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

On Thursday, May 23, 2002, 10:58:07 AM, Steven wrote:

SP> From: "Tantek Çelik" <tantek@cs.stanford.edu>
>> > The X11 colour names are an abomination that should have been stifled at
>> > birth, and adding them to CSS is a blemish on the otherwise excellent
SP> design
>> > of CSS.
>>
>> Would they not be a blemish on other W3C specs then?

SP> Absolutely! And I think the CSS group should lead the way and show how it
SP> *should* be done, rather than slavish following some brain-dead design from
SP> tens of years ago,

Steven,

Given that you are also proposing the (brain-dead, known flawed) HSV
system with its multitude of known problems in terms of usability you
are on rather shaky ground making thsese sorts of statements.

>> > To say that the X11 colour set and their names have been 'designed' is
SP> an
>> > insult to the word "design". It is just a mess.

Its a set of names, and they are widely used.

>> I don't know the history behind the X11 color set, nor do I know the
>> reasoning behind why they were implemented in the earliest graphical
>> browsers (Mosaic etc.), nor why they were widely adopted by web authors.

SP> Well, the need for named colours is clear.

No, not really. Anyone can make up their own names and declare them as
entities (in any XML document) and use them. For browsers that use a
real XML parser, of course. For external CSS files, people need to use
a preprocessor I guess.

SP> The first Mosaic was on X right?

Yes. In fact it was originally called xmosaic.

SP> So it is not surprising that they just used an existing name set that worked
SP> out of the box.

I suspect that was the reason.

>> But, that doesn't refute the fact that
>>
>> 1. the X11 color set has been widely consistently implemented for many
SP> years

Yes, true.

>> 2. colors in the X11 colors set have been used by web authors in documents
>> for many years.

Also true.

>> 3. the X11 color set has already been standardized by W3C[1]

That, too.

SP> I can name a dozen features of HTML that also fulfil all these properties,
SP> but that we are nevertheless ruthlessly consigning to the garbage can of
SP> history where they belong.

Cool. And for new formats starting with a clean slate, like XHTML 2,
that is exactly right.

>> One intent of CSS3 Color is provide some harmonization between the color
>> features in SVG and CSS.  The X11 colors are a part of that.

SP> And shouldn't be. Two messes do not make a clean design.

Ah, not only are X11 colors a mess but now, SVG is a mess too? Again,
thats pretty rich coming from the HTML working group. How many
implementations pass the XHTML test suite? Oh, yes, there isn't one.
HTML is one large, smelly mess, and worse due to the design flaws in
XHTML 1.x that allowed content to be served to existing browsers as if
that would somehow work, it is spreading non well formed content that
purports to be XML. So please, get you own house in order before
accusing others of "mess".

>> Yes, the X11 color set has already been accepted through the last call
>> process TWICE already (SVG 1.0, SVG 1.1) and for that reason alone they
SP> will
>> be included in CSS3 color - indeed, CSS implementations of X11 color
SP> predate
>> SVG implementations by quite some time.

SP> Nah. Now's the time to realise the mistake of including them in SVG and
SP> design a robust route towards a better solution.

Moving onto technical topics - what better solution would you propose?

>> > Let me just describe some of the problems:
>>
>> Problems acknowledged, and there is no intent to alter or add to the list
SP> of
>> X11 colors for exactly this reason.

SP> What a missed chance! You could have added some of the Crayola colours, and
SP> made the set even worse! What about Electric Lime, Purple Pizzazz, Razzle
SP> Dazzle Rose, Unmellow Yellow, Magic Mint, Radical Red, Sunglow and Neon
SP> Carrot? What about Macaroni and Cheese? We have a right to these colours!

If only XHTML clients actually had XML parsers, you could define them
yourself and use them.

>> > 1) Take the standard 6x6x6 internet non-dithering color cube that is
>> > implemented everywhere, and agree on names for each of the colors there.

ROFL. Non-dithering?

>> Not an undertaking the CSS wg is interested in I'm sure.

Well said.

SP> But they have been widely consistently implemented for many years and used
SP> by web authors in documents for many years. It is time to standardise them!
SP> Web authors need something to refer to!

They way they are used is already standardized, so thats a no-op.

>> > 2) Take a consistent naming scheme that properly addresses all
>> > dimensions of  the color space, and map this naming scheme
>> algorithmically to appropriate  colours. For instance:

>> A good attempt at improvement, but this is far too big an addition
>> to add to CSS3 Color at Last Call in my opinion. We should keep it as
>> a proposal for the next version.

Chuckle. Real colors that actually use color theory were too big a
change for CSS1, too big a change for CSS2, too big a change for CSS3
.... there is always a next version.


SP> 22/02 7:19 AM, In a reply to "Andrew Clover" <and@doxdesk.com>:

>> Could you name something else non standard that is as or more widely
>> supported?

SP> <embed> springs immediately to mind.

Easily declared in the internal DTD subset, and works in the existing
legacy browsers. XML content on non-XML clients, thats perfect XHTML
1.x, no?

>> The reasoning behind including the X11 colors is very much the same
>> as that that went into HTML 3.2 - codifying current practice.

SP> Which was right at that time. And that time is now past.

>> > There is no reason to
>> > make them a requirement of a CSS3 implementation.
>>
>> They are not a requirement of a CSS3 implementation.  CSS3 is a set of
>> modules.  Requirements for an implementation are documented through
>> profiles.  Example profiles are provided in the modules.

SP> So you're suggesting that if we don't want the X11 colours, we can just
SP> leave the colour module out?

Not the whole module, no. But a given profile can exclude them.

>> > Authors should not be encouraged to use the X11 colours.
>>
>> Why?

SP> 'Cause they're badly designed, badly distributed over the spectrum, wrongly
SP> named, include cultural references, ... see my earlier mail.

Rather like HSL is badly designed and badly distributed over the
'spectrum' (the color space). And has poor usability, something you
forgot to mention but a subject, I know, that is dear to your heart.

See my earlier emails. From 1996.

-- 
 Chris                            mailto:chris@w3.org

Received on Wednesday, 29 May 2002 16:35:11 UTC