Talk:How We Map
About this document
I have a bit of an issue with "Legal - means that you don't copy copyrighted data without permission." that it may be misleading by specifically singling out copyright as the only legal concept that makes data unusable for OSM. I would prefer something along the lines of "Legal - means that you do not copy or use data without explicit or implicit permission to integrate it in to our dataset" (from a style pov I would like to leave the "explicit ot implicit" away, but that may be too unclear). SimonPoole (talk) 12:43, 10 February 2015 (UTC)
- This is a good point, how about just "Legal - means you do not copy data into openstreetmap unless it is allowed" with a link to more context if nesc?
- Or, closer to the original wording, "Legal - means that you don't copy others' data without permission.
Truthful and Verifiable
I guess in this context 'verifiable' is short-hand for 're-surveyable' which isn't really a word. One can clearly see that imaginary data is imaginary (such as cities in the mid-Atlantic full of Happy Parks) without having to re-survey it. So maybe the wording is a bit redundant but i think the points are valid. Maybe 'surveyable' is a better word here than 'verifiable'? Jo Walsh.
The county boundary near me is I'm sure "truthful" (it's been imported I think from OS OpenData) but except for the odd signpost isn't really verifiable at all. Having both words in there with a slightly different explanation for each gets the message exactly right, IMHO. --SomeoneElse (talk) 10:28, 12 February 2015 (UTC)
"Take as much care over edits as previous mappers did in the same area"
I've been wondering for a while if there's a way to say something a bit like that somewhere. This document might not be the place (it's a nice 1-pager now, and really doesn't need any kind of bloat). However, there probably needs to be something, somewhere - maybe I'll write a diary entry or a mailing list post or something.
One example is that there are places adding a node "in roughly the right place" makes total sense. To take the recent example of bicycle repair stations, knowing that there's one "somewhere near this building" in a town with no on-the-ground mappers and no POIs mapped locally is really useful. Having the same node not accurately located somewhere that there are on-the-ground mappers and lots of existing POIs creates a false expectation of the accuracy of the new node.
Another example is where some things have been surveyed, some not yet - and someone crudely adds buildings, or landuse, or whatever, without looking in detail at what's already there. In the case of new mappers this is to be expected, and is just a learning issue - but some experienced mappers do it too. --SomeoneElse (talk) 10:54, 12 February 2015 (UTC)
- I think this is a huge topic with respect to data maintainence and definitely deserves a more elaborate discussion elsewhere. There is also a different side to this, namely where there is stuff that is mapped with great attention to detail but is outdated meanwhile and the question if and to what extent such mapping should be replaced by newer but possibly less precise and detailed data. Right now OSM as a project has relatively little experience in that - being only 10 years old and most stuff we map changing rather slowly. --Imagico (talk) 10:03, 13 February 2015 (UTC)
after: "There are tagging standards but they evolve instead of being pushed through." maybe add: "Be extra careful when removing existing tags that are now considered deprecated, some people may rely on them."
From Matthijs via the list:
"There is no consensus on this point, and including it in official 'How we map' guidelines would falsely suggest that there exists such a consensus. It is also not true that we map in this way: many tags that we are using have been proposed before they were used, see http://wiki.openstreetmap.org/wiki/Category:Proposed_features_%22Approved%22.
Also, the meaning of this sentence is not clear: what does it mean for a tagging standard to 'evolve', and what does it mean for a tagging standard to be 'pushed through'?"
- Diagree with Matthijs. Agree with proposed document. There may be exceptions but in the whole, tagging in OSM is not, and was never, according to pre-mediated rules - everywhere you look, people are encouraged to "just do it" and see where that goes. That's the spirit of OSM really. The "lets' first set up a grand set of tags before we begin our job" approach is what some other projects did and they failed. --Frederik Ramm (talk) 22:05, 12 February 2015 (UTC)
- I also disagree with Matthijs and like the proposed document statement. The biggest single thing that has allowed OSM to grow to the scale and diversity it has is the lack of a proscribed set of tagging rules. The tags used have evolved and should continue to do so to maintain the spirit, potential and diversity of interest that is displayed throughout OSM. Documentation of tags used is useful, but some process to either restrict tags available or mechanically edit diversity into uniformity is potentially very destructive to the quality of tagging and mapping, to the range details recorded and made available and, most importantly, to the enthusiasm of the community. Chillly (talk) 22:15, 12 February 2015 (UTC)
- I agree with Matthijs that this entire bullet point is too controversial to be included in a document that seeks to represent a basic consensus. Of course it depends on interpretation somewhat (does "pushed through" refer to wiki voting or mechanical edits, for example), but overall it is simply not true that we have very few rules. And yes, many of these rules started out as proposals.
- A better phrase would imo be: "OpenStreetMap has tagging standards, but they are open ended and frequently change based on community consensus." Basically, we want to show that we are open for a diversity of interests, and that the existing tags are not set in stone. I believe that this does also express the values mentioned by Chillly, doesn't it? Mechanical edits are already covered elsewhere, so this rule doesn't need to repeat that. --Tordanik 02:10, 13 February 2015 (UTC)
"Relevant - means that you have to use tags that make clear to others how to re-use the data"
Not sure I agree with the definition of "Relevant" here. That would make everything *relevant* that is tagged in a good way. I think those are two different things that we both want:
1. Use useful tagging scheme (that opens a whole can of worms what kind of tags are good) - this is a technical question *how* tags should look 2. Map only relevant stuff (another can of worms, but one we have to open one day, important for community cohesion in my opinion) - this is about *what* we tag
- Relevancy is a difficult topic and probably no consensus exists at the moment. IMO everything that is truthful and verifiable on a long term level (i.e. permanency as a component of verifiability) as well as legal (i.e. no privacy issues) is potentially relevant. It seems to me most relevance questions in substance (in contrast to tagging) could be solved by a stricter adherence to the verifiability principle. Beyond that restrictions would be hairy since this would significantly limit openness and diversity of the community in favour of cohesion which would be very problematic.
- But i agree ultimately there have to be limits (like for mapping grains of sand in the Sahara) - the question is only if such limits need to be imposed or if this is self regulating.
- In a way focussing relevancy on tagging makes sense since it is generally a good idea to use maximizing relevancy as a criterion for tag selection. But this is IMO too complex and abstract a topic for a general how-to-map guide.
- In summary I would probably phrase this point differently like:
- Meaningful - means that it should be clear from the tags you use how others can interpret and re-use the data.
- There is a temporal component to relevancy. Traffic light timing for example changes from time to time... is it a good idead to record that data? Brycenesbitt (talk) 05:58, 17 February 2015 (UTC)
Don't trust the Wiki rule
I think we need some kind of "don't trust the wiki" rule. There are just too many literal-minded people who take instructions on how to tag etc. from the wiki to literal.
- Perhaps this ties in a little with what SomeoneElse said above - take a close look at the area you're editing in, and if everyone seems to do things different than on the Wiki, what is in use always takes precedence. (This also neatly implies that you should never edit an area larger than you can take a close look at...) --Frederik Ramm (talk) 22:08, 12 February 2015 (UTC)
Self Promotion / Vandalism
- Correctly mapped local businesses, even if the business owner or SEO consultant places them.
We do not accept
- Internet businesses, as they do not have a relevant location.
In general the features mapped should be useful to others, and not purely self promotion. For this reasons mapping things like the home office of a consultant are strongly discouraged.
Let's not take the nicely worded, digestible page that was presented and turn it into a wiki-sized pile of drivel by discussing it it death. Just go for it. Chillly (talk) 22:18, 12 February 2015 (UTC)
But we have to be *seen* to generate a wiki-sized pile of drivel before just going for it :)
"Do not use OSM for agitation"
So maybe something like this would be nice to include. That is - don't try to push through your particular views/ideology/interests to the map/wiki. Also, mention that OSM is not for advertising. RicoElectrico (talk) 03:19, 13 February 2015 (UTC)
- I agree. This does not feel like an introductory page as it is, and we do have the Beginners Guide. --Jgpacker (talk) 21:33, 13 May 2015 (UTC)
Do not delete data unless you know (or have very strong reason to believe) that it is incorrect.
What if adding something like this?
"Do not delete data unless you know (or have very strong reason to believe) that it is incorrect. Even if you find something incorrect in some level, see if it's worth more trying to correct it than deleting it." - SergioAJV 04 May 2016