Conversation
Create constants for byte messages, implement asbytes so many methods can take Message and key objects directly and split get_string into get_text and get_binary. Also, change int handling to use mpint with a flag whenever the int is greater than 32 bits.
|
Can we expect a Python 3 compatible release of Paramiko before v2 @bitprophet? Apparently some people are successfully using it (see here and final comments on #16). |
|
Tested it while compiling a changed version of Fabric. +1 for Python 3 version |
|
Well, this pull request passes tests, but it breaks some tests for Here is a log for these tests (python 2.7): https://travis-ci.org/pashinin/fabric/jobs/16505811 I believe @scottkmaxwell can say something helpful. |
|
I haven't used |
|
What's blocking this from being merged? Anything I can help with? |
|
@pashinin I found the bug that caused tests to fail with Fabric. I just ran all of the tests with Python 2.7 and they succeed now. Added a new test to Paramiko for this, too. |
…-py25 Conflicts: paramiko/__init__.py setup.py tests/test_client.py
… we can skip starting server for test 5
|
Latest changes from master have been merged in. |
|
Thanks @scottkmaxwell. Now I have a repeated error from here: https://github.com/scottkmaxwell/paramiko/blob/master/paramiko/sftp_client.py#L581
I'm not sure what has a bug (my test code for |
|
@pashinin Any chance you can step thru the |
|
@dustinfarris @scottkmaxwell we are working on getting a new and updated site out first, to be able to communicate a roadmap etc. |
|
That's great news! Really looking forward to it. Let me know if I can help with either of those priorities. |
|
@scottkmaxwell Tested. The loop does not give exceptions. |
|
@pashinin How can I reproduce the problem? I'll have a look. |
|
I guess the problem is in my code. But if you want to help (I'm stuck) - see fabric/fabric#1050 (comment) |
…icated Key Exchange and user authentication
|
Coming to this thread without a lot of context*, I do see http://www.paramiko.org/ does mention a roadmap, but doesn't seem to actually have that roadmap. I guess that supports what @lndbrg posted earlier: that the next step is to communicate a roadmap, then merge in python 3 support? * I came to here after looking at where the open stack dependencies were sitting: https://gist.github.com/brettcannon/9009338 |
|
@robdennis There's a statement of @bitprophet about the integration of Python 3 support in #16:
|
All of the Py3 support changes updated to remove Python 2.5 support and its ugliness