[UPHPU] Genetically coded website
Justin Giboney
giboney at giboneydesigns.com
Mon Oct 22 13:36:41 MDT 2007
justin wrote:
>On 10/22/07, Craig Jackson <CJackson at cirris.com> wrote:
>
>
>>>Except that bright red + green + blue = white...
>>>
>>>
>>The average of #FF0000(255 0 0) #00FF00(0 255 0) #0000FF(0 0 255) = #555555(85 85 85)
>>That comes out with a gray color, which is why the color will not be white. That is only a compilation of three extreme color votes, but What forces the colors to turn brown though? Is it just slight descrepancies in which colors are chosen more? Don't know if this is very relevant, but just wondering how the color will turn brown?
>>
>>
>
>yeah. i was wrong about brown... (remember my disclaimer about it
>being late). it will go to a muted combination of the two highest
>voted colors, which in this case approached a brown/purple (~
>#550055). it could have just as easily evolved to dark teal (~
>#005555) or dark mustard (~ #555500).
>
>the distribution won't be exact. but right now it looks like a fairly
>even split between blue and red. since green wasn't quite as popular,
>it died at about gen 57. without any green in the mix, it gets closer
>to the red/blue split (in this case, heavier on the red) which makes
>brown.
>
>actually you can throw out anything past gen 40, because there's
>really no difference to the naked eye past that point.
>
>justin
>
>
So would using HSL then converting to RGB (for web browsers), and using
crossovers instead of inheritance, be better?
The only thing i can't get over, if the real way to do it is to randomly
generate items in a pool, run through each item, compare them, and
generate a new pool, then how do I make a site professional (good
looking) enough, yet stay ever growing with customers.
Justin
More information about the UPHPU
mailing list