Talk:Armchair mapping

From OpenStreetMap Wiki
Jump to navigation Jump to search

Discuss Armchair mapping here

Reflecting consensus

I've tried to be balanced and reflect community consensus on this page, or at least present two sides of the coin. What do people think? It's a tricky one because many people will read the page and not really grasp why it is so negative towards armchair mapping, meanwhile I know some people rail strongly against armchair mapping in any circumstances, and so for them this page is not denouncing it negatively enough.

Currently the page is missing some examples of armchair mapping projects. Adding these will swing it a little more towards the positive

Overall it's probably time we try to agree upon what is accepted and what isn't. This page should help us do that.

-- Harry Wood 15:34, 31 January 2012 (UTC)

Hmm ok from this point of view I guess it's ok, as you explicitly try to show it in black/white. Let's see if it turns out to be useful :) --!i! This user is member of the wiki team of OSM 16:07, 31 January 2012 (UTC)

I didn't think it gave a proper representation of this modern type of mapping. Added some positive aspects of armchair mapping as there were none previously. Also I do disagree with the disadvantages posted and could easily argue against all points and it implied you cannot. Did you mean "difficult to argue for" rather than "difficult to argue against"?

e.g. having a map discourages mappers. For me I would never have got involved in this project if my home city was a blank canvas. I discovered the project maybe 5 years but nothing was on the map and that discouraged me to join. The work and learning curve to populate blank city from no knowledge of the project is huge. It was only years later when we had many roads did I contribute.

And frankly saying "often" armchair maps vandalise the map is just offensive. --Rovastar 01:05, 6 March 2012 (UTC)

Well first of all "Vandalise" is a strong word. Vandalism is intentionally ignoring the consensus norms of the OpenStreetMap community. Simple mistakes and editing errors are not vandalism. I didn't use the word vandalise. You're making that up. What I've written is "Armchair mappers have often been found to be damaging or undoing the hard work of surveying mapping contributors". Sorry if you find that offensive but it's true. It's a simple statement intended to spur people to think and to direct armchair mappers to follow the guidelines.
A lot of the points are intended as justification for why we need guidelines. The page doesn't say "You must not do armchair mapping". In that sense I've really tried to reflect a balance of points of view, because there's plenty of people involved in OpenStreetMap who think just that. They think the guideline should be "Do not do armchair mapping". I have a more balanced view myself, but I can see their point of view.
I think your new "advantages" section is good and helps this page to be more balanced. I also had in mind adding some examples of armchair mapping initiatives too.
I think the guidelines section needs to remain near the top of the page though. That's really the main point of this page (whether you agree or disagree with armchair mapping there are ways it should be done more carefully)
-- Harry Wood 10:59, 6 March 2012 (UTC)
To be honest I felt the tone before was an old school don't do armchair mapping. I know you were trying to appease the old school mappers who do everything via the expensive GPS devices and use no sat imagery and who are over sensitive about their data and tbh think they don't want a completed map as they like editing it too much. So just trying to balance the page out, there are a lot of armchair mappers out there and I don't want to discourage them. damaging/vandalising are all negative terms and saying often when someone does this is is damaging has a tone of "don't do it". I changed "often" to "occassionally" as I think this is more realistic. --Rovastar 13:44, 6 March 2012 (UTC)

"difficult to argue against" is not really what I meant. Of course people repeatedly argue against the criticisms of armchair mapping. Arguing is not difficult :-) I've changed it to "difficult to prove or disprove" and also given that whole paragraph a bit of rewrite. People have differing views of the relative value of armchair mapped data. Perhaps that is where disagreement really lies. I think for pro-armchair mappers, the arguments about psychology of the new users seem wooly and insignificant because they see their data as almost as good as somebody mapping the neighbourhood locally - Harry Wood 13:42, 14 March 2012 (UTC)

I removed the negative vibe from the intro. Many beginners start off via armchair mapping experimentation and we shouldn't discourage them. We should just tell them in a friendly manner that outdoors has advantages and that they are encouraged to do that once they're comfortable with the workings of OSM. A "Your technique is inferior" vibe doesn't help. I added some more disadvantages to the disadvantages section. -- SwiftFast (talk) 07:06, 30 April 2017 (UTC)

I don't think you are helping people by removing what you call "negative vibe" (actually what you removed is the 100% factual statement that "It is controversial and not always widely accepted as a good idea."). People are going to encounter this attitude sooner or later, and the Wiki should educate them about this rather than suggesting otherwise. Also, speaking of "vibe", by splitting the disadvantages into "practical" and "philosophical" reasons you have made it sound as if only the relatively meagre list of practical points you are listing makes armchair mapping inferior. "Philosophical" reasons, who takes them seriously, eh? In my eyes what you call "philosophical reasons" should be explained in depth and in detail and a honest attempt should be made to educate people about this, rather than trying to relegate it to a footnote and generally make it sound as if armchair mapping differs from real mapping (which you changed from "normal" to "outdoor mapping" - to me this is as if you'd start to call football "outdoor football") in just a few technical points but was otherwise equal. --Frederik Ramm (talk) 15:17, 1 May 2017 (UTC)
Thanks for your comment. Regarding the statement (It is controversial and not always widely accepted as a good idea), I've reconsidered this and I think you're right. I suggest only a slight rewording to make it really 100% factual; "It is sometimes controversial". Because there are perfectly valid use cases (like HOTOSM, or filling an empty map (better than nothing), or marking the outlines of buildings that you know for sure they exist, or tracing roads that exist for sure).
I believe that adding "practical reasons" significantly strengthens the case for outdoor mapping. It gives real and tangible reasons against armchair mapping. I felt the previous "Disadvantages and controversy" revision failed to articulate properly why armchair is sometimes bad and the bullet points didn't get the message across. The philosophical points, whilst important, stem from the community's reaction to the practical points, which are the main problem. As for the name, I absolutely didn't mean to be demeaning when I wrote "philosophical". Philosophical reasons are important. If you have a better title which gives a better vibe of seriousness, by all means change it. I agree that the philosophical points need to be better written, though.
The only point I seriously disagree with you on is the "normal mapping" > "outdoor mapping" change. I think that change was needed. Implying that armchair isn't normal (and therefore bad) is somewhat demeaning. Let's not forget armchair mapping has lots of valid use cases. There's nothing "abnormal" when properly using armchair mapping.
-- SwiftFast (talk) 06:16, 2 May 2017 (UTC)
I've put something new in the top section. For a long time we said "It is controversial and not always widely accepted as a good idea", which is alright as a quick few words, leaving the detail for lower down the page. But it does seem to confuse/annoy people, and invite knee-jerk reactions. I sat through a presentation by Kevin Bullock in SOTM 2014 in which he quotes this wiki page (4min 40seconds in) and presents his misinterpretation that sentence. I don't know if he's willfully misinterpreting or he just never really understood the point (For one thing that sentence is not saying "do not do armchair mapping", for another thing the page is definitely not saying that hi res imagery is anything other than hugely valuable to mappers following whatever workflow) Removing the "negative vibe" is not a balanced solution here. I've rewritten something different at the top. A longer sentence which is essentially is presenting a very brief summary of advantages and disadvantages.
Another change: As Frederik says, putting things down under a "philosophical" heading feels like a demotion. Really the main objections to armchair mapping, are these (unfortunately somewhat difficult to convey) points around the effect of quashing local community building. The more "practical reasons" bullet points are supportive of this, but actually somewhat distracting from it.
-- Harry Wood (talk) 04:29, 9 August 2017 (UTC)
The current version looks good and balanced to me. Thanks! -- SwiftFast (talk) 16:30, 13 August 2017 (UTC)

Issues in Texas

Thank you so much for adding this page, Harry. I think that it's more than mere coincidence that this page appears a short time after I had an issue with an out of area mapper here in Texas. I really appreciate all of your help, and it's a great "rules of the road" page for the community. --Homeslice60148 20:48, 30 January, 2012 (UTC)

Glad you like it. It's certainly my intention that you should be able to link to this page to point out to armchair mappers that they should be taking more care and not trampling over your more valuable local contributions. It's great that more people are getting involved in mapping the U.S. properly, with local contributions nowadays. But in the U.S. we've actually been promoting armchair mapping for years now in the form of TIGER fixup, and there's still massive areas which can benefit from the attention of people operating "out of area", because the data needs fixing up and the aerial imagery is hi-res enough to allow it. So I guess there's a balance of opinions about this, but one thing's for sure: Armchair mappers should not go messing up valuable surveyed data. Perhaps if we can invent tools to help armchair mappers recognise areas of local survey-based mapping, then this will end up being particularly useful in the U.S. -- Harry Wood 01:16, 31 January 2012 (UTC)

Some notes I made

When I did a bit of armchair mapping to add missing road names in some parts of the UK from bing/os opendata sources, I set myself a few ground rules, detailed here. Sorry for the strange order - by adding newer stuff at the top you have to read the sections there in a strange order. --EdLoach 17:26, 13 March 2012 (UTC)

Imagery can be wrong

Because of the reasons you explain. BUT. The "survey on the ground" can be wrong as well sometimes. I've seen outdated or inconsistent street signs or highway numbers in the "real" world as well. This is quite rare but seen enough to say "the ground survey is not always the truth". In such cases, the tags 'old_name' or 'old_ref' are helpful. --Pieren 13:32, 14 March 2012 (UTC)

There's all kinds of discussion about street names as you know ( Talk:Names ) These are an awkward case. But most data should be Verifiable on the ground, and so "survey on the ground" is the way to get to the truth.
People who survey on the ground ("Normal mapping" as opposed to "armchair mapping") do not always do a perfect job of it. Mistakes are made.
How to fix those mistakes is an interesting question. Really the best way to fix them would be to go on the ground again. I'd be very reluctant to say that data should/could be fixed by armchair mappers. Note: I'm not saying things can't be fixed by using aerial imagery after the fact, by the same people who were on the ground. But using aerial imagery to fix stuff without going there is problematic really. You have no way to know if you are in fact screwing up valuable data.
-- Harry Wood 14:08, 15 March 2012 (UTC)

"Further information"

Somebody added a link to Roof modelling, and that gave me the idea that there ought to be a single page that gives not only an overview of things one can deduce from imagery, and what to watch out for when interpreting imagery, but also with example images (of different quality imagery!) and their plausible wrong and correct interpretations. Various general guidelines are mentioned in at least Mapping_techniques, Yahoo!_Aerial_Imagery, Yahoo!_Aerial_Imagery/Accuracy, Bing#Precision, Accuracy and Roof modelling - also a short bit at Beginners_Guide_1.1.2 - and each of these have a different set of guidelines. I'm quite certain these are not the only pages already written. Aerial Imagery caveats, maybe?Alv 18:18, 14 March 2012 (UTC)

Yeah. There's a page called Using Imagery which popped up recently (more recently than this Armchair mapping page). It needs work, and we need to figure out the scope and avoid too much duplication. "Using imagery" is more broad than "Armchair mapping" of course, because many/most mappers use imagery along with their data they have surveyed when they contribute (Using imagery but not armchair mapping). So then we have the problem that things like details of how to get the offset right, perhaps belong on both pages. -- Harry Wood 21:23, 28 April 2012 (BST)

"Disadvantages and controversy" is too philosophical

I suggest we just get straight to the point: Armchair mapping can provide less finegrained (no shops, POIs, etc), and sometimes outdated data. Beginners are encouraged to go outdoors if they want to get serious. Simple and to the point. The current wording is too philosophical and subjective for a Wiki. (list some more real disadvantages if I missed any) -- SwiftFast (talk) 07:10, 30 April 2017 (UTC)

I split it into two sub sections: Practical and philosophical. Practical first, because it is more convincing for most people. -- SwiftFast (talk) 07:16, 30 April 2017 (UTC)
I added quite a few more serious disadvantages, like not being able to map shops! -- SwiftFast (talk) 07:22, 30 April 2017 (UTC)
Why wouldn't you be able to add shops? True, you can't virtually go along the road and map everything, but you sure can search online for shops in the region and add them. -- Xerus (talk) 11:12, 7 May 2020 (UTC)

It's not necessarily black or white

I do both armchair work and field work. The combination is stronger than each separately. I do armchair work based on field work by others, vice versa. I find armchair work requires goed knowledge en "feel", which you get from goimg out and look at it, then you can expand the work knowing what you are doing. Confirmation is important, from as many reliable sources as possible. Changing the work of others without actual survey is a big no, but contact with someone who works the field while you do the mapping works fine. Talking about mapillary, routes, details on shops, roads, landcover, signs.

Use case route mapping: lots of waymarked trails and route networks are offered by official sources and have regular operators and maintenance schedules / methods. My approach is to enter them all as they are offered, which gives complete coverage with the same quality as is offered. Then systematically acquire as much detailed survey info as possible, using reports, community, imagery, mapillary, gpx, and monitoring the operator's website. OSM then invariably quickly surpasses the original source. When an issue rises, either because a quality assurance tool detects an error or someone has doubts, then actual field survey is needed. This requires going out there, or from the armchair I can contact the community, the operator, mappers I know are in the area or mappers I get from the history.

In short, armchair mapping better be office chair mapping, and it blends fine with focussed field work. Then you get best of both worlds. --Peter Elderson (talk) 13:11, 29 October 2018 (UTC)

Doing Research

One thing that isn't mentioned here at all is that you can search through official registers, shop websites and the like to find and add new features. For example I am currently looking for a driving school in my hometown, so as I go through them I also add them to OSM. That is actually the main way I contribute to OSM, and I totally don't understand why this isn't mentioned anywhere. - Xerus (talk) 11:10, 7 May 2020 (UTC)