Talk:Featured tile layers/Guidelines for new tile layers

With respect to the discussion on "non-commercial is preferred" in SWG. Perhaps rather than tying the preference to commercial vs non-commercial, it should be "comes from within the OSM community or not". One of the main reasons for having multiple tile layers is to show and advertise the power of what can be done with the data. At the same time it could also show off the power of the community and how easily one can create great new styles from the data by preferring . This would automatically give a certain degree of divide between commercial vs non-commercial however without explicitly saying so. It would also be easier to justify why that would be a relevant criterion. --Amm 00:40, 19 February 2011 (UTC)

What counts as "within the OSM community"? If they are using OSM data, and wanting their tiles to be on osm.org (2 requirements) then surly they are part of the community? - LastGrape/Gregory 22:42, 16 May 2011 (BST)

It seems as if guidelines to retiring tile layers are missing. So once in, always in?

If that is not the case, I don't quite understand why the cyclemap layer is still available. - User:SimonPoole 18:37, 20 May 2011 (MET)

Use of non-OSM data

There is not a particular practical need for this at the moment i suppose but i think to be in line with the mission of the OSMF the guidelines should include something like the following:

• In the must conditions: A tile layer must not use any non-free data.
• In the should conditions: A tile layer should not use non-OSM data where OSM data exists.

--Imagico (talk) 10:06, 22 May 2018 (UTC)

I agree that this would be a sensible and hopefully uncontroversial improvement. Is there any established mechanism to update the guidelines? They seem to have remained unchanged over the past years. --Tordanik 16:21, 23 May 2018 (UTC)

Since the current page is not identical to the text approved by the OSMF board originally anyway i think we can edit it if there is consensus on the change. This does not make it part of the official policy of course. Adopting a modified policy officially by the OSMF should probably go through the OWG.

Formally it would probably be necessary to more precisely define non-free data, in particular if this includes non-commercial licenses. --Imagico (talk) 16:55, 23 May 2018 (UTC)

I think it's clear enough for the most part (non-commercial data is generally considered unfree). But in other contexts (e.g. the contributor terms), the OSMF has relied on the Open Definition. So that might be a good resource if there's a need for more precise definitions. --Tordanik 16:29, 25 May 2018 (UTC)

Guidelines for current tile layers too

These should be guidelines for current tile layers too, not just new ones. Jidanni (talk) 22:35, 21 September 2018 (UTC)

The guidelines already apply to the existing tile layers, there is just no defined mechanism to evaluate this. But tile layers have been removed in the past (Mapquest) and if anyone thinks a tile layer - due to changes - no more meets the requirements you should bring this up on a suitable communication channel to be discussed. I don't see this being the case for any of the current tile layers though. --Imagico (talk) 09:28, 23 September 2018 (UTC)

International borders are a must, to prevent accidental deaths

International borders are a must, to prevent accidental deaths, and thus liability issues due to their purposeful ommision. Please see https://github.com/openstreetmap/openstreetmap-website/issues/2002 . Jidanni (talk) 22:38, 21 September 2018 (UTC)

I can just imagine the [fictional] news: "Cyclist shot dead due to hacked OpenStreetMap" One British cyclist was shot dead and another seriously injured as they mistakenly crossed the India Bangladesh border. "All the map showed was paddies and drainage ditches. How were we to know?" the injured one said from his hospital bed. OpenStreetMap spokesm iten said the origination, which provides free online maps to the public, was investigating a possible hack, where black hat editors had repositioned sections of the border of their rival neighbors. However as of late Thursday they were still unable to find where the new sections had been placed.

A spokeswoman from a North American cyclist association urged the American and Canadian governments to block access to osm.org until large sections of their common border were restored. "It is purely a safety issue." she said.

Last month users of the OSM brand web maps were also given a stern warning after wandering too near the border after getting off a bus in a central Asian country. [end fictional story] Jidanni (talk) 02:36, 23 September 2018 (UTC)

The OSMF terms of use (https://wiki.osmfoundation.org/wiki/Terms_of_Use#IX._Disclaimers) exclude any such liability - just as essentially any other map provider. Relying on the accurate representation or completeness of a map is universally ill-advised and there is nothing special about administrative boundaries or OSM in terms of risks. You can also drive along a road and drown because you fail to see the river the road crosses because it is not on the map and this can happen to you on a non-OSM map just as well.

I would also suggest you do not try to steer a discussion on a matter with made up fantasy scenarios - this is not in support of a productive discussion. --Imagico (talk) 09:28, 23 September 2018 (UTC)

Making up fantasy quotes is unwelcomed in discussion like this one Mateusz Konieczny (talk) 17:51, 23 September 2018 (UTC)

OK but that river wasn't in the database in the first place. And even if not a legal issue, still a moral and reputation issue. Jidanni (talk) 10:24, 23 September 2018 (UTC)

No, limited accuracy and incompleteness are universal risks that apply to any geodata use - both OSM and non-OSM. What to show and what not to show in a map are subjective design choices that depend on the purpose of the map (and the guidelines deliberately put no requirements on what this purpose can be). Claiming moral absolutes in this domain mistakes cartography for something that it is not. --Imagico (talk) 10:50, 23 September 2018 (UTC)

In short: this proposal is absurd and bad. Also even in case that it would make sense: it is hilarious that you care not about dead people but about a liability Mateusz Konieczny (talk) 17:49, 23 September 2018 (UTC)

Fine.

On the public facing website, e.g.,

• Hikers choose the hiking layer.
• Cyclists choose the cycling layer.
• Drivers choose the transport layer.
• etc.

Some will miss the e.g., tag=landmine and get blown up because it has been removed from their layer for aesthetic reasons. But that's OK. It's their own fault. Besides, one can't expect us to keep track of all the world's landmines anyway. Case closed. Jidanni (talk) 23:35, 23 September 2018 (UTC)

"One British cyclist was shot dead and another seriously injured"

I'm a British cyclist. I know lots of British cyclists. I am happy to report that British cyclists are not idiots and are perfectly capable of not straying across international borders. Please don't judge others by the standards of your own extreme muppetry --Richard (talk) 08:48, 24 September 2018 (UTC)

OK thanks everybody. Got this mail today:

From: support@thunderforest.com

I've added borders for all zoom levels to the OpenCycleMap layer today. [And] transport layer [soon]

Jidanni (talk) 02:30, 3 October 2018 (UTC)