If you set up a categorical hierarchy that is essentially MOCs of MOCs of MOCs, it would be cool if the nodes were color coded based on degree of separation from top MOC.
And for nodes that are linked to by multiple MOCs, their color could maybe be the multiplied rgb value.
I like this idea but I think that color coding could be more useful for tags, or link types than for degree of separation.
I think degree of separation would be best represented in a color attribute like opacity, color saturation, brightness or something like that. That way your color coding can represent multiple things at the same time. i.e. pick a color to represent something like a tag or link type, and then as you separate to further degrees of separation, then that same color can decrease in color saturation (i.e. become more grey).
I see what you mean. Good idea.
Your spin on it has more of a static defined color per tag, which I like, although it has some drawbacks.
If the system is more dynamic and based on what is viewable at the moment in the graph, the colors could be kept as different from each other as possible, however it wouldn’t be consistent.
In terms of whether to apply this to tags or top hierarchy MOC-like notes , I am starting to think your idea with tags may have some benefits. Honestly, my workflow has changed a good bit since I had this idea.
I do still like my original thought of the notes that have multiple colors (tags) get those colors multiplied, which would approach white for those that have many.
Thanks for the comment.
Agreed, I love the idea of dynamic automatically chosen color coding. Color coding is nice b/c it allows me to see things quickly at a glance, but it becomes more work because I have to decide what my color code is, and I have to memorize it, and manually maintain everything to match the color coding. If a color coding can be applied automatically then that removes all of that friction. Plus, the algorithm can choose colors that are as visually separated as possible. But like you say, the drawback is that the colors would no longer be consistent.
Honestly there are pros/cons to both approaches (dynamic vs static color coding). Ideally, it would be wonderful to have these be user configurable settings within the graph view. That way we can switch “modes” on the fly and each person can pick whichever approach matches their workflow.
My approach would be kind of a hybrid. Statically defined colors but dynamically applied saturation to represent a trait like degree of separation.