This page was inspired by the XKCD cartoon Self-Description. If you've not seen this cartoon already, take a look now, read the hover help on the cartoon image, and think about how it works. How could such a cartoon be drawn, or verified? Are there many possible "correct" forms? Or just one? If we create an "incorrect" form, and then let the different gauges try and adjust to compensate, does the resulting image converge on a "correct" form?
The image at the top sets up "permanent" black/white pixels. Try hitting "Set to cartoon 688", then move the sliders around a bit to see what happens in the result at the bottom. Then press the "Start automatic recalculation" button and see where it ends up. Then tweak the sliders again, and start the automatic recalculation cycle again. Enjoy!
Important note: this page uses SVG to render on Firefox (performance is OK) and Google Chrome (performance much better). It uses VML on Internet Explorer and the performance is TERRIBLE. Therefore this page is NOT RECOMMENDED for use with Internet Explorer. Google Chrome IS recommended. Other browsers on other platforms will select appropriate renderer layers, and I can't predict what the performance will be, so take care.
The following "gauges" will then be composited with the above "permanent ink" to product the result shown below.
Now compare your results with the original cartoon. Note that this page simplifies the cartoon slightly, so don't expect an exact match!
That's all, folks!