#10437: falkon browser -------------------------+----------------------- Reporter: ken@… | Owner: ken@… Type: enhancement | Status: assigned Priority: normal | Milestone: 8.3 Component: BOOK | Version: SVN Severity: normal | Resolution: Keywords: | -------------------------+-----------------------
Comment (by ken@…): Developments, but I'm not sure if I can call it progress ;) For the python, I definitely do not have the libTestPlugin.so which fedora remove from their build. The two optional modules are both part of the same (QT) git tree (no releases). The po files for hellopython are for a script which which requires PySide2 and is not installed, at least when the modules are missing. As to the tests - my runs on the other machine were with 3.0.0 installed whereas on the 8.1/5.11.0 I had run them on a fresh system before installing. Now that I retry (i.e. after the install) only one fails. The results are in Testing/Temporary/LastTest.log (this is created when ctest (from cmake) is run by 'make test'). {{{ ********* Start testing of WebViewTest ********* Config: Using QtTest library 5.11.0, Qt 5.11.0 (x86_64-little_endian-lp64 shared (dynamic) release build; by GCC 8.1.0) PASS : WebViewTest::initTestCase() FAIL! : WebViewTest::loadSignalsChangePageTest() Compared values are not the same Actual (((loadFinishedSpy.count()))): 2 Expected (loadFinishedEmitCount) : 1 Loc: [/tmp/falkon-3.0.1/autotests/webviewtest.cpp(87)] PASS : WebViewTest::cleanupTestCase() Totals: 2 passed, 1 failed, 0 skipped, 0 blacklisted, 15276ms ********* Finished testing of WebViewTest ********* <end of output> Test time = 15.70 sec ---------------------------------------------------------- Test Failed. "falkon-webviewtest" end time: Jun 09 20:42 BST "falkon-webviewtest" time elapsed: 00:00:15 }}} NB that test runs very slowly when it fails. There was a suggestion on one of the stack websites to use CTEST_OUTPUT_ON_FAILURE=TRUE when invoking make test, but for this failure, no extra information results. Using diff, and then view, between 3.0.0 and 3.0.1 was painful, 'view' thought it was latin1, so lots of gibberish in the po files but I saw one suspicious change. Using 'less' gave me sane UTF-8 but lacked the highlighting of old/new which 'view' provides. With 'git-diff' followed by 'view' it was easy to find: {{{ diff --git a/falkon-3.0.0/autotests/webviewtest.cpp b/falkon-3.0.1/autotests/webviewtest.cpp index a21e693..40b0a57 100644 --- a/falkon-3.0.0/autotests/webviewtest.cpp +++ b/falkon-3.0.1/autotests/webviewtest.cpp @@ -82,7 +82,7 @@ void WebViewTest::loadSignalsChangePageTest() view.setPage(page2); // WebPage: Workaround for broken load started/finished signals in QtWebEngine 5.10 - const int loadFinishedEmitCount = qstrncmp(qVersion(), "5.10.", 5) == 0 ? 2 : 1; + const int loadFinishedEmitCount = qstrncmp(qVersion(), "5.11.", 5) == 0 ? 1 : 2; QTRY_COMPARE(loadFinishedSpy.count(), loadFinishedEmitCount); QCOMPARE(loadStartedSpy.count(), 0); }}} It is only in the '''test''' and I wonder if the "== 0 ? 2 : 1;" was accidentally changed (presumably the 5.11. needed to be changed). Will explore that later. And in case anybody wonders, when I was at school we had to show our workings in exams. By writing this down, I'm hoping that maybe I'll be able to find it when I've forgotten how I reached these opinions. -- Ticket URL: <http://wiki.linuxfromscratch.org/blfs/ticket/10437#comment:11> BLFS Trac <http://wiki.linuxfromscratch.org/blfs> Beyond Linux From Scratch -- http://lists.linuxfromscratch.org/listinfo/blfs-book FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page