Bug#499344: Wrong cursor when oxygencursors are installed
reassign 499344 kdebase-workspace thanks To restate the bug report, crystalcursors includes all the important cursors, including wait/busy, what is this?, pointing hand, text selection, and resizing cursors, but the corresponding cursors from Oxygen are used nonetheless. I suspect this is because there are at least two names for each cursor, and KDE 4 uses the cursor with the preferred/new name from the default theme before any cursor available under an alternative name in the selected theme (actually I think it's just those MD5 hashes of the hardcoded built-in cursors that are used). I think that's a bug in KDE, but I'm not sure which package is responsible. I picked kdebase-workspace because it's kdebase-workspace-data that depends on oxygencursors. Please pass the bug on if I guessed wrong. Alternatively, or as a workaround, all cursor packages might have to provide all their cursors under the new names as well. That would also work around the swapped resize cursors bug 548804. -- Magnus Holmgrenholmg...@debian.org Debian Developer signature.asc Description: This is a digitally signed message part.
Bug#499344: reassign 499344 to crystalcursors
On söndagen den 8 augusti 2010, you stated the following: > KDE only tells X the paths where to look for cursors Also reported at https://bugs.kde.org/show_bug.cgi?id=213615, with a perhaps better analysis. I still don't think it's a bug in any cursor package; maybe in Qt? -- Magnus Holmgrenholmg...@debian.org Debian Developer signature.asc Description: This is a digitally signed message part.
Bug#548804: crystalcursors: the corner resize cursors are swapped
reassign 548804 kdebase-workspace
forwarded 548804 https://bugs.kde.org/248599
tags 548804 + upstream patch
thanks
On söndagen den 8 augusti 2010, Magnus Holmgren wrote:
> Basically all the cursor packages except oxygencursors display the same
> bug. It appears that it has to do with the hash symlinks
> fcf1c3c7cd4491d801f1e1c78f10 and c7088f0f3e6c8088236ef8e1e3e7,
> which according to the author of Crystalcursors is part of the "QT
> symlinks mess". oxygencursors does not have these symlinks; apparently
> KDE/Qt uses size_fdiag and size_bdiag instead.
>
> It's almost as if those two hashes were mixed up in KDE 4, perhaps in
> QString XCursorTheme::findAlternative(const QString &name) const in
> xcursortheme.cpp?
Suggested patch if this is the case is attached. If it's not the case, the
world (of cursor theme makers) should be told.
--
Magnus Holmgrenholmg...@debian.org
Debian Developer
--- a/kcontrol/input/xcursor/xcursortheme.cpp
+++ b/kcontrol/input/xcursor/xcursortheme.cpp
@@ -82,8 +82,8 @@ QString XCursorTheme::findAlternative(co
// Note that the MD5 hash for left_ptr_watch is for the KDE version of that cursor.
alternatives.insert("size_ver", "81600681408080010102");
alternatives.insert("size_hor", "028006030e0e7ebffc7f7070c0600140");
-alternatives.insert("size_bdiag", "c7088f0f3e6c8088236ef8e1e3e7");
-alternatives.insert("size_fdiag", "fcf1c3c7cd4491d801f1e1c78f10");
+alternatives.insert("size_bdiag", "fcf1c3c7cd4491d801f1e1c78f10");
+alternatives.insert("size_fdiag", "c7088f0f3e6c8088236ef8e1e3e7");
alternatives.insert("whats_this", "d9ce0ab605698f320427677b458ad60b");
alternatives.insert("split_h","14fef782d02440884392942c11205230");
alternatives.insert("split_v","2870a09082c103050810ffde0204");
signature.asc
Description: This is a digitally signed message part.
Bug#499344: reassign 499344 to crystalcursors
On måndagen den 9 augusti 2010, I stated the following: > On söndagen den 8 augusti 2010, Sune Vuorela stated the following: > > KDE only tells X the paths where to look for cursors > > Also reported at https://bugs.kde.org/show_bug.cgi?id=213615, with a > perhaps better analysis. I still don't think it's a bug in any cursor > package; maybe in Qt? I think you could have been more helpful, but anyway, I now understand that it's libXcursor that is responsible for the fallback to the default theme, and that neither KDE nor Qt can do much about it (except that Qt could have used standard X11 cursor names when running under X11). Perhaps lower the alternatives priority for oxygencursors so that if the administrator installs additional themes besides Oxygen, one of them will be the default by default, so to speak? -- Magnus Holmgrenholmg...@debian.org Debian Developer signature.asc Description: This is a digitally signed message part.
Bug#499344: Wrong cursor when oxygencursors are installed
On söndagen den 8 augusti 2010, Magnus Holmgren wrote: > Alternatively, or as a workaround, all cursor packages might have to > provide all their cursors under the new names as well. That would also > work around the swapped resize cursors bug 548804. It appears that, at least for the time being, all cursor packages will have to provide the Qt cursor names as symlinks for them to work in KDE when Oxygen is the default theme. Here is a list: center_ptr up_arrow watch wait xterm ibeam fleur size_all hand pointing_hand v_double_arrow size_ver h_double_arrow size_hor fd_double_arrow size_bdiag bd_double_arrow size_fdiag question_arrow whats_this sb_h_double_arrow split_h sb_v_double_arrow split_v crossed_circle forbidden grabbing closedhand fleur openhand (?) -- Magnus Holmgrenholmg...@debian.org Debian Developer signature.asc Description: This is a digitally signed message part.
Bug#328362: pmk: postinst fails, missing depends?
found 328362 0.9.3s2-2.1 thanks I can reproduce this bug too. Apparently pmksetup segfaults after reading pmkcpu.dat. Version 0.10.1 seems to fix the bug. I can take maintainership of the package and upload the new version. -- Magnus Holmgren[EMAIL PROTECTED] (No Cc of list mail needed, thanks) pgpkTnx7fGoc1.pgp Description: PGP signature
