Fix color issues

The issue with the color was that a little bug regarding to the bit shifting operations in the native color converter. This fixes the issue by shifting the additional required 8 bits to become the right color by conversion.

native/color.c

@@ -6,14 +6,14 @@ void pyfb_initcolor_u32(struct pyfb_color* cptr, uint32_t value) {
 
     // initialize the 16 bit color
 #if __BYTE_ORDER == __LITTLE_ENDIAN
-    unsigned int r = (value & 0xFF000000) >> 16;
-    unsigned int g = (value & 0x00FF0000) >> 8;
-    unsigned int b = (value & 0x0000FF00);
+    unsigned int r = (value & 0xFF000000) >> 24;
+    unsigned int g = (value & 0x00FF0000) >> 16;
+    unsigned int b = (value & 0x0000FF00) >> 8;
     u16_color      = (uint16_t)((r >> 3 << 11) + (g >> 2 << 5) + (b >> 3));
 #else
-    unsigned int r  = (value & 0x000000FF) << 16;
-    unsigned int g  = (value & 0x0000FF00) << 8;
-    unsigned int b  = (value & 0x00FF0000);
+    unsigned int r  = (value & 0x000000FF) << 24;
+    unsigned int g  = (value & 0x0000FF00) << 16;
+    unsigned int b  = (value & 0x00FF0000) << 8;
     u16_color       = (uint16_t)((r << 3 >> 11) + (g << 2 >> 5) + (b << 3));
 #endif
 
@@ -35,8 +35,7 @@ void pyfb_initcolor_u16(struct pyfb_color* cptr, uint16_t value) {
     g               = g * 255 / 63;
     b               = b * 255 / 31;
 
-    u32_color = (uint32_t)((r << 16) | (g << 8) | b);
-    u32_color = u32_color << 8;
+    u32_color = (uint32_t)((r << 24) | (g << 16) | (b << 8));
 
     // now set the alpha
     u32_color &= 0xFFFFFF00;// clear alpha channel if something is in there
@@ -49,8 +48,7 @@ void pyfb_initcolor_u16(struct pyfb_color* cptr, uint16_t value) {
     g               = g * 255 / 63;
     b               = b * 255 / 31;
 
-    u32_color = (uint32_t)((r >> 16) | (g >> 8) | b);
-    u32_color = u32_color >> 8;
+    u32_color = (uint32_t)((r >> 24) | (g >> 16) | (b >> 8));
 
     // now set the alpha
     u32_color &= 0x00FFFFFF;// clear the alpha channel if something is in there