Open joshuap opened 3 years ago
this is related 👍 to #122
Addressed with #123
@Kelvin4664 @subzero10 I don't think this was addressed with #123, because this issue was to add local variables at the stack frame level in the traceback, whereas #123 added local variables for the top-level call. Is that right? If so, we should keep this open, assuming it's technically feasible.
@Kelvin4664 @subzero10 I don't think this was addressed with #123, because this issue was to add local variables at the stack frame level in the traceback, whereas #123 added local variables for the top-level call. Is that right? If so, we should keep this open, assuming it's technically feasible.
Oh, you are right! Good catch!
I believe Python makes method arguments available to each stack frame in the traceback. Is that correct? If so, I'd like to include the arguments in the Honeybadger backtrace, as we do for Elixir (see
filter_args
option).This feature should be enabled when a new
filter_args
is set tofalse
. The default should betrue
so that args are not included by default. We'll ship this as an opt-in feature in the next minor version; I may enable it by default in the next major version.Also, what other data is available in the traceback/stack frame that we could be including?
Resources
notice.error.backtrace[].args
in /v1/notices schema