Description of problem: Package enki fails to build from source in rawhide. Version-Release number of selected component (if applicable): 16.04.0-2.fc25 Steps to Reproduce: koji build --scratch f25 enki-16.04.0-2.fc25.src.rpm Additional info: This package is tracked by Koschei. See: https://apps.fedoraproject.org/koschei/package/enki From build.log: … test_dev_null (test_core.test_opened_files_dock.Rename) ... ERROR test_dev_null_os_fail (test_core.test_opened_files_dock.Rename) ... ERROR test_flags (test_core.test_opened_files_dock.Rename) ... ok test_os_fail (test_core.test_opened_files_dock.Rename) ... ERROR test_same_path (test_core.test_opened_files_dock.Rename) ... ERROR test_success (test_core.test_opened_files_dock.Rename) ... ERROR test_1 (test_core.test_project.Test) Parse words ... ok test_1 (test_core.test_uisavefiles.Test) ... ERROR test_2 (test_core.test_uisavefiles.Test) ... /usr/bin/xvfb-run: line 166: 23461 Segmentation fault (core dumped) DISPLAY=:$SERVERNUM XAUTHORITY=$AUTHFILE "$@" 2>&1
I'd rather not block the python-qt5 update on *just* failing tests. Is there any verifiable runtime impact?
(In reply to Rex Dieter from comment #1) > I'd rather not block the python-qt5 update on *just* failing tests. Is > there any verifiable runtime impact? Those tests are usability tests emulating some of the general use cases. If they fail, the installed package can be assumed to misbehave.
Still, I'd rather not speculate. (We already have some evidence to the contrary, in the past where some of these tests failed in ways that were false positives)
When I call enki-16.04.0-2.fc24.noarch with python3-qt5-5.6-1.fc24.x86_64 : CRITICAL:root:Traceback (most recent call last): File "/usr/lib/python3.5/site-packages/enki/core/openedfilemodel.py", line 293, in _onDocumentOpened document.documentDataChanged.connect(self._onDocumentDataChanged) TypeError: decorated slot has no signature compatible with Document.documentDataChanged[]
Report it to your upstream, it appears enki's use of that Document.documentDataChanged api is no longer compatible. Or could be a PyQt5 bug, introducing an incompatible change, but I'm not sure if this is a public api or not, your upstream should be better suited to answer that question.
Document with its documentDataChanged refers here to enki.core.Document with a custom signal https://github.com/hlamer/enki/blob/master/enki/core/document.py#L151 http://pyqt.sourceforge.net/Docs/PyQt5/signals_slots.html#PyQt5.QtCore.pyqtSignal Did there anything change in PyQt5 sources?
Please include newest upstream fixees in rawhide. changeset: 12c3868cce84 date: Tue May 03 12:18:24 2016 +0100 Documented the slot signature checking as a potential incompatibility. changeset: 2a1416f9a2d3 date: Tue May 03 08:34:18 2016 +0100 Fixed connections to decorated slots to enforce the slot signature and not silently ignore the decorator. sip.voidptr and bytes arguments to the decorator now generate an appropriate C++ type. https://www.riverbankcomputing.com/static/Downloads/PyQt5/ChangeLog-5.6.1.dev1605031218
I built the latest PyQt5 snapshot locally, and the enki tests (still) fail the same way.
I found the problem in enki source, will prepare a patch and send it to upstream.
raphgro's enki-16.04.0-3.fc25 completed http://koji.fedoraproject.org/koji/buildinfo?buildID=761422
enki-16.04.0-3.fc24 has been submitted as an update to Fedora 24. https://bodhi.fedoraproject.org/updates/FEDORA-2016-a65c5ce647
enki-16.04.0-3.fc24 has been pushed to the Fedora 24 testing repository. If problems still persist, please make note of it in this bug report. See https://fedoraproject.org/wiki/QA:Updates_Testing for instructions on how to install test updates. You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2016-a65c5ce647
enki-16.04.0-3.fc24 has been pushed to the Fedora 24 stable repository. If problems still persist, please make note of it in this bug report.