W3C home > Mailing lists > Public > www-style@w3.org > June 2002

RE: X11 Colors (was Last call comments on CSS3 module: color)

From: Ian Tindale <ian_tindale@yahoo.co.uk>
Date: Wed, 19 Jun 2002 12:44:09 +0100
To: <www-svg@w3.org>, <www-style@w3.org>
Message-ID: <000101c21786$a2db6920$fc00a8c0@solstice>

One of the aspects of those X11 named colours is that they’re not canonical in any real way. They’re simply a flat-filed equally meaningful/less unprioritised mass of interesting words, some having connotations and others not. There’s little semantic value. In fact it could be said that the semantics stop at the point of using a colour name.

If reality were to be written all over again, it’d be nicer to have chosen an almost dot-notation approach, where it is possible to ascertain which data value you need by having a rough and big blunt stab at a small selection of rough big blunt colours that are more likely to be universally ‘meaningful’, like ‘red’, ‘blue’, or ‘pink’. Ah, but then you get ‘cyan’, ‘magenta’ – it’s possible to find people who don’t know what these are (probably). Anyway, imagine you had these big blunt colours, and you were then to refine it in iteration form using property/descriptor-like fashion: So you want a blue, eh? What sort – light or dark? Light blue, thanks. This light? Lighter please. Okay, how saturated – pale or strong? Paler than that please. No, even paler. And on it goes – you end up with something like blue.light.er.pale.pale.er (of course, this wouldn’t work as is). This is a rough gist of what I mean about retaining some degree of semantic carrier, rather than leavingoutside the door as soon as we hit the colour names, which is presently the case.

Incidentally, when I was a magazine art editor*, using QuarkXpress all day, I’d define a set of colours for a particular article, and work from those. Many designers in slightly more rigid roles have to set up colour sets, for client work perhaps, and in doing so they’ll take one of several approaches. QuarkXpress doesn’t let you have every single colour displayed in the colour palette at once. You have to mix your own and give it a name – it could be derived from Pantone, it could be derived from RGB or CMYK or any other colour model available. It becomes an ‘instance’ of the ‘symbol’, as it were. Typically a designer will call the colour after its component values, like “blue-C100/M60/Y0/K10” or suchlike. Also, typically other designers might call a colour anything they fancy (bit like the X11 mentality). 

Personally, however, I elected to ‘objectify’ my colour naming in my magazine work, as I invariably made extensive use of style sheets in Xpress. I’d call my colours things like “Boxout-border” or “Standfirst-text” or “Crosshead-text” or “line-X” and “line-Y” or suchlike. This way I could immediately alter whatever the actual visual result was, which was abstracted from the usage, but I’d used the colour name as a usage tagging method instead of a reminder of how they looked.

I’m quite intrigued by the possibility of stylesheet cascade giving rise to ‘chained modifier’ colour specification.

--
*Mayfair Magazine, if you’re interested – fleshtone contones being a notoriously difficult colour to get through repro without undue quantising occurring.
-- 
Ian Tindale
Received on Wednesday, 19 June 2002 07:44:31 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:14 GMT