Re: GdkColor to Hexadecimal conversion

["David Necas (Yeti)" <yeti physics muni cz>]

On Wed, Aug 17, 2005 at 12:40:13PM +0200, Iago Rubio wrote:
> On Wed, 2005-08-17 at 20:21 +1000, Nick Watts wrote:
> > Easiest way to convert a GdkColor to its hexadecimal equivalent?
>
> If you mean an #RRGGBB string for web use: 
>
> gchar* // please free me when you're done
> gdk_color_to_rgb_hex_string(GdkColor* color)
> {
>       gchar* rgb;
>       gfloat r, g, b;
>
>       r=color->red;
>       g=color->green;
>       b=color->blue;
>
>       r=(r/65535)*255;
>       g=(g/65535)*255;
>       b=(b/65535)*255;
>
>       rgb = g_strdup_printf("#%02x%02x%02x",(gint) r,(gint) g,(gint) b);
>
>       return rgb;
> }

Isn't that a complicated way to calculate a biased result?
I mean: To get the same rounding, you could just use

    rgb = g_strdup_printf("#%02x%02x%02x",
                          color->red*255/65535,
                          color->green*255/65535,
                          color->blue*255/65535);

and don't bother with floats at all.

But such a rounding seems biased.  There are 257 values that
give 0, another 257 give 1, etc., but only one (65535) gives
255.  Even simple

    rgb = g_strdup_printf("#%02x%02x%02x",
                          color->red >> 8,
                          color->green >> 8,
                          color->blue >> 8);

maps the same number of values (256) to each of the numbers
0..255.  Or is there any specific reason for rounding to zero?
I can imagine reasons for discrimination of both border
values (0 and 255), but why just 255?

Yeti


--
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?