Re: Negative yStride

In a just world, strides would be ptrdiff_t rather than size_t.


On Apr 29, 2014, at 12:54 PM, Brendan Bolles <[email protected]> wrote:

> I'm reading into an image buffer where the lines are ordered from bottom to top, so I'm using a negative yStride like so:
> 
> frameBuffer.insert(name,
>                    Slice(Imf::FLOAT,
>                            last_line_ptr,
>                            xStride,
>                            -rowbytes, // yStride
>                            xSampling, ySampling, fill) );
> 
> 
> But one thing I noticed is that Slice's yStride is a size_t, which is unsigned.  So instead of a modest negative number, the debugger shows it as a huge unsigned integer.
> 
> It seems to work fine, but I find it a little off-putting.  Should I be worried?
> 
> 
> Brendan
> 
> 

--
Larry Gritz
[email protected]