Test for Bug 4543: Testers needed

[Charles Davis]

Vincent Povirk wrote:
> Now on windows (with s/if(0)/if(1)/) I get lots of:
> 
> metafile.c:740: Test failed: unexpected EMF bits: got 01, expected 0xff
> 
> metafile.c:740: Test failed: unexpected EMF bits: got 00, expected 0xff
> 
> metafile.c:740: Test failed: unexpected EMF bits: got 00, expected 0xff
> 
> ...
> 
> You, uh, do realize metafiles are a vector format, right? The data
> isn't really going to be all 1's.
Right. I thought, though, that the GetEnhMetaFileBits() function would
return the bits as they are drawn by playing the metafile. Silly me. It
returns the raw bits in the metafile itself. (The use of the term "bits"
confused me.) That test is going to go away then.
> 
> If I comment out the for loop and initialize i to 1, the test succeeds
> on Windows.
> 
There's one test that succeeds if there are exactly two BitBlt() records
in the metafile. The test_BitBlt_record() function just increments i
(through a pointer) every time it encounters a BitBlt() record. What
happens when i is initialized to 0?

Chip