I have read through the editing and import guidelines for OSM. Contributing our data would require a considerable number of hours to do properly. Doing it properly is the only way we're interested in contributing.

What kind of protection can be had for any data we might share?

Note: Receiving "notifications" whenever someone edits data we've contributed, with the option to revert the changes, does not interest us. Taking the time to share the data via the OSM guidelines is already a significant burden and we want a 'fire-and-forget' approach. But if it can't be protected from wrong (if well meaning) edits then we don't see much point in sharing in the first place.

Further clarification: There seem to be places where a portion of our data, or data product, has made it into OSM. We are fine with that (since we are not responsible for what someone might do with our data via OSM) but we have also noted subsequent changes to our data that has decreased it's accuracy. Hence my question. No point in contributing data when we have no time to play data-cop to keep it in good condition.

asked 29 Mar '16, 18:09

_jk's gravatar image

accept rate: 0%

edited 30 Mar '16, 15:59


Some thoughts: Adding a remark into description of each way saying that when the way was imported all the points were with accuracy of for example +-0.5m gives a good indication to future editors to trust the locations of the points and not to move them if not truly necessary.

Your imported data will for sure change regarding the tags on the ways and points as I am sure there are things people do wish to add that is not already tagged. Not knowing the details of your data it might be as simple information as surface, reference numbers, speed limits and access rights.

As more detail (smaller roads, footways, paths, speed pumps and all kind of strange things you did not intend to map youreselves) is added over time your imported ways itself do need to change by getting new nodes in points where added roads do link or where new features are.

(29 Mar '16, 20:33) RM87

The question has been closed for the following reason "Excessive trolling" by SimonPoole 30 Mar '16, 11:05

OSM is open, so I'm afraid you can't do that. A look at the CT may tell you more about contributing to OSM. https://wiki.osmfoundation.org/wiki/License/Contributor_Terms

permanent link

answered 29 Mar '16, 19:01

yvecai's gravatar image

accept rate: 9%


There's nothing you can do to prevent other people from making changes.

You can of course work with them to make sure they understand that their changes are harmful, but there's no mechanism to lock data, and probably quite a lot of resistance to the idea if you survey the community.

One way to do the fire and forget might be to make the data available under a broadly open license and let someone else do the work to integrate it (or maybe people will just use it as a reference). No ongoing burden there.

permanent link

answered 29 Mar '16, 18:15

maxerickson's gravatar image

accept rate: 29%

edited 29 Mar '16, 18:57


OSM has its own mechanisms of producing quality. These mechanisms will be different from yours, and the quality they produce will be different from yours. It is important to understand that OSM is first and foremost a group of humans who curate a world wide data set. Bypassing these humans - trying to share something "through" OSM where you and only you control the data - would be taking the most valuable aspect of OSM out of the equation. Not only is this not possible; it also betrays considerable ignorance about what OSM is in the first place, and what made it great. OSM is a project where data has been surveyed because official data wasn't made available, not a melting pot for official data from different sources. That doesn't mean that your data isn't welcome - au contraire! But your data is like a glass of water that you pour into the ocean that OSM already is. You can release it but you can't hold on to it after.

(29 Mar '16, 20:06) Frederik Ramm ♦

Why would uncontrolled mean poor quality data ? Why do you expect that people will alter the data you added and replace it with "poor quality" ? I feel insulted by this statement. I think almost every OSM contributor will try to improve bad data and when the data is good, it will remain unaltered. Please do not think that only your agency is capable of producing quality data.

(30 Mar '16, 07:15) escada

This may have started as an honest question, but it is completely clear that you are all now just feeding a troll.

Closing the question.

(30 Mar '16, 11:05) SimonPoole ♦

You are welcome to suggest ways to improve and or keep data quality. Simply locking it is not an option as already explained multiple times.

(30 Mar '16, 18:25) scai ♦

First of all, and most importantly, that's great that your agency wants to contribute your data, Thank you! I'm glad to hear that other groups are taking an interest in OSM, and I'm gland your group wants to do it properly.

I'm not very knowledgeable at all about data imports, after a bit of reading I suspect that you know more than me. Despite that, I would like to underline that it's important to consult with the local community before doing any imports.

As others have said there is not a technical mechanism in place to prevent detrimental edits. I don't think this would be possible to implement without preventing edits that are beneficial. OSM continues to improve everyday, and it's important to allow all objects represented in the database to continue to improve. OSM's solution to the problem of detrimental edits is relying on the editors ensure that other editors are improving the map. While well meaning edits may, as you say, be "wrong", on the whole the map improves with every passing day.

I don't quite understand what benefit you'd get if it was somehow possible to "lock" the roads and trails you are suggesting you import. The vast majority of edits are beneficial so preventing edits is only going to be detrimental to the map.

So to answer your question, as long as the import is accepted by the community, then you are welcome to "fire and forget" it, but it is necessary to trust the community to do the right thing with your data. They will.

permanent link

answered 29 Mar '16, 19:32

keithonearth's gravatar image

accept rate: 0%


As others have pointed out, one of the most important aspects of OSM is that the data in it is available to the general population to use, and even edit, if they see a need to do that. Your high accuracy data, which is certainly welcome, may have other properties besides positioning that will need to be changed by future mappers. The way your objects are tagged, for example, will almost certainly need modification at some point in the future because tagging methodology is constantly evolving. I've seen huge data imports from very respectable sources, state agencies usually, whose tags are completely wrong. The professional GIS specialists who contributed those data were probably convinced they had everything right too. But in the end, some one of us amateurs will have to sit down and correct their work.

(29 Mar '16, 23:58) AlaskaDave

@_jk, there is so much more to OSM than positional data. You "seriously doubt anyone is going to beat that", but there is more data that you may not have considered. What if a new path joins up to it later? No, not allowed. They add street lighting? No, you can't add that, this is _jk's private data. A route is run over your way, sharp intake of breath Accept that when you upload that to OSM, you agree that it can and will be edited. You can't put it on a publically editable medium then get upset when people publically edit it.

(30 Mar '16, 00:57) Deanna Earley

@_jk, Please don't take my comment as a reason to withhold your data. We want your data. I was simply pointing out that protecting it from the OSM community is not only not possible but not desirable.

(30 Mar '16, 00:58) AlaskaDave

@_jk, "Just thought we would look into contributing to OSM quality but it's not wanted, like, at all. I get that now." You misunderstand. It is very much wanted, but once you make it available, you can not control it what happens to it.

(30 Mar '16, 01:01) Deanna Earley

What are the reasons for all OSM data being so-so, and isn't it in your interest in making other data better?

(30 Mar '16, 05:27) emj

There is no such thing as perfect data. People can always improve upon your work (because you never bothered to map some detail, because a new path is added/blocked/resurfaced, etc). Suggesting that trails could remain unchanged for 10 years would be quaint idea in most parts of the world. Precise data is even more vulnerable to this kind of bitrot; not too long ago somebody mapped a golf course with centimeter accuracy, which sounds great until you realize that this accuracy was casually twarthed by vegetation growth and plate tectonics.

Therefore, no data is untouchable in OSM. This is not a weakness, this is how OSM thrives. It does come at the price of occasional temporary deterioration, but the balance stays overwhelmingly positive. The "how to lock some OSM data" question comes back with some regularity, and the answer from the community is pretty much always that it would be a bad idea and isn't supported.

I find it strange that you think that contributing your data to OSM would only be worth it if said data remains untouched by anyone but you. What are you trying to achieve by puting your data in OSM ? One typical reason is to make the data more broadly available, and that goal remains fullfiled even if some heavy-handed contributor moves a trail by a meter. The general public is still better off if trail data is available in OSM, even if some of it got deteriorated. They can still get at your "perfect" data if they know where to find it and don't need the extra context and improvements. Which source they choose depends on their needs.

If you want this import to be a one-time job but will keep improving your version of the dataset, a good solution is to document your import process so that anybody can redo the import/merge process later. OSM has plenty of QA tools, including some which compare OSM data to other sources.

OSM won't sit still for 10 years, or even a day. This can feel daunting, but this is the right thing to do. Embrace the flow :)

permanent link

answered 30 Mar '16, 00:08

Vincent%20de%20Phily's gravatar image

Vincent de P... ♦
accept rate: 19%


Plate tectonics goes faster than you might think. Centimeter-accurate positions will be thrown off by tectonics within one year pretty much anywhere on Earth. Half-meter accuracy will take somewhere between 5 and 20 years depending on where you are. If your agency aims to make very acccurate measurements on a multi-year timeframe, it should not laugh at tectonics.

(30 Mar '16, 08:59) Vincent de P... ♦

Many in the OSM community would take offense in the statement that we "garble data". We see the OSM model as a way to achieve supperior completeness and accuracy compared to other providers, and take quality seriously. There are occasional hiccups, but the general trend is overwhelmingly positive. As with all wikis, "it doesn't work in theory, but it works in practice".

If you want your data to thrive in OSM, you need to learn to give up your position of absolute authority and trust the community. OSM is not a hosting platform for your immutable blob of data : it will integrate it, work with it, improve it.

(30 Mar '16, 09:21) Vincent de P... ♦

Show us the data, it matters more than opinions:

  1. link to your data here
  2. release your data as CC-0 or a more free license.
  3. publish your data easily downloadable
  4. update that data at least weekly, provide reasons for the changes
  5. match all your data to OSM tags, think hard.
  6. do an importer for OSM, but don't do the import yourself.

Since OSM is a very good data producer and one of the biggest data consumers, I do not think your time spent on this will be wasted. At least if you think other people than you should use your data.

permanent link

answered 30 Mar '16, 05:12

emj's gravatar image

accept rate: 15%

Follow this question

By Email:

Once you sign in you will be able to subscribe for any updates here



Answers and Comments

Markdown Basics

  • *italic* or _italic_
  • **bold** or __bold__
  • link:[text](http://url.com/ "title")
  • image?![alt text](/path/img.jpg "title")
  • numbered list: 1. Foo 2. Bar
  • to add a line break simply add two spaces to where you would like the new line to be.
  • basic HTML tags are also supported

Question tags:


question asked: 29 Mar '16, 18:09

question was seen: 2,482 times

last updated: 30 Mar '16, 22:16

powered by OSQA