Skip to content
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

natural=mud doesn't render in rivers #115

Closed
chillly opened this issue Aug 14, 2013 · 24 comments
Closed

natural=mud doesn't render in rivers #115

chillly opened this issue Aug 14, 2013 · 24 comments

Comments

@chillly
Copy link

chillly commented Aug 14, 2013

natural=mud now renders around the coastline, but not in rivers. The example below shows the join between coast and river and the cutoff is obvious.

Example: http://www.openstreetmap.org/#map=16/53.6995/-0.4381

@CloCkWeRX
Copy link
Contributor

Is this simply because the riverbank is a closed polygon, and

      polygon-fill: @water-color;
      polygon-gamma: 0.75;

renders after/conflicts with

    polygon-pattern-file: url('symbols/mud.png');

?

Or is it just missing an appropriate comp-op: overlay as per https://www.mapbox.com/tilemill/docs/guides/comp-op/ ?

@gravitystorm
Copy link
Owner

Whatever the problem is, it's not a comp-op issue.

@mrwojo
Copy link
Contributor

mrwojo commented Oct 18, 2013

Move the glacier style from the top of #water-areas and it works as expected.

It's the order of symbolizers attachments, I suppose. The glacier style at the top makes ::natural before the ::waterway attachment of riverbanks. Mud is ::natural and is therefore drawn over by ::waterway.

@mrwojo
Copy link
Contributor

mrwojo commented Dec 18, 2013

I have the above fix for this in 26c9d4e. But I also have questions:

Should natural=glacier also render over riverbanks? The current behavior renders glaciers under riverbanks (like natural=mud; they're the same natural attachment).

Also, what about other water areas? Should mud/glacier always render over features like natural=water and landuse=reservoir? Or should it follow the usual polygon sorting patterns (z order then area)?

@dieterdreist
Copy link

2013/12/18 mrwojo notifications@github.com

Should natural=glacier also render over riverbanks? The current behavior
renders glaciers under riverbanks (like natural=mud; they're the same
natural attachment).

Also, what about other water areas? Should mud/glacier always render over
features like natural=water and landuse=reservoir? Or should it follow the
usual polygon sorting patterns (z order then area)?

natural=mud is a really dumb tag that doesn't quite fit into the system. +1
for not supporting it at all. "natural" describes geographic features, a
glacier is one, "mud" isn't. Better use landcover or surface as keys for
the mud value and use something like natural=wetland with the appropriate
subtag for "muddy areas".

@malcolmh
Copy link

malcolmh commented Mar 6, 2014

I note that the Wiki page for natural=mud now redirects to the natural=wetland page, so I assume that this is the preferred tagging. However, it appears that the renderer does not differentiate between wetland=... cases, rendering the reed-like pattern for all. This is inappropriate for those cases where there is no vegetation, such as wetland=tidalflat. If that could be fixed, perhaps using the same rules as natural=sand/mud to render the same pattern for wetland=tidalflat + surface=sand/mud, then the need for the old natural=sand/mud goes away.

@dieterdreist
Copy link

Am 06/mar/2014 um 20:47 schrieb Malcolm Herring notifications@github.com:

then the need for the old natural=sand/mud goes away.

natural=sand / mud is quite a strange tag, fitting not really in the tagging system (it should be surface or landcover). We shouldn't encourage its use by rendering it

@Rovastar
Copy link
Contributor

Rovastar commented Mar 7, 2014

Feel free to contact the tagging mailing list where they can direct you to how do proposals and get the voting changed tags that you disagree with.

Never used natural=Mud (very onclear on the wiki what should be used - nothing on the wetlands page) but have used Natural=Sand is used a fair bit even if it was just for mapping golf courses.
I do not see the reason not to change this.

Regarding the wetland tag. We do not store this in our database so it is unlikely this will be changed anytime soon. (And for wetlands=mud only has 16 hits on taginfo....)

@malcolmh
Copy link

malcolmh commented Mar 7, 2014

On 7 Mar 2014, at 17:12, Rovastar notifications@github.com wrote:

We do not store this in our database

Who is “we” and which database are you referring to?

@Rovastar
Copy link
Contributor

Rovastar commented Mar 7, 2014

The database we use which generates the rendering
"we" = this project.
See the project description:
https://github.com/gravitystorm/openstreetmap-carto

@dieterdreist
Copy link

Am 07/mar/2014 um 18:12 schrieb Rovastar notifications@github.com:

but have used Natural=Sand is used a fair bit even if it was just for mapping golf courses.
I do not see the reason not to change this.

this is a little bit OT here, but there are nice tagging details for golf courses in the wiki, suggesting golf=bunker for these, together with natural=sand and the latter explicitly for the renderer (that's also what it is IMHO: tftr)

@Rovastar
Copy link
Contributor

Rovastar commented Mar 7, 2014

:nod: Yeah After a discussion Harry Wood and I added those descriptions for golf course to the wiki a year or so ago. As previously people were mapping bunkers as beach. It obviously is sand. :-)

@mrwojo
Copy link
Contributor

mrwojo commented Mar 11, 2014

@Rovastar, your messages aren't showing up at GitHub, which makes this conversation a little hard to follow there.

Who is “we” and which database are you referring to?

The database used for rendering is defined by the openstreetmap-carto.style file. If a key isn't in there, then it can't be used (for now).

But it looks like we have everything we need to render based on wetland=* and/or surface=*. We're currently not using those keys but they should be there in the database.

@Rovastar
Copy link
Contributor

Stupid github

@Rovastar
Copy link
Contributor

Hopefully my messages appear now.

@malcolmh
Copy link

I have opened an issue to address the wetland=* rendering style: #387

@matthijsmelissen
Copy link
Collaborator

Is this the same issue as https://trac.openstreetmap.org/ticket/5020?

@malcolmh
Copy link

Yes it is.

@matkoniecz matkoniecz changed the title natural:mud doesn't render in rivers natural=mud doesn't render in rivers Sep 16, 2014
@malcolmh
Copy link

malcolmh commented May 8, 2015

If the waterway=riverbank tag is changed to natural=water, then the rendering is correct. These two tags ought to produce the same result, but don't.

@imagico
Copy link
Collaborator

imagico commented May 8, 2015

I see no indication of areas tagged natural=water and waterway=riverbank being treated any differently - if they are please open a new issue and show an example. Note any special issues with natural=mud will be solved by #1497.

This problem here (different handling of coastline and water areas in relation to landcovers) is still unsolved - it will require use of water polygons for the ocean rendering.

@malcolmh
Copy link

malcolmh commented May 8, 2015

Look here: http://www.openstreetmap.org/#map=14/53.7058/-0.5772. I changed one of the waterway=riverbank polygons to natural=water and the mud re-appeared.

@imagico
Copy link
Collaborator

imagico commented May 8, 2015

#1497 is not yet rolled out - if there was any difference before it should be gone after this change.

By the way - if you change waterway=riverbank to natural=water please always add water=river.

@malcolmh
Copy link

malcolmh commented May 8, 2015

OK, I will wait until roll-out to see if this issue is then cleared. Thanks!

@matthijsmelissen
Copy link
Collaborator

Closing this now, please let us know if there is still a problem after #1497 is rolled out.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

9 participants