-
Notifications
You must be signed in to change notification settings - Fork 244
New issue
Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? # to your account
Regression with IronPython on RPyC v4.1.0 #337
Labels
Bug
Confirmed bug
Comments
comrumino
added
the
To Start
Description reviewed and a maintainer needs "to start" triage
label
Jul 9, 2019
Thanks! I'll try to test this moving forward since I now have it configured. |
comrumino
added
Bug
Confirmed bug
and removed
To Start
Description reviewed and a maintainer needs "to start" triage
labels
Jul 14, 2019
YuvalEvron
pushed a commit
to weka/rpyc
that referenced
this issue
Oct 27, 2019
* Added warning to _remote_tb when the major version of local and remote mismatch (tomerfiliba-org#332) * Added `include_local_version` to DEFAULT_CONFIG to allow for configurable security controls (e.g. `include_local_traceback`) * Update readme.txt * Added break to client process loop when everything is dead * Increased chunk size to improve multi-client response time and throughput of large data tomerfiliba-org#329 * Improved test for response of client 1 while transferring a large amount of data to client 2 * Cleaned up coding style of test_service_pickle.py * Updated issue template * added vs code testing cfgs; updated gitignore venv * Changed settings.json to use env USERNAME * Name pack casted in _unbox to fix IronPython bug. Fixed tomerfiliba-org#337 * Fixed netref.class_factory id_pack usage per tomerfiliba-org#339 and added test cases * Added .readthedocs.yml and requirements to build * Make OneShotServer terminates after client connection ends * Added unit test for OneShotServer. Fixed tomerfiliba-org#343 * Fixed 2.6 backwards incompatibility for format syntax * Updated change log and bumped version --- 4.1.1 * Added support for chained connections which result in netref being passed to get_id_pack. Fixed tomerfiliba-org#346 * Added tests for get_id_pack * Added a test for issue tomerfiliba-org#346 * Corrected the connection used to inspect a netref * Refactored __cmp__ getattr * Extended rpyc over rpyc unit testing and removed port parameter from TestRestricted * Added comment explaining the inspect for intermediate proxy. Fixed tomerfiliba-org#346 * Improved docstring for serve_threaded to address when and when not to use the method. Done tomerfiliba-org#345 * Release 4.1.2 * Fixed versions referred to in security.rst * link docs instead of mitre * set up logging with a better formatter * fix bug when proxy context-manager is being exited with an exception (#1) * logging: add a rotating file log handler * fix bug when proxy context-manager is being exited with an exception (#1) * logging: add a rotating file log handler
# for free
to join this conversation on GitHub.
Already have an account?
# to comment
a63c27e re-introduces incompatibilities of RPyC with IronPython that were initially referenced in #10. The issue is that in IronPython,
str == unicode
, and the CPython side unboxes objects of that type tounicode
. This in turns leads to various problems in functions liketype()
that expectstr
objects.A simple fix is to replace this line with:
to bring back the behavior that was there before. I feel like it's safer to do the cast during unboxing rather than in
class_factory
, to prevent the id_pack to ever exist as two different types, but that's just a suggestion.This bug shows up when connecting IronPython 2.7.9 to CPython 2.7.16 on Windows 10 64bits, rpyc v4.1.0.
Server = IronPython:
Client = CPython:
Traceback:
The text was updated successfully, but these errors were encountered: