Re: [OSM-talk] tagging trailblazes / marked paths
This is the sort of map I envision: http://www.uwsp.edu/cnr/Schmeeckle/Map/images/schmeeckle_map.jpg As an aside, I like the style of that map for doing walking routes (e.g. on Freemap) Wonder how easy it would be to generate using GD / PDF libraries etc? Nick ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
Nick Whitelegg wrote: This is the sort of map I envision: http://www.uwsp.edu/cnr/Schmeeckle/Map/images/schmeeckle_map.jpg As an aside, I like the style of that map for doing walking routes (e.g. on Freemap) Wonder how easy it would be to generate using GD / PDF libraries etc? That I don't know, but if you're curious, here's the same area in my slightly customized mapnik render (modified to understand foot/bicycle/horse=designated, and to render paths on top of roads, and with a catchall rule for any paths which have no designation.) http://web.hawkesnest.net/osm.html?lat=44.53762lon=-89.56218zoom=16layers=B and in osmarender: http://www.informationfreeway.org/?lat=44.53753819714547lon=-89.56241392105782zoom=16layers=B000F000F I tried using generate_image.py to create an image of the same area, but it just showed up blank grey... -Alex Mauer hawke ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
On Wed, Aug 6, 2008 at 6:43 PM, Alex Mauer [EMAIL PROTECTED] wrote: In fact, highway=path makes it easier on renderers. That's a bold statement for someone who has no experience running a renderer. Without using highway=path, renderers need to understand every single specialized way. [...] Most renderers can render all paths generically. If a new access method is added, no change is needed. What an absolutely terrible idea. This is astounding daft. If I have chosen to render paths for cyclists, horse riders and pedestrians on my map, why on earth would I want to accidentally render every other variant when someone adds it? If I wanted to render every possible future linear feature without knowing what it was I would use an elsefilter on planet_osm_line and be done with it. Specialized maps like cyclemap only need to add special rendering for their area of interest. For example, a cycle map can render any highway=path the same, and only highlight those which are for bicycles. Specialised maps, nay, every map would need to keep track of every single possible tag that you can add to highway=path just in case someone adds something new that you don't want to render, or you thought was dangerously misleading. There's good reasons why every new feature gets a new tag - it's so that you don't end up accidentally rendering things in a confusing manner. There's very little to be gained from lumping lots of things that you'd never want to render identically - no sane map would render cycle paths, footpaths and snowmobile-only trails identically. So what you're suggesting actually *raises* the bar for renderers since it's now twice as hard to render just footpaths. 1. highway=[anything]way. Renderers need to know about every type of [thing]way. Impossible to tag a multiple-use way (or ridiculously complex anyway -- highway=bicyclefoothorseskisnowmobileway? I'm not going to waste time discussing with someone who can't refrain from adding strawman arguments to everything he discusses. So in summary - regardless of the discussions of whether the tagging scheme is better for the contributor, or from a data correctness point of view, please don't start bandying around wild statements about it being easier on renderers. Cheers, Andy ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
Alex Mauer wrote: In fact, highway=path makes it easier on renderers. Without using highway=path, renderers need to understand every single specialized way. snowmobileway, skiway, nordicskiway, telemarkskiway, alpineskiway, elephantway, etc. When someone introduces a new specialized way, the renderers need to be updated to understand it. I work on a (non-OSM) map renderer, and understanding every single specialised way (that you want to render) is the only sensible way to do it. Renderers have to decide up front what to show and discard everything else, as what they select completely depends on what purpose the map is meant to have. E.g., a road map renderer won't include overhead power lines - but a countryside walking map might include them for orientation. Rendering every single arbitrary way gives you output like Ito's OSM Mapper tool - fantastically useful for visualising data, but very much a specialised map. -dair ___ [EMAIL PROTECTED] http://www.refnum.com/ ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
Andy Allan wrote: What an absolutely terrible idea. This is astounding daft. If I have Yes, I am clearly mad. I appreciate that. chosen to render paths for cyclists, horse riders and pedestrians on my map, why on earth would I want to accidentally render every other variant when someone adds it? If I wanted to render every possible future linear feature without knowing what it was I would use an elsefilter on planet_osm_line and be done with it. Huh? There's a difference between any future linear feature and any sort of path. Say you've got a place with a variety of paths: bike trails, walking trails, ski trails. Now say that you want to make a map useful for biking that area, but you still want to show the other paths. (so that turning at the second left is still accurate) So you render the bike paths in a green broken line. Now, does it make more sense to have single rule for all the other kinds of path that you don't care about to render as a grey broken line, or does it make more sense to have separate extra rules to render footway, bridleway, and four kinds of skiway all in that way? And then someone maps the snowmobile trail that also goes through the area. Is it better that it's now rendered like all the other special-use paths that you don't wish to highlight, or is it better to have to add another rule for snowmobileways? There's good reasons why every new feature gets a new tag - it's so that you don't end up accidentally rendering things in a confusing manner. There's very little to be gained from lumping lots of things that you'd never want to render identically - no sane map would render cycle paths, footpaths and snowmobile-only trails identically. So what Incorrect. See above. If one is making a ski or a horse map, why should one care whether some other paths are for foot, bicycle, or snowmobile? But one would still want to render them just to show that they're there. you're suggesting actually *raises* the bar for renderers since it's now twice as hard to render just footpaths. Not really. If it's highway=footway or foot=designated, render it as a footpath. Hey, that's how it already should work. Convenient! 1. highway=[anything]way. Renderers need to know about every type of [thing]way. Impossible to tag a multiple-use way (or ridiculously complex anyway -- highway=bicyclefoothorseskisnowmobileway? I'm not going to waste time discussing with someone who can't refrain from adding strawman arguments to everything he discusses. That's no strawman. See http://wiki.openstreetmap.org/index.php/Proposed_features/Cycle_and_Footway So much for: the obligation to research alternative options (rather than just campaigning for one) surely lies with the proposers. Just by mentioning one of those alternative options, you immediately ignore anything else I have to say. Did you even read the rest of the message? The other two options I considered were much better, and I stated straight away that that one is terrible. -Alex Mauer hawke ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
On Thu, Aug 7, 2008 at 5:10 PM, Alex Mauer [EMAIL PROTECTED] wrote: Andy Allan wrote: What an absolutely terrible idea. This is astounding daft. If I have Yes, I am clearly mad. I appreciate that. chosen to render paths for cyclists, horse riders and pedestrians on my map, why on earth would I want to accidentally render every other variant when someone adds it? If I wanted to render every possible future linear feature without knowing what it was I would use an elsefilter on planet_osm_line and be done with it. Huh? There's a difference between any future linear feature and any sort of path. Say you've got a place with a variety of paths: bike trails, walking trails, ski trails. Now say that you want to make a map useful for biking that area, but you still want to show the other paths. (so that turning at the second left is still accurate) So you render the bike paths in a green broken line. Now, does it make more sense to have single rule for all the other kinds of path that you don't care about to render as a grey broken line, or does it make more sense to have separate extra rules to render footway, bridleway, and four kinds of skiway all in that way? And then someone maps the snowmobile trail that also goes through the area. Is it better that it's now rendered like all the other special-use paths that you don't wish to highlight, or is it better to have to add another rule for snowmobileways? If the point is to show all possible paths, then you'll also want to similarly show all the roads as well? In which case an else rule on highway=* would solve the problem. So the only distinction created by highway=path is that it is of type path which is a sufficiently broad spectrum of features from tiny trails to wide tracks that it isn't actually much of a distinction at all. Incidentally (and completely irrelevant to the discussion), I just found a few ways in Dorset, England, annual snow fall maybe one or two days a year, which had recently been converted from highway=track to highway=path,foot=yes,motorcar=no,ski=no,snowmobile=no. I'm fairly sure that other than the lack of snow, skiing isn't actually banned. Does anyone know why they might have done this? A preset somewhere maybe? (anonymous user so I can't ask them). Dave ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
Dave Stubbs wrote: If the point is to show all possible paths, then you'll also want to similarly show all the roads as well? In which case an else rule on highway=* would solve the problem. The point is to show all possible paths and highlight one particular subset of them, yeah. This is the sort of map I envision: http://www.uwsp.edu/cnr/Schmeeckle/Map/images/schmeeckle_map.jpg Note that it is useful to differentiate roads from paths on that sort of map, so a catchall on highway=* wouldn't be sufficient. And before someone says it, I'm not trying to duplicate that map in OSM. So the only distinction created by highway=path is that it is of type path which is a sufficiently broad spectrum of features from tiny It's not there to distinguish one kind of path from another, it's there to distinguish a path from something which isn't a path, such as a road. Does anyone know why they might have done this? A preset somewhere maybe? (anonymous user so I can't ask them). Looks like the JOSM paths preset to me. If someone used that to change it to a path and thought they had to fill in all the access restrictions, that would likely be the result. no is probably correct, since it means not permitted or unsuitable -- if it gets so little snow, it's probably unsuitable for skiing. -Alex Mauer hawke ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
On Wednesday 06 August 2008 00:45:20 Dave Stubbs wrote: On Tue, Aug 5, 2008 at 10:32 PM, Alex Mauer [EMAIL PROTECTED] wrote: Jon Burgess wrote: The only thing I see an issue with is introducing the specific 'highway=path' tag. I see this as an unnecessary complication. I guess it's a matter of perspective. I see it as a simplification: instead of having three categories for one physical feature (and still needing to twist reality in order to fit them in (highway=footway+foot=no+ski=yes, anyone?) you have only one category. From a quick glance at the examples given I think they are all covered with combinations of highway=cycleway|footway|track with the other tags Except the first, second, third, fourth, fifth, sixth, ninth, and tenth. Yeah. you propose like foot=y/n, motorcar=y/n or tracktype=gradeN etc. I propose none of those tags. the first two are part of the initial revision of access=*, and the last I do not propose nor agree with. I really don't see what highway=path adds. To quote the wiki page: A generic path. Either not intended for any particular use, or intended for several different uses. For the nth time, bridleway/cycleway/footway do not cover these. Gotcha. Excepth that, assuming you /can/ walk on it, that's what the rest of us have been using highway=footway for since the dawn of time (well, dawn of map features maybe. well, last couple of years at least). I beg to differ here. What do you mean /can/ walk on it? You can basically walk everywhere. Heck, people are walking up to the top of Mount Everest, but I'd be hard pressed to designate highway=footway along the vertical walls of ice along that route. So that's not what the rest of us have been using it for. -Inge ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
On Tue, Aug 05, 2008 at 06:33:10PM -0500, Alex Mauer wrote: So it really depends on interpretation. In particular, footways have a particular legal status in the UK which doesn't apply to every place that you can walk. Just as a point of information, this isn't actually true. As far as I am aware, the only UK legal use of the term footWAY is to refer to what I would call a pavement and you might call a sidewalk. The particular legal status to which you refer is actually applied to the legal term footpath, and the OSM tag highway=footway in the UK does not, of itself, imply that a path is a Public Right of Way (and hence a legal footpath). s ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
OK, perhaps nonsensical was too strong. Against the intent of the highway tag certainly, and I'd add defeating the purpose of the access series of tags as well. I hope you agree with my point that the legal accessibility of a way doesn't belong in the highway key, especially when we have a separate key for it. The highway tags are a mishmash of different concepts and properties of different kinds. There's physical, legal, and intended use jammed in there in different parts. If you're looking for proper separation of concerns then don't use the highway tag at all. But yes, in general I think things like legal right of way etc are best kept elsewhere, especially when we have a separate tag for it as you say. Hence why I assume highway=footway does not imply any kind of right of way, merely that this is a path, and people walk on it. So that leaves us with intended use. As far as I'm concerned if there's a path there, and people walk on it, then it's for people to walk on, so unless it has another obvious intended use I tag it as footway. Dave ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
On Wed, Aug 6, 2008 at 5:07 AM, Alex Mauer [EMAIL PROTECTED] wrote: Incorrect. You neglected to account for the existing tags on those 509k/425k. There's actually a net gain (reduction) in the number of tags needed. The simplest cases (cycleway/footway/bridleway) are identical, obviously. But now a specialty route which is not a c/f/b is both more intuitive (no highway=footway+foot=no needed) and requires fewer tags (highway=path + snowmobile=designated instead of highway=footway + foot=no + snowmobile=designated for example). highway=footway + foot=no is simply garbage, and shows that you don't really understand how the tagging is supposed to work. The footway/cycleway/bridleway is just three very common examples. If you particular thing doesn't fit into any of them (e.g. these snowmobile things, or ice-climbing pitches) then there's no need to crowbar them in with such conceptual acrobatics. Highway=snowmobileway would be a single-tag solution for snowmobile tracks that you aren't allowed to do anything else on. Upheaval? People in charge of renderers being asked why highway=path, cycleway=designated doesn't show up when highway=cycleway does, when they could spend time on more useful things which add value to the maps. Dual-tagging regime? See preceding sentence. I also refer you to the instances of the work or in the example page you keep linking to. Neither have happened here Untrue. And IMO if someone knew of a less disruptive, more intuitive change to make, they should have mentioned it during the 6 months that the proposal was in the wiki. I'm not obliged to spend my time patiently explaining the counterarguments to every proposal on the wiki - the obligation to research alternative options (rather than just campaigning for one) surely lies with the proposers. Cheers, Andy ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
Stephen Gower wrote: On Tue, Aug 05, 2008 at 06:33:10PM -0500, Alex Mauer wrote: So it really depends on interpretation. In particular, footways have a particular legal status in the UK which doesn't apply to every place that you can walk. call a pavement and you might call a sidewalk. The particular legal status to which you refer is actually applied to the legal term footpath, You're right, I should have written footpath. My point still stands though. -Alex Mauer hawke ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
[OSM-talk] tagging trailblazes / marked paths
Hello, Now that the highway=path has been moved to the official features page, is there any more or less agreed way of tagging marked paths? I see a lot of different proposal pages on this, but no real consensus. I myself have been tagging my local area using trailblazed=yes, but it would be nice to use some generally agreed tag. Igor -- http://igorbrejc.net ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
Brejc wrote: Hello, Now that the highway=path has been moved to the official features page, is there any more or less agreed way of tagging marked paths? I see a lot of different proposal pages on this, but no real consensus. I myself have been tagging my local area using trailblazed=yes, but it would be nice to use some generally agreed tag. Take a look at http://wiki.openstreetmap.org/index.php/Key:trail_visibility Combined with highway=path, does that cover what you need to map? -Alex Mauer hawke ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
Igor Brejc wrote: Now that the highway=path has been moved to the official features page, is there any more or less agreed way of tagging marked paths? I see a lot of different proposal pages on this, but no real consensus. I myself have been tagging my local area using trailblazed=yes, but it would be nice to use some generally agreed tag. It was approved on the basis of a tiny vote on the wiki and I would say there is zero chance of most people switching from the tags that have been in use for several years to some new scheme that, as I understand it, requires about five tags for each path. Given that most of the UK examples on the wiki were actually wrong by their own definition last time I looked I certainly plan to stick to what we've always done. Tom -- Tom Hughes ([EMAIL PROTECTED]) http://www.compton.nu/ ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
Hi, Tom Hughes wrote: It was approved on the basis of a tiny vote on the wiki and I would say there is zero chance of most people switching from the tags that have been in use for several years to some new scheme that, as I understand it, requires about five tags for each path. Given that most of the UK examples on the wiki were actually wrong by their own definition last time I looked I certainly plan to stick to what we've always done. A very similar thing seems to have happened recently with the Crossing tag. I've never been a friend of that voting business but it seems to get more absurd every day. Is it perhaps time now to have a vote on abolishing votes altogehter - or should we continue to let people vote on whatever they like and ignore the results? It's fine with me but seems to irritate newbies who lack the stubbornness that speaks from your above paragraph ;-) Bye Frederik -- Frederik Ramm ## eMail [EMAIL PROTECTED] ## N49°00'09 E008°23'33 ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
Tom Hughes wrote: It was approved on the basis of a tiny vote on the wiki and I would Uh, what? 34 votes is one of the largest votes of any proposed/approved feature on the wiki. say there is zero chance of most people switching from the tags that have been in use for several years to some new scheme that, as I understand it, requires about five tags for each path. Then I think you misunderstand it. Take a look at http://wiki.openstreetmap.org/index.php/Tag:highway%3Dpath/Examples -- most require two tags at most. The only one which reaches five additional tags is the last one. Which doesn't fit into the bridleway/cycleway/footway paradigm anyway, and is one of the most complex examples to be found. You don't like highway=path, fine. If your tagging needs are met by bridleway/cycleway/footway, then I'm glad for you. But it's not adequate for all situations. Don't make up bullshit just to trash-talk that which you don't understand. -Alex Mauer hawke ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
Frederik Ramm wrote: Hi, Tom Hughes wrote: It was approved on the basis of a tiny vote on the wiki and I would say there is zero chance of most people switching from the tags that have been in use for several years to some new scheme that, as I understand it, requires about five tags for each path. Given that most of the UK examples on the wiki were actually wrong by their own definition last time I looked I certainly plan to stick to what we've always done. A very similar thing seems to have happened recently with the Crossing tag. I've never been a friend of that voting business but it seems to get more absurd every day. Is it perhaps time now to have a vote on abolishing votes altogehter - or should we continue to let people vote on whatever they like and ignore the results? It's fine with me but seems to irritate newbies who lack the stubbornness that speaks from your above paragraph ;-) Bye Frederik I'm not touching the voting-no voting issue :). I'm just trying to think ahead and consider how this data that we're so happily entering will be used for purposes other than just rendering (getting back to my hiking-buddy SW idea I mentioned a few times before). If we have 10 different ways of describing the same thing it's going to be difficult to implement and maintain SW that uses this data. Igor -- http://igorbrejc.net ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
Alex Mauer wrote: Brejc wrote: Hello, Now that the highway=path has been moved to the official features page, is there any more or less agreed way of tagging marked paths? I see a lot of different proposal pages on this, but no real consensus. I myself have been tagging my local area using trailblazed=yes, but it would be nice to use some generally agreed tag. Take a look at http://wiki.openstreetmap.org/index.php/Key:trail_visibility Combined with highway=path, does that cover what you need to map? -Alex Mauer hawke ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk Which tag value would I use for a path through the forest that is clearly visible, but with no markings? There are a lot of those in Slovenia. Igor -- http://igorbrejc.net ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
Frederik Ramm wrote: I've never been a friend of that voting business but it seems to get more absurd every day. Is it perhaps time now to have a vote on abolishing votes altogehter - or should we continue to let people vote on whatever they like and ignore the results? Do you have a better suggestion? I like Andy Allan's modifications to the Key:crossing page, suggesting that it be used for documenting current usage, with renderers working from that. So all you have to do to add a key or value is to use it. It's unfortunate that current usage is so hard to find, particularly outside of Europe... -Alex Mauer hawke ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
On Tue, 2008-08-05 at 15:20 -0500, Alex Mauer wrote: Tom Hughes wrote: It was approved on the basis of a tiny vote on the wiki and I would Uh, what? 34 votes is one of the largest votes of any proposed/approved feature on the wiki. say there is zero chance of most people switching from the tags that have been in use for several years to some new scheme that, as I understand it, requires about five tags for each path. Then I think you misunderstand it. Take a look at http://wiki.openstreetmap.org/index.php/Tag:highway%3Dpath/Examples -- most require two tags at most. The only one which reaches five additional tags is the last one. Which doesn't fit into the bridleway/cycleway/footway paradigm anyway, and is one of the most complex examples to be found. You don't like highway=path, fine. If your tagging needs are met by bridleway/cycleway/footway, then I'm glad for you. But it's not adequate for all situations. The only thing I see an issue with is introducing the specific 'highway=path' tag. I see this as an unnecessary complication. From a quick glance at the examples given I think they are all covered with combinations of highway=cycleway|footway|track with the other tags you propose like foot=y/n, motorcar=y/n or tracktype=gradeN etc. I really don't see what highway=path adds. The one exception is for snowmobile, for that I'd suggest possibly adding highway=snowmobile instead. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
Igor Brejc wrote: Which tag value would I use for a path through the forest that is clearly visible, but with no markings? There are a lot of those in Slovenia. It's probably not necessary to tag it specially at all as I expect this is the default, but it looks like trail_visibility=excellent (Unambiguous path or markers everywhere) would be the one to use. -Alex Mauer hawke ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging trailblazes / marked paths
On Tue, Aug 5, 2008 at 9:40 PM, Alex Mauer [EMAIL PROTECTED] wrote: Frederik Ramm wrote: I've never been a friend of that voting business but it seems to get more absurd every day. Is it perhaps time now to have a vote on abolishing votes altogehter - or should we continue to let people vote on whatever they like and ignore the results? Hmm. It's a tough one, especially when 34 people (out of an electorate of, lets say, the 5,750 people editing last month) only narrowly agree on restructuring some of the most widespread tags in the db. The wiki, for better or worse, is most likely seen as authoritative by most of those 5,750 people and so I'd like it if everyone was a bit more cautious about changing what's said on there, especially when it comes to changing existing conventions/features, posting stuff that's contrary to established use, or confusing or complex topics. I like Andy Allan's modifications to the Key:crossing page, suggesting that it be used for documenting current usage, with renderers working from that. So all you have to do to add a key or value is to use it. It's the way I like it, although the downside is that I'll often start tagging and rendering new stuff and forget to document it (or even add it to the key) (Ahem. Notice is hereby given that tagging cafes with fryup=yes is likely to get you a nice fork-in-a-sausage symbol on the cycle map :-) http://www.gravitystorm.co.uk/osm/?zoom=16lat=6711490.40418lon=-11750.44101layers=B00 ) It's unfortunate that current usage is so hard to find, particularly outside of Europe... If anyone needs some real figures for these discussions, I'm more than happy to help. I only realised recently that tagwatch only covers Europe, which is unfortunate. So below is the list of ways using the highway tag with more than 100 instances. You can see that the highway=footway/cycleway/bridleway/pedestrian totals 509,920 instances (path has a respectible, but tiny by comparision, 2165). And everyone should remember the 509k instances mean that *lots* of different contributors use these tags and understand them; and there are *lots* of renders, routing algorthims and whatnot understand and use them too. So what are the advantages of the change? One scheme that covers the corner cases along with the most common occurences. And the disadvantages? Confusion for many contributors, every data user needing to understand two sets of tagging styles, the most common cases (the 509k) needing twice as many tags as before, and the corner cases are still fairly corner needing a small handful of tags. So in my opinion, the problem is the main tagging scheme wasn't well enough documented (a canal towpath is hardly a pedestrian precinct, which I came across today) to prevent arguments and misunderstandings, but the proposed upheaval and/or dual tagging regimes is overkill. A way to tag the corner cases that don't fit in well would have been much preferable. And you can throw in the term cost/benefit here as well, but I'm sure everyone gets my point by now. Cheers, Andy residential | 12605937 service | 2200128 unclassified | 106 | 529868 secondary| 510636 track| 417479 tertiary | 391691 footway | 320536 primary | 268507 motorway_link| 181060 cycleway |97914 pedestrian |84164 motorway |81050 trunk|70124 trunk_link |27778 primary_link |20594 steps|19919 living_street|13936 road | 8500 bridleway| 7306 unsurfaced | 6313 minor| 4010 path | 2165 construction | 1675 FIXME| 889 secondary_link | 738 byway| 630 footpath |
Re: [OSM-talk] tagging trailblazes / marked paths
Dave Stubbs wrote: Gotcha. Excepth that, assuming you /can/ walk on it, that's what the rest of us have been using highway=footway for since the dawn of time (well, dawn of map features maybe. well, last couple of years at least). If it happened to have another purpose (ie: bikes or horses) then it got upgraded to cycleway or bridleway. If that's not what you thought highway=footway meant then I guess the docs for highway=footway need updating (again). From Tag:highway=footway: For designated footpaths, i.e. mainly/exclusively for pedestrians. That is a perfectly reasonable definition in my opinion. However, I see a distinction among intended for, allowed, and not forbidden. So it really depends on interpretation. In particular, footways have a particular legal status in the UK which doesn't apply to every place that you can walk. And even in the US there's a difference between a path and a path built specifically for people to walk on. So that means 2 to 9 are fully covered by the existing map features (ie: footway/cycleway/bridleway/track/service) 2-4: clearly not mainly/exclusively for pedestrians, since it's not for anything in particular. It's just a path; certainly no cycleway or bridleway, even though bicycles and horses don't appear to be forbidden. 5: it's a cycleway and a footway. Calling it only one of those gives a priority which doesn't exist. That problem is fixed with the designated value for access. 6: the purpose of the path is impossible to determine. All we can see is what is forbidden. We can perhaps hope that horses and bicycles are allowed to use it, I guess. It's still not a bridleway. 7,8: covered by f/c/b 9: again, we can't tell what it's for, just what's forbidden. Definitely not a bridleway since horses are forbidden. Not a cycleway either though, since it's not /for/ bicycles. Aside: I don't think track/service matter at all for this purpose. Nonsensical is a matter of opinion clearly. You can't just say things are nonsensical and hope that means something. It happens to make perfect sense. You might not like it, and there might be a better way, but that's not really the same thing. OK, perhaps nonsensical was too strong. Against the intent of the highway tag certainly, and I'd add defeating the purpose of the access series of tags as well. I hope you agree with my point that the legal accessibility of a way doesn't belong in the highway key, especially when we have a separate key for it. -Alex Mauer hawke ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
