natrius
commented
3 months ago I tried to find something for that problem. While researching i found this article https://storytellingmitdaten.com/die-richtige-verwendung-von-farben-in-datenvisualisierungen/ that suggests that no more than 6 colours should be user because above that it gets hard to distinguish. But you have 6 - so i tried where its possible to choose 9 colours https://colorbrewer2.org/#type=qualitative&scheme=Set1&n=9 and it looks easy to distinguish in my opinion.
Might be worth using that?
The color palette is problematic imho.
The problem may be more or less apparent depending on your eyes, but also depending on different monitors. While I am not color blind, and I have a good quality monitor (for photography), I find it very difficult to discern the reddish colors.
https://zlant.github.io/parking-lanes/#19/49.78873/9.93285
Even with unfolded legend:
The Peterplatz left side seems to be pinkish (residents), while in the Bergmeistergasse it looks between no-stopping, not applicable and residents.
Also here
the parking lot area has a hue I cannot find in the legend. The next similar would be no stopping. This is partially due to the transparency of the rendering and mixing with the gray underneath.
Suggestion: The colors should be well separated, without being a random swatch of colors. Green should be "free"/no payment, blue for payment, and red-yellow for "not allowed" in general.
"Fire red" for absolutely no standing like a fire lane, red for no stopping, yellow for loading only (which also seems for this use-case equal to no parking).
So, "for customers" will be likely not somewhere that you could choose for a trip into the city, so should be somewhere between green and red, avoiding the above used colors, so maybe a dark yellow, i.e. brown?
"private" parking lots are missing (see #150), they could be even darker than customers (maybe brown-black?)
Free parking should clearly be the brightest green (the one that is currently used for customers). All restrictions (time limit, customers only) should tend to a darker, or more blueish, or reddish color.
The "not applicable" is unclear (possibly a tagging error) - it should have the most "smells funny" color: glaring pink. (Maybe dotted or so to indicate a data problem)