What I found really useful about this was how the presentation laid it out in a way where I could really understand it. For the first time, I got a simple set of concepts behind choosing colors. Without going back to the article, just what I remembered:<p>1) Choose complements.
2) Mix them with each other until they dull down a bit
3) Make some other neutral color ( possibly by mixing them in equal parts ? ) and then you have 3 colors, and you've got a basis.<p>( OK -- just considering is there any connection between color theory and the basis of a vector field? )<p>So what I liked about this was it taught me something I didn't know.<p>I accept there are a lot of comments by experts on here about how this simplified presentation is wrong in such and such a way. I don't think that decreases its utility for me.<p>To a novice, just having something tangible to understand is the first step. I understand how to experts, it seems like the wrong advice. But I felt I understood it, and that's invaluable.<p>If you don't know what I mean by this, maybe it's better explained by something a professor said to me in grad school, "The trick to teaching is to trick people into thinking they understand it. Don't show them all the complexity. Show them a simplified model. Let them come to it in stages."<p>I think that's true. It stuck with me and it seems true. It also matches what I see my science education was like looking back -- a presentation of a series of models, over a number of years, that gradually layered in more and more complexity, but at a pace that was comfortable. This "chain of models" mostly agreed with each other, but the points where they contradicted each other didn't to my memory negatively impact learning, instead they served as signposts that made each model more distinct and aided in remembering them.<p>Also, there is this great feeling of being able to say, "Yes from the perspective of model X that is true, but when you factor in what model Y says, you get a different picture. Here, let me show you." That always sounds like a true expert. I like that and it also seems like having a bunch of models, not all in perfect agreement, in fact strengthened by the points where they do differ, gives overall a far richer picture of the topic.<p>The strongest idea that stayed with me from my science education was that "everything is a model" and "none of them are true" but some of them are useful. Different models, with different levels of faithfulness to evidence.<p>My main critique of what I will label the "dismissive expert answers" is that they posit a inflexible model of the "absolute truth", which, I say is, given by itself: not that useful ( at least to learn the topic ), and not really true ( because well, no model can be ).<p>I am sorry if this length comment is more suited as a blog post. Maybe I should just make blog posts instead of comments and link to the article. Maybe that's better than anyone. I'm not trying to offend any experts or do anything ego based, just push back against the what I feel were dismissive answers that missed another side of the value of this article. I wanted to share what I thought that other side of the value of this article was.<p>It does seem like a highly personal comment, just from my point of view. I hope it's been useful to some others to read it!