(RQT) character/state/comment

[Jim Croft]

>I believe there is an intermediate way, and that is providing
>templates for properties togehter with states. The more templates
>that could be offered, the easier the setup of a character definition
>would be. But the templates would be copied, not linked, so that e.g.
>too many color states could be copied from the template, and then
>reduced or changed as needed.

As an n-dimensional continuum, colour has always been dificult to quantify
- the number and size of available templates is mind-blowing.  And so much
is subjective - true believers know the colour turquoise is a green,
heretics insist it is a blue; and where are the boundaries around and
within the violet/indigo/purple/mauve/lilac impression?

petals white/pink/red/orange/yellow/green/blue/purple is simple enough, but
there are problems where you might want to draw a distinction betwen
scarlet and crimson, or the intensity of the hue (almost white, but did it
get there by losing yellow or losing green or by gaining a little bit of
red?).

In the interests of precision and flexibility it would seem sensible to
adopt a property/scheme/value model like the metadata bods tend to do.
That way you could be as precise as you like with Royal Hort. Soc.,
Munsell, Panatone, etc. colour charts, or R/B/G and C/M/Y ratios and people
could work out exactly what you thought you were dealing with.

The only trouble is the data becomes completely unintelligible an
meaningless to the average human - a perfect no-win situation.

But there are probably instances where you would want to do such a thing,
and our model should accommodate it.

As Kevin has pointed out several times, the notion of universal lexicons or
vocabularies is incredibly seductive, but often involve a lot of work and
running around that rarely results in the promised joy and bliss.  Sort of
like a global currency and free trade - a good idea, but not likely to
happen.  Our best bet is to resist the tempation to build the complete
description of all things right away, but to model and plan exactly what it
is we are tyring to do; the tools and raw ingredients will follow.

jim


__________________________________________________________________________
Jim Croft ~ jrc at anbg.gov.au ~ http://www.anbg.gov.au/people/croft.jim.html
ph 02-6246-5500 ~ fx 02-6246-5248 ~ GPO Box 1600 Canberra ACT 2601