On 5 September 2016 at 18:35, Matthias Kuhn <[email protected]> wrote: > Hi > > I asked on Qt bug tracker > > https://bugreports.qt.io/browse/QTBUG-47192?focusedCommentId=330334&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-330334 >
Thanks Matthias! I was planning on doing that too but my Qt account seems broken - I can't login and can't reset it... sigh. Nyall > Matthias > > On 09/05/2016 08:23 AM, Mathieu Pellerin wrote: >> Denis, >> >> There's an additional way to mitigate this: >> 1/ requesting Qt devs to apply the patch fixing this issue ( >> http://code.qt.io/cgit/qt/qtbase.git/patch/src/corelib/tools/qlocale.h?id=726fed0d67013cbfac7921d3d4613ca83406fb0f >> ) onto Qt 5.6, which after all is a LTS version; >> 2/ requesting distributions - such as Ubuntu and Fedora - to apply the >> patch onto the packages maintained and offered through their repositories. >> >> Regarding #2, I've filed such a request earlier today: >> https://bugs.launchpad.net/ubuntu/+source/qtbase-opensource-src/+bug/1620173 >> >> It's a shame we can't override the QVariant conversion to QString with >> some code of our own to patch the issue on Qt5 < 5.7. >> >> Math >> >> On Mon, Sep 5, 2016 at 12:23 PM, Denis Rouzaud <[email protected] >> <mailto:[email protected]>> wrote: >> >> Hi Mathieu, >> >> For reference, here is the Qt issue >> https://bugreports.qt.io/browse/QTBUG-47192 >> <https://bugreports.qt.io/browse/QTBUG-47192> >> >> It's the conversion from QVariant to QString which is broken. >> >> There are several ways to deal with this: >> * fix this for some versions of Qt (no idea how) >> * require Qt 5.7+ >> * ship like this with a warning notice >> >> The discussion already arised and the latter seems to be prefered: >> https://lists.osgeo.org/pipermail/qgis-developer/2016-August/044127.html >> >> <https://lists.osgeo.org/pipermail/qgis-developer/2016-August/044127.html> >> >> Denis >> >> >> >> On 09/05/2016 04:39 AM, Mathieu Pellerin wrote: >>> Greetings, >>> >>> While testing QGIS compiled against Qt5 5.6.1 (the version that >>> will be offered in Ubuntu's next release in a month or so), I ran >>> into a rounding issue, whereas the round() expression engine >>> function simply fails to round values to the stated number of >>> decimals. >>> >>> As seen in this screenshot[1], using this expression "round( >>> 0.972743..., 2 )" to round the double value to two decimals >>> returns 0.96999999999997. >>> >>> Nyall says the issue has been fixed in Qt5 5.7. For Qt5 < 5.7, is >>> there a contingency plan for this problem? Could we get >>> distributions to include a fix in their 5.6.x packages (if that’s >>> possible)? >>> >>> Math >>> >>> [1] http://i.imgur.com/7v2Z24Z.png >>> >>> >>> >>> _______________________________________________ >>> Qgis-developer mailing list >>> [email protected] <mailto:[email protected]> >>> List info: http://lists.osgeo.org/mailman/listinfo/qgis-developer >>> <http://lists.osgeo.org/mailman/listinfo/qgis-developer> >>> Unsubscribe: http://lists.osgeo.org/mailman/listinfo/qgis-developer >>> <http://lists.osgeo.org/mailman/listinfo/qgis-developer> >> >> >> >> >> _______________________________________________ >> Qgis-developer mailing list >> [email protected] >> List info: http://lists.osgeo.org/mailman/listinfo/qgis-developer >> Unsubscribe: http://lists.osgeo.org/mailman/listinfo/qgis-developer >> > _______________________________________________ > Qgis-developer mailing list > [email protected] > List info: http://lists.osgeo.org/mailman/listinfo/qgis-developer > Unsubscribe: http://lists.osgeo.org/mailman/listinfo/qgis-developer _______________________________________________ Qgis-developer mailing list [email protected] List info: http://lists.osgeo.org/mailman/listinfo/qgis-developer Unsubscribe: http://lists.osgeo.org/mailman/listinfo/qgis-developer
