[cairo] Extending Cairo API to support reusable paths

[Matt Woodrow]

I'd agree in some ways, but this is keeping in style with how the 
current surface pointers are passed.

Creating an otherwise empty wrapper object or a typedef could make the 
code more readable at least.

Matt

On 20/04/10 4:14 PM, M Joonas Pihlaja wrote:
>    
>>      /*
>>       * Creates a backend specific path object from the current path data
>>       */
>>      const void *
>>      (*build_path_object)    (cairo_path_fixed_t     *path,
>>                   cairo_gstate_t         *gstate);
>>      
> [snip more void pointers]
>
> This gives me pimples.  The returned pointer type ought to be typed,
> so that the responsibility of keeping track of types is with the
> implementor rather than the user of these APIs.  It's okay to be a
> completely opaque type imho, but not one that says "whatevah, I do
> what I want!" one.
>
> Cheers,
>
> Joonas
>
>