[13:06] [telegram] It looks like redefining `_FORTIFY_SOURCE` in Qt broke veyon: [13:06] [telegram] https://launchpadlibrarian.net/706208072/buildlog_ubuntu-noble-amd64.veyon_4.7.5+repack1-1build3_BUILDING.txt.gz [13:20] [telegram] Do you mean the qtbase upload, or the settings now in dpkg and gcc-13 that set `_FORTIFY_SOURCE`=3 as the new default? (re @mitya57: It looks like redefining _FORTIFY_SOURCE in Qt broke veyon: [13:20] [telegram] https://launchpadlibrarian.net/706208072/buildlog_ubuntu-noble-amd64.veyon_4.7.5+repack1-1build3_BUILDING.txt.gz) [13:23] [telegram] I mean the qtbase upload. That's where =2 comes from, I guess. [13:24] [telegram] :( [13:26] [telegram] Reverting that will a) cause many things to crash on start, and b) qtbase would never migrate as multiple autopkgtest against it would fail [13:26] [telegram] I guess the best thing would be an upstream fix for qtbase to not crash things if built with =3 [13:27] [telegram] I have no clue on that though [13:28] [telegram] Maybe we should find where qtbase stores information that it was built with =2 and delete it? [13:28] [telegram] If that is a thing, then yeah [13:32] [telegram] ``` [13:32] [telegram] $ reverse-depends src:veyon [13:32] [telegram] No reverse dependencies found``` [13:32] [telegram] so not a disaster if we have to push that out of the release pocket for a bit [13:33] [telegram] Ok. Let's see if some other package breaks with the same symptoms. [13:33] [telegram] yeah [13:35] [telegram] FYI, I also saw that if you don't build qtbase with =2, then qtlocation FTBFS [13:36] [telegram] qtlocation was NCR for the icu transition and failed with the original =3 qtbase [13:36] [telegram] then built ok on retry after qtbase was fixed [13:43] [telegram] Ack [13:45] [telegram] @mitya57 also if you want to trying building some things with different options, I have the ppa: [13:45] [telegram] https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/4979 [13:45] [telegram] Ack. In fact I don't have much time today, maybe I'll play with it later. [13:46] [telegram] Of course. Thanks as always for what you have time for :D [21:19] [telegram] @mariogrip Hi! Do you have any ideas about this error in lomiri-ui-toolkit? [21:19] [telegram] https://launchpadlibrarian.net/706284141/buildlog_ubuntu-noble-amd64.lomiri-ui-toolkit_1.3.5010+dfsg-3build1_BUILDING.txt.gz [21:19] [telegram] ``` [21:19] [telegram] ucstylehints.cpp: In member function ‘void UCStyleHintsParser::verifyProperty(const QQmlRefPointer&, const QV4::CompiledData::Binding*)’: [21:19] [telegram] ucstylehints.cpp:54:23: error: invalid use of non-static member function ‘QV4::CompiledData::Binding::Type QV4::CompiledData::Binding::type() const’ [21:19] [telegram] 54 | if (binding->type == QV4::CompiledData::Binding::Type_Object) {``` [21:20] [telegram] (It's with Qt 5.15.12.) [21:49] [telegram] heyo! [21:50] [telegram] Jumping in for Marius real quick: seems current git has a fix for it: https://gitlab.com/ubports/development/core/lomiri-ui-toolkit/-/blob/main/src/LomiriToolkit/ucstylehints.cpp#L54 (re @mitya57: @mariogrip Hi! Do you have any ideas about this error in lomiri-ui-toolkit? [21:50] [telegram] https://launchpadlibrarian.net/706284141/buildlog_ubuntu-noble-amd64.lomiri-ui-toolkit_1.3.5010+dfsg-3build1_BUILDING.txt.gz [21:50] [telegram] ucstylehints.cpp: In member function ‘void UCStyleHintsParser::verifyProperty(const QQmlRefPointer&, const QV4::CompiledData::Binding*)’: [21:50] [telegram] ucstylehints.cpp:54:23: error: invalid use of non-static member function ‘QV4::CompiledData::Binding::Type QV4::CompiledData::Binding::type() const’ [21:50] [telegram] 54 | if (binding->type == QV4::CompiledData::Binding::Type_Object) {)