[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
Guilherme Polo ggp...@gmail.com added the comment: Finally I'm looking into this again. So, for now, I decided to only move the tk load tests to Lib/lib-tk/test/test_tkinter under a new module named test_loadtk. Lib/test/test_tcl remains almost the same, except it no longer it contain those tests related to tk loading. Patch attached. May I reassign it to me David ? -- Added file: http://bugs.python.org/file14329/moving_loadtktests.diff ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
R. David Murray rdmur...@bitdance.com added the comment: On Sun, 21 Jun 2009 at 14:57, Guilherme Polo wrote: Patch attached. May I reassign it to me David ? Absolutely. -- title: test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped - test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
Guilherme Polo ggp...@gmail.com added the comment: Running tk tests through both Lib/test/test_tk.py and Lib/test/regrtest.py show the desired behaviour (from what I understood from your description and from what I tested). It has been committed now, r73495 (trunk). Should 2.6 and 3.0 really receive this patch ? I've removed them from the list in this issue, please add back if it is really needed. -- assignee: r.david.murray - gpolo versions: -Python 2.6, Python 3.0, Python 3.2 ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
R. David Murray rdmur...@bitdance.com added the comment: No, I don't see any reason to bother backporting. From my understanding we're not backporting anything to 3.0 at this point anyway. -- ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
Guilherme Polo ggp...@gmail.com added the comment: Fine, closing then. Committed as r73497 on py3k. -- status: open - closed ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
Changes by R. David Murray rdmur...@bitdance.com: -- assignee: gpolo - r.david.murray priority: - low stage: - needs patch versions: +Python 2.7, Python 3.2 ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
Guilherme Polo ggp...@gmail.com added the comment: Uhm, I don't agree with this TkinterTest name since it is only doing a minor test on _tkinter._flatten. This same class is indicated in tests_gui but it doesn't require gui to run, it should be indicated only in tests_nogui. Then there is TclTest which I also don't agree with the name, it is only testing tk loading so a different name would be better here. And finally the file is named test_tcl.py which to me suggests something that won't require gui, but that is exactly what it does by doing test_support.requires('gui'). Any chance you can improve these names ? -- assignee: - gpolo ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
R. David Murray rdmur...@bitdance.com added the comment: On Tue, 24 Mar 2009 at 18:24, Guilherme Polo wrote: Any chance you can improve these names ? Absolutely. I agree with you; I was trying to stay parallel to the original names, but it felt wrong. And since I don't know TCL/TK/tkinter, I didn't know what flatten was testing. I'll make it non-gui. -- ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
R. David Murray rdmur...@bitdance.com added the comment: I tried using TestSkipped first, but since I was throwing it from an individual testcase and not from the top level, it still showed up as an error in test_tcl (I may have done it wrong, of course). Moving the tests makes more sense anyway, so I've revised my patch to do that. I didn't move all of the tests because as far as I could understand the test.test_tk code, if I moved them all they'd only get run if the GUI resource was enabled, and that seems wrong. So I just moved the tests that specifically involved tk. I ran the test suite under a variety of DISPLAY settings and non-settings, and it seemed to do the right thing with the patch applied. I have pushed the change up to Launchpad as well. -- Added file: http://bugs.python.org/file13308/test_tcl.patch ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
R. David Murray rdmur...@bitdance.com added the comment: Oops, I forgot to remove the changes I'd previously made to testLoadTk. I've updated the patch file and pushed to Launchpad, after re-running my tests. -- Added file: http://bugs.python.org/file13310/test_tcl.patch ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
Changes by R. David Murray rdmur...@bitdance.com: Removed file: http://bugs.python.org/file13308/test_tcl.patch ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
R. David Murray rdmur...@bitdance.com added the comment: Patch uploaded as bzr branch to Launchpad at bzr+ssh://bazaar.launchpad.net/~rdmurray/python/bug5450-test -- message_count: 2.0 - 3.0 ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
Guilherme Polo ggp...@gmail.com added the comment: Please use test_support.TestSkipped instead of showing a skip message using print. -- message_count: 3.0 - 4.0 nosy: +gpolo nosy_count: 1.0 - 2.0 ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
Guilherme Polo ggp...@gmail.com added the comment: It should be okay to move tk tests to somewhere in Lib/lib-tk/test/test_tkinter/ (answering the final question). Even the rest of test_tcl could be relocated to that place. -- message_count: 4.0 - 5.0 ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
New submission from R. David Murray rdmur...@bitdance.com: If the DISPLAY environment variable is set, the test_tcl test 'testLoadTk' gives a traceback. In the same circumstance, test_tk and test_ttkguionly display the error and are skipped. Although this is a nit rather than a problem, for consistencies sake I think the testLoadTk test should also be skipped rather than show up as a failure. I've attached a patch against the trunk that does this, with the caveat that it is an individual testcase being skipped so it doesn't get recorded in the 'skipped tests' summary. I don't think this is a problem since test_tk will likely show up in the skipped tests if this one is skipped, and also there are circumstances in which testLoadTkFailure is skipped completely silently. Perhaps the real problem is that the tk tests should be in test_tk instead of test_tcl? -- components: Tests, Tkinter files: test_tcl.patch keywords: patch message_count: 1.0 messages: 83339 nosy: bitdancer nosy_count: 1.0 severity: normal status: open title: test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped type: behavior versions: Python 2.6, Python 3.0, Python 3.1 Added file: http://bugs.python.org/file13273/test_tcl.patch ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue5450] test_tcl testLoadTk fails if DISPLAY defined but connect fails, instead of being skipped
R. David Murray rdmur...@bitdance.com added the comment: That first sentence should have read If the DISPLAY environment variable is set but the display cannot be opened. -- message_count: 1.0 - 2.0 ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue5450 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com