Hello
I'd like to be able to set breakpoints (in interpreted code, at least).
What I find useful for debuggers to offer:
- being able to set breakpoints also in files for which I don't have the permission for modifications.
- being able to remember the selected breakpoints over source code changes (like reloads of the source file into the running system); but maybe emacs can handle this by itself.
Probably the position of the breakpoints will be given by line/column numbers in the source file (through an emacs mode or similar). Or maybe S-EXPR related positioning (SXPath?) can be a good alternative.
What would be the steps necessary to achieve this? I may help in the implementation or in researching the best ways for handling of breakpoints from emacs.
BTW can compiled Gambit code be sensibly debugged with a stepper? Using GDB? Or maybe by using the portable scheme debugger? Or inserting some other trickery like some form of inserted checkpoints which stop execution if some flag is set? This may especially turn out to be useful in connection with continuation serialization, since serializing continuations is only really practical for compiled code because of their size; so a system which is depending on serialization will want to have all code be compiled at least in live usage, and it would be useful to be able to debug live systems too.
Thanks Christian.