On 9/25/07, James Hawkins truiken@gmail.com wrote:
On 9/25/07, Christian Gmeiner christian.gmeiner@gmail.com wrote:
I am running make test an a fresh git fetch; git rebase origin and am getting this:
../../../tools/runtest -q -P wine -M comctl32.dll -T ../../.. -p comctl32_test.exe.so monthcal.c && touch monthcal.ok err:monthcal:MONTHCAL_WindowProc unknown msg 2006 wp=00000000 lp=00000000 err:monthcal:MONTHCAL_WindowProc unknown msg 2005 wp=00000001 lp=00000000 err:monthcal:MONTHCAL_WindowProc unknown msg 2006 wp=00000000 lp=00000000 err:monthcal:MONTHCAL_WindowProc unknown msg 2005 wp=00000000 lp=00000000 err:monthcal:MONTHCAL_WindowProc unknown msg 2006 wp=00000000 lp=00000000 err:monthcal:MONTHCAL_WindowProc unknown msg 2005 wp=00000001 lp=00000000 monthcal.c:818: Test succeeded inside todo block: Expected 131073, got 131073 make[2]: *** [monthcal.ok] Fehler 1
This patch removes the todo_wine.. as it seems to work now. If this patch gets not accepted, please tell me why.
Becuase it doesn't pass for other people, particularly Julliard; otherwise, it would have been fixed by now.
-- James Hawkins
The monthcal layout is definitely incorrect. See http://bugs.winehq.org/attachment.cgi?id=4485 for an example.
Commit 7495d814954420c16e21de40c3031a9c95385f56 (gdi32: Using a bitmap font as the fallback sans serif is a very bad idea.)
change the fonts used for many people. This made the test pass for some simply because they got lucky. I see this because the monthcal test fails on my work machine but works on my home machine.
I can pick a new coordinate to fix this, but then the new hit test will be just as brittle as the current one. Anyone have better ideas on how to make these coordinate based tests more robust?