Introduction of waterway=pressurised

[flacombe]

Expected behavior

Render waterway=pressurised [eventually filter with specific usage=, tunnel= or location=* values] as a thin blue line cased in strong grey [or any colour suitable for render consistence] solid lines.
Use casing dashes for tunnels

Waterway=pressurised is a new value to map features where water flows under pressure (which weren't mapped so far)
https://wiki.openstreetmap.org/wiki/Tag:waterway%3Dpressurised

Actual behavior

No rendering of any waterway=pressurised

Links and screenshots illustrating the problem

https://www.openstreetmap.org/way/243333033 (overground penstock)
https://www.openstreetmap.org/way/243333009 (underground tunnel)

[matkoniecz]

https://taginfo.openstreetmap.org/tags/waterway=pressurised has 127 uses worldwide, it is too early to render it

Please, open a new issue once usage is significant (#1630 has some discussion about what is "significant usage") but it is certainly counted in thousands (except extreme outliers like capital=yes).

[flacombe]

Hi

I don't get how it can be more widely used if renders and editors make it bogus and unconsistent with other tags.

There will surely be editing wars and misunderstandings if render doesn't match with wiki.

Can you take this in your scope prior to wipe months of work please ?

[matkoniecz]

if render doesn't match with wiki

Note that renderers are not obligated to match the wiki, especially "proposed rendering" section.

Any wiki page that has something like "renderers must...", "rendering of XYZ is obligatory", "XYZ must not be rendered" is wrong.

If one spots page like this and for some reason is unwilling to fix it - please notify me at https://wiki.openstreetmap.org/w/index.php?title=User_talk:Mateusz_Konieczny&action=edit&section=new or notify someone else about such mistake.

[imagico]

For clarification: Historically absolute use numbers have not been considered a relevant criterion for deciding on use of a tag or not. If a tag is consistently used by a broader mapping community, has a verifiable meaning and rendering it in a certain way creates positive mapping incentive are the main factors.

@gustavecha - the tags you proposed to consider in rendering are part of a new and highly complex tagging system. It would not be responsible to push and further interpret this system by rendering it here in a certain way before the global mapper basis had a chance to independently decide if an how to practically adopt the proposed tags and how to further develop their meaning. History has shown that waterway rendering is highly prone to influence mappers to tag for the renderer - see #2346 for a simple example. Adding further not well thought through stuff has the potential to do a lot of damage.

[dieterdreist]

2018-04-19 16:00 GMT+02:00 Gustave <[email protected]>:

I don't get how it can be more widely used if renders and editors make it bogus and unconsistent with other tags. Can you take this in your scope prior to wipe months of work please ?

it is generally a problem that most people only map things if they are shown on the map (newbie and occasional users typically only map things which have a preset). In this case of a new waterway value, I agree with the team's decision to wait for more usage.

[matkoniecz]

I don't get how it can be more widely used if renders and editors make it bogus and unconsistent with other tags.

1. note that it threshold to support in editors is typically much lower ("considered as a good idea to tag"), it may include rendering in editor map style

2. anybody may make any rendering, it includes rendering of waterway=pressurised

[matkoniecz]

For clarification: Historically absolute use numbers have not been considered a relevant criterion for deciding on use of a tag or not. If a tag is consistently used by a broader mapping community, has a verifiable meaning and rendering it in a certain way creates positive mapping incentive are the main factors.

It is certainly not a sole and only criterion but very low usage (say, below 1000 worldwide) usually indicates that tag has (currently) no broad support from community of mappers.

To explain:

Please, open a new issue once usage is significant

This is not a promise that tag will be rendered, just that at this moment waiting to see how this new tagging scheme is considered by OSM community is IMHO preferable rather than rushing to implement it.

New tag used 50 times worldwide will (except really extreme cases) not be rendered in this style. New tag used 50 000 times worldwide may be rendered depending on other factors, most important ones mentioned by @imagico .

[jeisenbe]

Now used 2172 times on ways. Related to rendering man_made=pipeline and tunnel=flooded

[matkoniecz]

I closed initially based on low usage count, so I will reopen.

Note that is does not mean that it should or will be rendered, but that it does warrant review larger than "according to taginfo it is basically unused"

[jeisenbe]

Current usage:

1051 ways with man_made=pipeline and location=underground - http://overpass-turbo.eu/s/Mdx - mostly in Europe, but there are a good number in Japan and some on every continent.

667 ways with man_made=pipeline and location=overground - http://overpass-turbo.eu/s/Mdz

22 ways with man_made=pipeline and tunnel but without location

120 ways with tunnel but without man_made - http://overpass-turbo.eu/s/MdB

44 ways without tunnel=* or man_made=pipeline - http://overpass-turbo.eu/s/Mdt - most of these still have "usage=penstock" or another "usage" tag, and many have "location=underground" or "location=overground".

The location=overground could be rendered similarly to other types of man_made=pipeline above ground - see issue #640

There are 1193 features which are underground pipelines or tunnels would need to be considered separately, since we probably will not render most underground pipelines.

[flacombe]

Thank you Joseph for those figures

22 ways with man_made=pipeline and tunnel but without location

This is pretty uncommon situations. The only possibility is to find a pipeline inside a human accessible tunnel, it exists only in a few places.

120 ways with tunnel but without man_made

This sounds right : no pipeline but a pressurized tunnel

44 ways without tunnel=* or man_made=pipeline - http://overpass-turbo.eu/s/Mdt - most of these still have "usage=penstock" or another "usage" tag, and many have "location=underground" or "location=overground".

man_made=pipeline is required for usage=penstock, which can be a precious hint to fix them

Ok with #640

As mentioned before, waterway + usage=headrace or waterway + usage=penstock have often great water flows. I'd be in favour they appear with a light blue line, eventually cased or dashed even underground.
I agree that other underground pipelines shouldn't be rendered here.

[jeisenbe]

Would you be interested in making a PR to add this feature, along with `man_made=pipeline`, once you are finished with the power=portal PR?

[jeisenbe]

#4070 will add pipelines with location=overground, which will address some waterway=pressurised features

[flacombe]

Tests and demos in #4070 look very promising and are great for pipelines
Thank you Joseph.

[jragusa]

@flacombe: what about the rendering of waterway=pressurised since pipeline and flooded tunnel are now rendered ?

[flacombe]

Hi @jragusa
Recent rendering of man_made=pipeline and your good contribution on tunnel=flooded make this issue almost solved, thank you.

Just to be sure, is tunnel=flooded that is rendered or the combination waterway=canal + tunnel=flooded?
Perfection would be to distinguish casing or dashes upon waterway value, but showing waterway=pressurised + tunnel=flooded the same as waterway=canal + tunnel=flooded is enough

However I completely second what is proposed in #3354

All the best

[jragusa]

Currently, only waterway=canal + tunnel=flooded are rendered.

[flacombe]

Well, it would be great to extend it to waterway=pressurised.

pressurised or canal don't inform us on tunnel size.
These pressurised tunnels are +4m wide
https://www.openstreetmap.org/way/310730345
https://www.openstreetmap.org/way/231511649

While this canal is 1.5m wide
https://www.openstreetmap.org/way/371282336

[jeisenbe]

Only one of those three ways includes a width= or diameter= tag. How do you know the size? Is this information in an external public data set?

[flacombe]

First of all it's possible to measure ones with open air outlets, at downside outlet.

First pressurised tunnel corresponds to this map showing internal diameters (i missed to add diameter on osm object).
https://wiki.openstreetmap.org/wiki/File:Hydropower_underground_water_drains_map.jpg
Second one was got on a similar map.

Last one was measured directly on outlet (and missed to add it as well)

[jeisenbe]

There are now 176 waterway=pressurized with tunnel=* but without pipeline=* according to http://overpass-turbo.eu/s/TmH -

Many are 5 to 20 km long hydroelectric features which cut across the topography. I'm not sure if these should be rendered on a general-purpose map:
E.g.:

[flacombe]

It's up to maintainers, depending on what they want.

These ones are really large ones, water flows at 30m3/s or more inside.
We can usually see them on 1:25000 topographic maps.
When following overground pipes which appear on osm carto, people may ask "where does the water come from?". You can see how it looks on openinframap.org currently showing pipelines only.