Re: Color Palette Incompatibilities

Murray Altheim (
Fri, 6 Sep 1996 14:25:08 -0500

Message-Id: <v02110111ae562593a141@[]>
Date: Fri, 6 Sep 1996 14:25:08 -0500
To: Stewart Brodie <>
From: (Murray Altheim)
Subject: Re: Color Palette Incompatibilities

Stewart Brodie <> writes:
>Magenta & Cyan are the standard names for these colours.  Quite why
>people (Microsoft?) felt it necessary to rename them is a mystery.
>Fuchsia is extremely vague (fuchsias come in a variety of colours).

Comes from the original dyestuff for magenta (fuchsine), not the flower. I
think the original color list came from SGI, not MS. But I too have a
problem with "Cantaloupe Mist"-type color names...

Earlier this year I proposed a simplified color naming specification,
complete with hex values, etc. in reaction to the ugliness of the CSS color

My original proposal was meant as a _simple_ method of creating a color
namespace, with hex values for specifying colors, since no other method is
supported by current browsers.

Chris Lilley (working at W3C) responded in depth (we had a long discussion
about it, if anyone remembers), which if I remember correctly ended with a
scientifically accurate conclusion, but one I don't think anyone can
currently use. I still maintain that when one is working in mainstream
computer systems with RGB monitors, that "common understanding" is that
red-green-blue are the recognized primaries, with yellow-magenta-cyan the
secondaries; on this I based my proposal. As Chris correctly pointed out,
non-gamma-corrected RGB values are not platform-independent and therefore
not suitable as an accurate specifier. He also disagreed with my RGB/CMY
naming scheme and the fact that I didn't include a saturation modifier. I
disagreed with his Munsell-based color names. As I mentioned, we had a long
discussion about this which I don't see the need to rehash now.

As anyone who has done any work in color or color theory, this issue gets
quite muddled very quickly, and is not one that lends itself to a simple
solution, which is what I was aiming at. Possibly adding a one or two-level
saturation modifier to my proposal would provide a workable solution, but
we really need browser support for CIE colors, which are truly platform
independent. Mapping color names to CIE will still be problematic.

I haven't followed any W3C developments in CNS color specification since
June. Possibly Chris has moved farther ahead with his solution.


     Murray Altheim, Program Manager
     Spyglass, Inc., Cambridge, Massachusetts
     email: <>
     http:  <>
            "Give a monkey the tools and he'll eventually build a typewriter."