I've had two problems today when creating the pattern "Illumn'd Manuscript" -- twice, colors were changed. In the first case, #B08D34 was changed to #B08D35; in the second case, #4C3C3C was changed to #4D3C3C.

When I noticed that one color had changed, I edited the pattern successfully. Then I noticed that the second color had changed, but I was unable to edit it successfully. I'd paste in the correct color again, but when I paged down to click the group name and then pressed Add Pattern, the color had changed on the finished pattern. I tried it several times.

I've been using Copy / Paste for both these problems, so it's not just that I enter the wrong hex codes.

I'm using Firefox 3.0.10 on a Windows XP machine.

I'm having the same problem. And not for the first time...
But the colours didn't change dramatically, so I didn't bother too much :)

I may have found the problem. If I paste in a hexcode and then hit Tab, the hex code changes. For example:

111111 [tabkey] 121212
AAAAAA [tabkey] ABABAB
ABCDEF [tabkey] ADCEF0
So (at lease in these cases) pressing tab causes one bit to be added to the R B and G componenets.

I'd just stop using the tab key, except I use it so extensively in Adobe Illustrator to see the effect of changing parameters, that I don't think I can break the habit.

Again, Firefox 3.0.10 and Windows XP.

I'm having even more problems today. I paste in a color, and even hitting tab doesn't transfer it to the selected COPASO color. I finally was able to get it to work by pressing Ctrl-Shift after pasting in the hex code. ???

Firefox 3.0.10, Windows XP.

There's a bug in the applet that does the automated colour-model value conversion; the last place I expected to run into this common bit of bad code was on this site. Because the 'helpful' colour creation area wants to be able to accomplish two things at once (dynamically updating results as-you-type, and 'converting' between RGB/HSV/HTML/CMYK) it loses the ability to do either with any amount of precision.

For example, I've been trying every way I know how to finagle the New Colour submission to accept (232,29,0). I can get it to maintain the typed values in any one colour model (RGB being, oddly, the most troublesome) until I go to the Add button, whereupon the inaccurate conversions either mistake it for a nearby, existing colour, and it won't let me add it as new. I've generated three distinct erroneous matches this way, none of which in their quoted values match my inputted values prior to the Add button and subsequent miscalculation of values.

The only way to solve this is going to be for the site to employ competent coders with experience in colour science and colour representation, and re-work the site's mechanisms so that a user chooses a colour model to input values to which will take precedence over nearest-neighbour matches when that colour is then mapped into a model with a different gamut, and/or can recognize unique formulae within intended spectra.

@Restroom, we're a pretty polite group here. I don't think your criticism of any specific individuals belongs in an open forum -- that should have been handled confidentially via a message to the site owner. Stating a problem is exactly what this forum was designed for; but using value-judgment words such as "bad" and "competent" is out of place.

Thanks for the feedback restroom. To be clear we don't employ any decent coders... only exceptional and talented coders. The code isn't the problem... it's the color science. We're always more than happy to get support in the scientific and precise color areas.

If you have some specific information to share, feel free to post it here or send me an email.

Cheers,
+D

Chris is a seriously good coder, stuff just pops up here and there. No such thing as an app that can't be improved, no matter who made it ;)

Darius, I'd bet that someone in your building at Microsoft has a great set of equations for translating between color spaces, maybe you could ask around?