[Python-Dev] PEP 553: Built-in debug()
Guido van Rossum
guido at python.org
Wed Sep 6 13:19:48 EDT 2017
99% of the time I use a debugger I use pdb.set_trace(). The pm() stuff is
typically useful for debugging small, simple programs only -- complex
programs likely hide the exception somewhere (after logging it) so there's
nothing for pdb.pm() to look at. I think Barry is wisely focusing on just
the ability to quickly and programmatically insert a breakpoint.
On Wed, Sep 6, 2017 at 10:00 AM, Nathaniel Smith <njs at pobox.com> wrote:
> On Wed, Sep 6, 2017 at 7:39 AM, Barry Warsaw <barry at python.org> wrote:
> >> On Tue, Sep 5, 2017 at 7:58 PM, Nathaniel Smith <njs at pobox.com> wrote:
> >> This would also avoid confusion with IPython's very
> >> useful debug magic:
> >> https://ipython.readthedocs.io/en/stable/interactive/
> >> and which might also be worth stealing for the builtin REPL.
> >> (Personally I use it way more often than set_trace().)
> > Interesting. I’m not an IPython user. Do you think its %debug magic
> would benefit from PEP 553?
> Not in particular. But if you're working on making debugger entry more
> discoverable/human-friendly, then providing a friendlier alias for the
> pdb.pm() semantics might be useful too?
> Actually, if you look at the pdb docs, the 3 ways of entering the
> debugger that merit demonstrations at the top of the manual page are:
> pdb.run("...code...") # "I want to debug this code"
> pdb.set_trace() # "break here"
> pdb.pm() # "wtf just happened?"
> The set_trace() name is particularly opaque, but if we're talking
> about adding a friendly debugger abstraction layer then I'd at least
> think about whether to make it cover all three of these.
> Nathaniel J. Smith -- https://vorpus.org
> Python-Dev mailing list
> Python-Dev at python.org
> Unsubscribe: https://mail.python.org/mailman/options/python-dev/
--Guido van Rossum (python.org/~guido)
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Python-Dev