-
Notifications
You must be signed in to change notification settings - Fork 819
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Render greenhouse buildings as light brown #2538
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The problems with being too dominant are not unique to greenhouses, but an issue with all buildings when they're in that kind of density
Do you have similar examples? |
DeltaE to mud is 3.2 which is too small for comfort. Similarity to landuse=residential is also too close for a completely different class of feature - you can see that in your example where the user might easily read this as yet another class of urban land cover. I don't think you are going to be able to squeeze in another area color in our already very overcrowded color scheme, especially with relatively weak and subtle colors. |
@pnorman These are not too similar, one should try with z13-z14 to make it comparable. On z17+ there are much more elements as you can see and buildings are not as visible as when there are no icons. Also the building color on z15+ is made lighter than on earlier zoom levels. |
@imagico At z13-z14 it looks like something more important than towns around, which is worse kind of misguiding users than what you suggest. Brown is basically only building-related and already used in variants, so I'm not afraid of extending it more - it will make using brown even more clear and consistent rule, which would be good. |
I disagree on both the assessment of the specific situation and the interpretation of the meaning of colors in general. Colors are a continuum, there is no such thing as a line between brown colors and non-brown colors. We have three building area colors currently in this style, all of which are more similar to other non-building area colors than to each other (see also #2353) so there is no such thing as a building-brown that can be extended with additional variants. While i think it would be good to render building=greenhouse in a distinct way (though not necessarily via color) i don't think the sample area presented shows a particular problem in its current appearance. I have not been to this particular area but in general greenhouses like these are very specific, prominent and dominating features in a landscape and it is not unreasonable to render them significantly more prominently than a generic landuse. |
There's also no such thing as a line between green and non-green colors. Green however "should be reserved for vegetation related features". Yet it's also used for sport and leisure, which are something else, and nobody complains it could be misunderstood. And green has even no continuum, as brown does, we have just a set of unrelated green shades. Doesn't it disturb you as a bigger problem? |
The vegetation related green tones in this style form a distinct group in color space, i.e. they are (with a few exceptions) closer to each other than to the rest of the area colors. Some of the recent changes (playground and pitch/track) blurred the boundary by moving not strictly vegetation related borderline green tones towards the center of vegetation related greens. I explained this also in the original version of the guidelines document on: |
Before that the problem was playgrounds were too close to the pools (the same problem for blue this time), so nothing new arised and the problem is milder now, but it's side tracking the issue. Let's get back to the core - sport was green also before any shading, and it's not more intuitively green than buildings brown. Green set is blurred now and was blurred before. Brown is not and was not - hence "features rendered with similar color should actually be similar in purpose and meaning" (not my words, but I agree) is more in favor of brown than green. |
I don't see this discussion leading anywhere - i explained my viewpoint, in particular
The color differences can be calculated by anyone interested - both these and the green and blue tones for comparison. |
Maybe I don't understand your vocabulary - if shades are "colors" for you? Then all light colors and all dark colors would be alike. That doesn't make sense for me. |
I don't think i talked about shades in this context. |
The only difference here is using lighter kind of brown - but still brown - and you claim it will be too close to other light colors. So I assume you were talking about lightness, not about colors. |
deltaE measures visual differences in both lightness and chroma/hue, see https://en.wikipedia.org/wiki/Color_difference. If anything the difference in lightness is more meaningful because display devices provide better distinction in lightness than in chroma/hue. This is why the wood/water color combination is so annoying despite the colors having a relatively large difference (deltaE is 27.5). |
I think that what is a bit forgotten in the whole discussion regarding colors, lightness, chroma, deltaE or whatever other measure one would desire to use to describe differences, is that people not only "read" the contents of a map by looking at "color", but use other (visual) clues as well. E.g. greenhouses tend to:
So there are more clues that set apart certain objects and allows recognition. This is just one example, similar things hold for other types of objects. |
...and locating it on mud would be highly surprising (while water in the woods is quite common case), so what's the use of measuring deltaE in such case? On the pictures there are much more typical and that's what is interesting, not the corner cases. [UPDATE:] On residential area it's a bit less visible than normal buildings - but that was the point - yet visible enough for me: |
+1 I don't think the proposed solution is the best solution for the given problem. |
sent from a phone
… On 21 Mar 2017, at 22:23, Matthijs Melissen ***@***.***> wrote:
The problems with being too dominant are not unique to greenhouses, but an issue with all buildings when they're in that kind of density
+1
-1, it's a feature that dense building structure leads to dense rendering
|
Resolves #738.
Near Hague, z13:
Before
After
Warsaw, z17
Before
After