Gerald Pfeifer wrote:
Not sure how to else handle this; possibly remove this whole subtest?
Gerald
ChangeLog: Simplify code in dlls/imm32/tests/imm32.c (and avoid false warning).
diff --git a/dlls/imm32/tests/imm32.c b/dlls/imm32/tests/imm32.c index 1951242..3e35ac1 100644 --- a/dlls/imm32/tests/imm32.c +++ b/dlls/imm32/tests/imm32.c @@ -210,7 +210,7 @@ static void test_ImmNotifyIME(void) {
/* behavior differs between win9x and NT */ ret = ImmGetCompositionString(imc, GCS_COMPSTR, resstr, sizeof(resstr));
- ok(ret || !ret, "You'll never read this.\n");
- ok(TRUE, "You'll never read this.\n");
We could use broken() to make sure we follow the wanted behaviour for Wine. From the patch that changed that item it should be:
ok(!ret || broken(ret), .....)
If you look however at test.winehq.org it looks like hardly any box meets the proper requirements to run these tests at all.
So I'd say:
ok(!ret, "After being cancelled the composition string is empty.\n");
(Tested on my W2K3 and Vista boxes).
-
Paul Vriens