[issue4951] failure in test_httpservers

2009-04-25 Thread Tarek Ziadé
Tarek Ziadé added the comment: Applied in r71878 and r71879. Thanks EnvironmentVarGuard is definitely the way to go, I'll use it. -- status: open -> closed ___ Python tracker __

[issue4951] failure in test_httpservers

2009-04-25 Thread Mark Dickinson
Mark Dickinson added the comment: EnvironmentVarGuard certainly looks like a good thing to use here, provided that it works... -- ___ Python tracker ___

[issue4951] failure in test_httpservers

2009-04-25 Thread Walter Dörwald
Walter Dörwald added the comment: Hmm, EnvironmentVarGuard seems to be broken: import os from test import support with support.EnvironmentVarGuard() as env: env.unset("HOME") env.set("HOME", "bar") print(os.environ.get("HOME")) I would have expected that to print the original HOME vari

[issue4951] failure in test_httpservers

2009-04-25 Thread Walter Dörwald
Walter Dörwald added the comment: There's an EnvironmentVarGuard context manager in support.py that IMHO should be used for recording changes to the environment variables. Or a new context manager that does what your patch does could be put into support.py. There might be other tests that benefi

[issue4951] failure in test_httpservers

2009-04-24 Thread Mark Dickinson
Mark Dickinson added the comment: Okay, I think I've got this figured out. It's a problem in the distutils tests: the mechanism for saving and restoring environment variables in Lib/distutils/tests/test_util.py doesn't work: after the line self.environ = os.environ in the setUp me

[issue4951] failure in test_httpservers

2009-04-24 Thread Mark Dickinson
Mark Dickinson added the comment: On my machine, the failure seems to come down to this: Macintosh-4:py3k dickinsm$ MACOSX_DEPLOYMENT_TARGET=10.4 ./python.exe Python 3.1a2+ (py3k:71845M, Apr 24 2009, 19:51:44) [GCC 4.0.1 (Apple Inc. build 5490)] on darwin Type "help", "copyright", "credits" or

[issue4951] failure in test_httpservers

2009-04-24 Thread Mark Dickinson
Mark Dickinson added the comment: Here's the verbose test output: Macintosh-4:py3k dickinsm$ ./python.exe Lib/test/regrtest.py -v test_distutils test_httpservers test_distutils test_formats (distutils.tests.test_bdist.BuildTestCase) ... ok test_simple_built (distutils.tests.test_bdist_dumb.Bui

[issue4951] failure in test_httpservers

2009-04-24 Thread Mark Dickinson
Mark Dickinson added the comment: This looks like an interaction with test_distutils. On OS X 10.5.6 I get: Macintosh-4:py3k dickinsm$ ./python.exe Lib/test/regrtest.py test_distutils test_httpservers test_distutils test_httpservers Traceback (most recent call last): File "/private/var/fol

[issue4951] failure in test_httpservers

2009-03-31 Thread Ned Deily
Ned Deily added the comment: Also seeing on OS X. Without having looked at the code yet, I wonder if this might be related to Issue1711605. -- nosy: +nad ___ Python tracker ___

[issue4951] failure in test_httpservers

2009-03-15 Thread Ger Luijten
Ger Luijten added the comment: Reported Issue 5494 Failure in test_httpservers on Linux with other error message. Maybe there is a relation, maybe not. Greetings, Ger -- nosy: +gerluijten ___ Python tracker ___

[issue4951] failure in test_httpservers

2009-01-14 Thread Antoine Pitrou
New submission from Antoine Pitrou : This one seems to pop up rather frequently on the buildbots: == FAIL: test_post (test.test_httpservers.CGIHTTPServerTestCase)