-2007, John Swensen wrote:
>
> | John W. Eaton wrote:
> | > On 29-Oct-2007, John Swensen wrote:
> | >
> | > | I fixed up these suggestions also.  The patches are attached.
> | >
> | > | Index: debug.cc
> | > | ===================================================================
> | > | RCS file: /cvs/octave/src/debug.cc,v
> | > | retrieving revision 1.26
> | > | diff -r1.26 debug.cc
> | > | 22d21
> | > | < 
> | > | 29a29,30
> | > | > #include <set>
> | > | > 
> | >
> | > Thanks, can you please send a context diff?
> | >
> | > jwe
> | >
> | >   
> | Sorry, I forgot the -uN option to cvs diff this last time.  Attached are 
> | the context sensitive patches.
>
> I applied the patch with a few additional changes.  The most
> significant was to change the structure of the singleton class to
> match the way other singleton classes in Octave work.  I don't think I
> broke anything, but I didn't do extensive testing.  Can you please
> update and make sure everything still works as you expect?
>
> Thanks,
>
> jwe
>
>   
Thanks.  It is kindof nice to have finally made a more-than-trivial 
contribution to octave.  I also learned a lot about how you expect 
things to be done, coding styles, etc.  So hopefully my next 
contribution won't require so many iterations.  I will test the cvs 
version and get back to you.  Is the test directory reserved for certain 
types of tests, or should I write a test_debug.m file?  My next step is 
to allow a breakpoint structure to be passed into dbstop().

John Swensen