)]}'
{"tox.ini":[{"author":{"_account_id":679,"name":"Kevin L. Mitchell","email":"klmitch@mit.edu","username":"klmitch"},"change_message_id":"226a9cd0b442c5f042f77b4e54006d38ae9beb9a","unresolved":false,"context_lines":[{"line_number":1,"context_line":"[tox]"},{"line_number":2,"context_line":"minversion \u003d 1.8"},{"line_number":3,"context_line":"envlist \u003d pep8,py27-constraints"},{"line_number":4,"context_line":"skipsdist \u003d True"},{"line_number":5,"context_line":""},{"line_number":6,"context_line":"[testenv]"}],"source_content_type":"text/x-properties","patch_set":1,"id":"da6ed579_0e839005","line":3,"updated":"2016-01-13 20:31:26.000000000","message":"1. I\u0027m not so certain about dropping py34 from this.  For one thing, I believe there is a known testr issue that demands that py34 be run before py27; for another, I believe we\u0027re expending the necessary effort to ensure that the py34 tests run, even if some have to be skipped because we\u0027re not fully compatible yet, and it would be nice for developers that have python 3.4 installed to see the output from that.\n\n2. I do like the idea of running pep8 before the other unit tests, although I am a little worried that its output may become obscured by the output from the py27 tests.","commit_id":"98188ee572cc7cab3206a94450dff9c5fd9e60bf"},{"author":{"_account_id":1,"name":"James E. Blair","email":"jim@acmegating.com","username":"corvus"},"change_message_id":"f225a9e74ff08164a917f42b5e2f1c75083bcd01","unresolved":false,"context_lines":[{"line_number":1,"context_line":"[tox]"},{"line_number":2,"context_line":"minversion \u003d 1.8"},{"line_number":3,"context_line":"envlist \u003d pep8,py27-constraints"},{"line_number":4,"context_line":"skipsdist \u003d True"},{"line_number":5,"context_line":""},{"line_number":6,"context_line":"[testenv]"}],"source_content_type":"text/x-properties","patch_set":1,"id":"da6ed579_91c68598","line":3,"updated":"2016-01-14 19:06:21.000000000","message":"I proposed this because it was conveyed to me that nobody just runs bare \u0027tox\u0027 because it does too many things, and that all the developers actually just run \"tox -e py27\".  This is an attempt to improve the situation, so that in the future, we might reasonably expect developers to just run \"tox\" and then that becomes a defacto standard of \"the set of tests we think should be run locally\" -- and that set can more easily mutate over time.  I personally think think that if the prevailing attitude among developers is to minimally test before submission to gerrit, then the more minimal this is the more useful it is.\n\nHowever, I have no skin in the game -- I\u0027m just trying to help out if I can.","commit_id":"98188ee572cc7cab3206a94450dff9c5fd9e60bf"},{"author":{"_account_id":679,"name":"Kevin L. Mitchell","email":"klmitch@mit.edu","username":"klmitch"},"change_message_id":"54c3f1714a76d0063ad9fceca0e2bba7d784860e","unresolved":false,"context_lines":[{"line_number":1,"context_line":"[tox]"},{"line_number":2,"context_line":"minversion \u003d 1.8"},{"line_number":3,"context_line":"envlist \u003d pep8,py27-constraints"},{"line_number":4,"context_line":"skipsdist \u003d True"},{"line_number":5,"context_line":""},{"line_number":6,"context_line":"[testenv]"}],"source_content_type":"text/x-properties","patch_set":1,"id":"da6ed579_2a587b87","line":3,"in_reply_to":"da6ed579_91c68598","updated":"2016-01-14 20:43:48.000000000","message":"Yep, I understand, and it\u0027s why I didn\u0027t vote on this change :)  I wanted to see the perspective of some other reviewers before I decided how I wanted to vote on this one :)  (And incidentally, my typical tox usage almost always involves explicitly specifying the environments I want to run, even on personal projects where I\u0027m the sole developer…)\n\nMy thinking is that, given the new feature in tox that allows it to skip environments for which it doesn\u0027t have the correct Python interpreter, we don\u0027t need to drop the py34 tests from the envlist, although we\u0027d need to set skip_missing_interpreters to True.  Beyond that, my only other concern is the pep8 output getting obscured by the test output, and I think that\u0027s a pretty minor concern.","commit_id":"98188ee572cc7cab3206a94450dff9c5fd9e60bf"}]}