[OSM-talk] tagging roads
Hi, I have noticed intense discussion about changing how roads should be tagged. It seems that some people devised their own way how to apply different values for highway tag and now attempt to force this on everybody. On the other hand some people are attempting to introduce new values for highway tag (http://wiki.openstreetmap.org/wiki/Proposed_features/residential_narrow and http://wiki.openstreetmap.org/wiki/Proposed_features/residential_narrow/old). Both approaches seem wrong. Changing how roads should be classified will break existing OSM data. Introducing new values will only add to confusion choosing road classification. It seems that both issues are caused by desire to achieve certain rendering in mapnik or TAH. I guess best way to solve such problems would be to render streets differently based on value of tag width. This would encourage use of tag width. This way we would also gain useful data that can be used by routing software. Another tag that is currently not supported by rendering engines is surface. I think renderers should make distinction between paved and unpaved roads. Otherwise you are risking that people will classify unpaved road as track just to achieve desired rendering effect. I am aware that highway=unsurfaced is rendered differently. But this value is supposed to be deprecated. Besides, unsurfaced roads can have different classifications (http://wiki.openstreetmap.org/wiki/Highway#Exceptions_to_physical_attributes). I also propose extending instructions for road classification to use width tag when road is narrow. Instructions should also suggest what is considered narrow. For instance less than 4 meters for residential and unclassified roads, less than 6 meters for tertiary roads, less than 8 meters for secondary and primary roads. For trunks and motorways this limit should probably apply to lane width. Proposed limits on when certain type of road is considered narrow should match rendering rules. Blaž ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
--- On Sun, 2/8/09, Blaž Lorger wrote: > I also propose extending instructions for road > classification to use width tag I agree with everything else you wrote except width since I really don't want to get a tape measure out and measure widths of roads, using lanes=* to estimate widths would be more sensible and is already in use. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Sun, Aug 2, 2009 at 10:59 AM, John Smith wrote: > I agree with everything else you wrote except width since I really don't want > to get a tape measure out and measure widths of roads, using lanes=* to > estimate widths would be more sensible and is already in use. +1 I prefere to add "narrow=yes" combined with residential or unclassified (or any other) highways. It should be interpreted as "between 75% to 50% narrower than the default width of this highway type". Pieren ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Sunday 02 August 2009 10:59:08 John Smith wrote: > --- On Sun, 2/8/09, Blaž Lorger wrote: > > I also propose extending instructions for road > > classification to use width tag > > I agree with everything else you wrote except width since I really don't > want to get a tape measure out and measure widths of roads, using lanes=* > to estimate widths would be more sensible and is already in use. Unfortunately lanes only specifies number of lanes. In general every road that is not one way has at least 2 lanes, even if it is narrow, say 3.5 meters. But you are right. Actually measuring road width is cumbersome and dangerous. We really don't want any OSM mapper to be nominated for Darwin award. :-) Still, you must get width data somehow. But, since width is more important in case of narrow road, you can limit gathering of width data only on narrow roads. Which has additional benefit that width for narrow road is easier to estimate than for wide road. Obviously it would be a good idea to add recommendation that only width of narrow roads should be estimated to tagging instructions. Some good practices on how to accurately estimate road width could also came handy. Which leaves us with problem how accurate is width data. Wiki suggest using tag est_width. But this means that software needs to check two tags: width and est_width. Maybe additional tag (width_estimated=yes|no) could solve this problem. Default value for tag would be yes, since we can assume that someone that did not bother to acquire accurate data won't bother to add the tag to record such fact. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
--- On Sun, 2/8/09, Blaž Lorger wrote: > Unfortunately lanes only specifies number of lanes. In > general every road that > is not one way has at least 2 lanes, even if it is narrow, > say 3.5 meters. Even one way roads can have multiple lanes. Like dual carriage ways :) > Some good practices > on how to accurately estimate road width could also came > handy. I feel you could get road widths accurately enough based simply on the number of lanes and the type of highway. For example, a residential road with 2 lanes, one in each direction is usually about 6-7m wide. Motorways are usually the numbers of lanes times 3m + 3m each side of the road way to allow for break downs, or maybe you could even have lanes=* breakdown_lanes=* to make this more accurate, but in any case a 3 lane motorway will be about 15m wide (3 * 3 + 3 + 3). ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
--- On Sun, 2/8/09, Pieren wrote: > I prefere to add "narrow=yes" combined with residential or > unclassified (or any other) highways. It should be > interpreted as > "between 75% to 50% narrower than the default width of this > highway > type". +1 ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
Am Sonntag 02 August 2009 11:49:11 schrieb Blaž Lorger: > On Sunday 02 August 2009 10:59:08 John Smith wrote: > > --- On Sun, 2/8/09, Blaž Lorger wrote: > > > I also propose extending instructions for road > > > classification to use width tag > > > > I agree with everything else you wrote except width since I really don't > > want to get a tape measure out and measure widths of roads, using lanes=* > > to estimate widths would be more sensible and is already in use. > > Unfortunately lanes only specifies number of lanes. In general every road > that is not one way has at least 2 lanes, even if it is narrow, say 3.5 > meters. Yes, but equally unfortunately width only specifies width, and not the number of lanes. Therefore, it would be nice to have both. Regards, Marc signature.asc Description: This is a digitally signed message part. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Sunday 02 August 2009 11:49:06 Pieren wrote: > On Sun, Aug 2, 2009 at 10:59 AM, John Smith wrote: > > I agree with everything else you wrote except width since I really don't > > want to get a tape measure out and measure widths of roads, using lanes=* > > to estimate widths would be more sensible and is already in use. > > +1 > > I prefere to add "narrow=yes" combined with residential or > unclassified (or any other) highways. It should be interpreted as > "between 75% to 50% narrower than the default width of this highway > type". > > Pieren Wiki clearly states why tag narrow=yes would be a bad idea (http://wiki.openstreetmap.org/wiki/Key:width#Using_relative_sizes). Basically, what is wide for bicycle is narrow for a car, what is wide for a car is narrow for a truck ... Besides, I am not much bothered by narrow roads when driving a car. On the other hand some drivers will drive very slow when lane width drops under twice the width of their vehicle. Whose estimation of whether road is narrow or not is appropriate? Specifying actual or at least estimated width is only way to ensure that data is universally usable. I guess heavy truck driver would not be amused when navigation system would direct him in 3m wide residential street, while car driver would only see it as a problem when passing another car. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
--- On Sun, 2/8/09, Blaž Lorger wrote: > Specifying actual or at least estimated width is only way > to ensure that data > is universally usable. I guess heavy truck driver would not > be amused when > navigation system would direct him in 3m wide residential > street, while car > driver would only see it as a problem when passing another > car. Actually the only way to get consistent road width estimates is to calculate it based on lanes, otherwise you will get a wide range of subjective values because people will estimate differently no matter what you do or say or write on a wiki. Counting lanes = easy estimating road width = nightmare ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Sunday 02 August 2009 11:57:12 John Smith wrote: > --- On Sun, 2/8/09, Blaž Lorger wrote: > > Unfortunately lanes only specifies number of lanes. In > > general every road that > > is not one way has at least 2 lanes, even if it is narrow, > > say 3.5 meters. > > Even one way roads can have multiple lanes. Like dual carriage ways :) > > > Some good practices > > on how to accurately estimate road width could also came > > handy. > > I feel you could get road widths accurately enough based simply on the > number of lanes and the type of highway. > > For example, a residential road with 2 lanes, one in each direction is > usually about 6-7m wide. > > Motorways are usually the numbers of lanes times 3m + 3m each side of the > road way to allow for break downs, or maybe you could even have lanes=* > breakdown_lanes=* to make this more accurate, but in any case a 3 lane > motorway will be about 15m wide (3 * 3 + 3 + 3). Main point here is to identify narrow roads. This is where road width matters the most. It is obvious that narrow roads will have 2 lanes. Unfortunately, those lanes will be narrow (2m or less). So number of lanes will in no way indicate road width, especially for narrow roads. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Sun, Aug 2, 2009 at 12:18 PM, Blaž Lorger wrote: > Wiki clearly states why tag narrow=yes would be a bad idea > (http://wiki.openstreetmap.org/wiki/Key:width#Using_relative_sizes). > Basically, what is wide for bicycle is narrow for a car, what is wide for a > car is narrow for a truck ... The wiki says "without clear definition" about narrow. We just need a clear definition on the wiki and mine is a proposal. It is just proportional to the default width for all highway types for all countries. The advantage is that you don't have to tag all highways, just the ones you consider narrower. Otherwise, if you use the width=x , you have to do it on all roads or it will never be used by any application since they don't know if this segment is wider or narrower than the 95% of the highways where this width is missing. Here we could start again the discussion about the default width or default lanes and so on per highway type per country. That's the advantage of the key narrow, it is easy to use for mappers and it works for all countries. Pieren ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
--- On Sun, 2/8/09, Blaž Lorger wrote: > Main point here is to identify narrow roads. This is where > road width matters > the most. It is obvious that narrow roads will have 2 > lanes. Unfortunately, > those lanes will be narrow (2m or less). So number of lanes > will in no way > indicate road width, especially for narrow roads. If the main issue is to identify narrow roads, you explicitly state narrow compared to what, and narrow=yes simple. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Sun, 2 Aug 2009, John Smith wrote: > --- On Sun, 2/8/09, Blaž Lorger wrote: > > I also propose extending instructions for road > > classification to use width tag > > I agree with everything else you wrote except width since I really don't > want to get a tape measure out and measure widths of roads, using lanes=* > to estimate widths would be more sensible and is already in use. this is a sensible suggestion, but may i note that there is the "road reserve" which is a certain size - here in Australia actually surveyed in chains ( a very old, old measure) and then a visible road inside that which may be much narrower so i'm not really sure what is the road width when i see these ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Sunday 02 August 2009 12:36:48 Pieren wrote: > On Sun, Aug 2, 2009 at 12:18 PM, Blaž Lorger wrote: > > Wiki clearly states why tag narrow=yes would be a bad idea > > (http://wiki.openstreetmap.org/wiki/Key:width#Using_relative_sizes). > > Basically, what is wide for bicycle is narrow for a car, what is wide for > > a car is narrow for a truck ... > > The wiki says "without clear definition" about narrow. We just need a > clear definition on the wiki and mine is a proposal. It is just > proportional to the default width for all highway types for all > countries. The advantage is that you don't have to tag all highways, > just the ones you consider narrower. > Otherwise, if you use the width=x , you have to do it on all roads or > it will never be used by any application since they don't know if this > segment is wider or narrower than the 95% of the highways where this > width is missing. > Here we could start again the discussion about the default width or > default lanes and so on per highway type per country. That's the > advantage of the key narrow, it is easy to use for mappers and it > works for all countries. Let's see: 1. There is no clear definition what is narrow. 2. There is no specification for default width of road type. 3. If narrow=yes is not applied everywhere where it should be it is equally useful/useless as with width tag. 4. At the end it is always up to the individual mapper to decide what is narrow. While 1 meter is 1 meter. 5. Should definition of "default road width" ever change. All narrow=* tagging will be completely useless and will have to be reevaluated from scratch. Actually it will be useless before that due to subjective nature of value assigned to tag. 6. You will actually require large number of values for "narrow" to even approach granularity offered by one simple tag "width". Either you will have to have narrow=no|foot|bicycle|motorcycle|car|suv|lgv|hgv|... Vale yes could not be used, since it does not specify how narrow the road is or it could be equivalent for narrow=car. 7. You must prepare clear enough instructions how to select value for narrow, to reduce subjective factor to minimum. 8. You must get renderers to support it. On the other hand, if you use tag width, which is already established tag if I may add, you have to accomplish following: 1. You must get renderers to support it. 2. Prepare some guidelines how to estimate road width. Of course using measuring equipment is always preferred but less realistic. 3. Deprecate tag est_width and always store width data in tag width. Add additional tag that would state accuracy of width data. This is really not necessary, but will be easier to use by the software and probably also easier to map. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Sun, Aug 2, 2009 at 1:39 PM, Blaž Lorger wrote: > Let's see: > 1. There is no clear definition what is narrow. > 2. There is no specification for default width of road type. > 3. If narrow=yes is not applied everywhere where it should be it is equally > useful/useless as with width tag. Adding narrow=yes to your unclassified highway meaning "it is 50% to 75% narrower than the usual unclassified highway width in my country" is universal and doesn't have to be added everywhere. When you add oneway=yes to a road, do you add oneway=no to all others ? When it is not attached, then it is the usual width of the unclassified highway in your country. And this can be used by any renderer who could draw thinner roads when the tag is present. But renderers will never draw the exact width=* of a road excepted at the very high zoom levels. > 4. At the end it is always up to the individual mapper to decide what is > narrow. While 1 meter is 1 meter. You always have some subjectivity when you map, look the other thread about residential vs unclassified. Below you say yourself it must be estimated, so your 1 meter can be 1.5 meters for someone else. Your just give the "impression" that a number is more accurate than an adjective but it is just an impression (excepted if you really measure the width with a tape). > 6. You will actually require large number of values for "narrow" to even > approach granularity offered by one simple tag "width". Either you will have > to > have narrow=no|foot|bicycle|motorcycle|car|suv|lgv|hgv|... Vale yes could not > be used, since it does not specify how narrow the road is or it could be > equivalent for narrow=car. I only suggest narrow=yes. I don't understand what means narrow=foot|etc. It is narrower than the default width of the road. It has nothing to do with vehicules. > 7. You must prepare clear enough instructions how to select value for narrow, > to reduce subjective factor to minimum. no values for narrow, it is a rough estimate that the road you are adding as "residential" is not the usual residential width in your country but is narrower. A typical example here in Europe is in towns or villages centers where some old streets are very narrow and usually oneway because two cars cannot cross. Tagging them as highway=service is improper because it is real streets and not alleys. > 8. You must get renderers to support it. > > On the other hand, if you use tag width, which is already established tag if I > may add, you have to accomplish following: > 1. You must get renderers to support it. > 2. Prepare some guidelines how to estimate road width. Of course using > measuring equipment is always preferred but less realistic. > 3. Deprecate tag est_width and always store width data in tag width. Add > additional tag that would state accuracy of width data. This is really not > necessary, but will be easier to use by the software and probably also easier > to map. Pieren ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Sunday 02 August 2009 14:40:09 Pieren wrote: > On Sun, Aug 2, 2009 at 1:39 PM, Blaž Lorger wrote: > > Let's see: > > 1. There is no clear definition what is narrow. > > 2. There is no specification for default width of road type. > > 3. If narrow=yes is not applied everywhere where it should be it is > > equally useful/useless as with width tag. > > Adding narrow=yes to your unclassified highway meaning "it is 50% to > 75% narrower than the usual unclassified highway width in my country" > is universal and doesn't have to be added everywhere. When you add > oneway=yes to a road, do you add oneway=no to all others ? When it is > not attached, then it is the usual width of the unclassified highway > in your country. And this can be used by any renderer who could draw > thinner roads when the tag is present. But renderers will never draw > the exact width=* of a road excepted at the very high zoom levels. To my knowledge there is no such thing as usual highway width. There are certain standards for width of newly built roads, but those usually increase over time, which means you will be forced to periodically reevaluate *ALL* "narrow" tagging. Obviously you haven't read my original message carefully. I suggested that only two widths are used by renderer and that border value is determined by renderer based on highway type. Absence of width tag is interpreted as road having "usual" width. This is exactly same as absence of tag narrow. Having actual road width is always more useful than having just some subjective estimate whether road is too narrow or not. Besides rendering you can use it to improve routing based on actual vehicle width/size. > > > 4. At the end it is always up to the individual mapper to decide what is > > narrow. While 1 meter is 1 meter. > > You always have some subjectivity when you map, look the other thread > about residential vs unclassified. Below you say yourself it must be > estimated, so your 1 meter can be 1.5 meters for someone else. Your > just give the "impression" that a number is more accurate than an > adjective but it is just an impression (excepted if you really measure > the width with a tape). Well yes, but with "width" it is only estimation error. While with narrow you must decide what is usual width for specific type of road, you make estimation error for road width, you make calculation error in percentage of usual road width and you must decide whether calculated percentage width is low enough to justify narrow=true. some subjective factor is inevitable, but at least it should be kept as low as possible. Besides in case of dispute or for whatever reason, road can actually be measured. Whether road is narrow or not is always matter of opinion, there is no way to improve accuracy here. > > 6. You will actually require large number of values for "narrow" to even > > approach granularity offered by one simple tag "width". Either you will > > have to have narrow=no|foot|bicycle|motorcycle|car|suv|lgv|hgv|... Vale > > yes could not be used, since it does not specify how narrow the road is > > or it could be equivalent for narrow=car. > > I only suggest narrow=yes. I don't understand what means > narrow=foot|etc. It is narrower than the default width of the road. It > has nothing to do with vehicules. Again, what is default width of the road? And it has everything to do with vehicles, because what you need to now is whether your vehicle, which can be 1, 2, 4 ... meters wide will be able to drive along road in question. Specifying narrow=yes|no can only be used to render a map. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
2009/8/2 Blaž Lorger : > Again, what is default width of the road? this depends on the type of road and on the footways/pavements and other present. Look here at the pictures on the right to identiy different elements: http://de.wikipedia.org/wiki/Stra%C3%9Fenquerschnitt some standards/proposed widths can be found here (again, it's in German but you can look at the pictures and tables): http://www.isl.uni-karlsruhe.de/module/strassenanlagen/querschnitt/querschnitt.html > And it has everything to do with > vehicles, because what you need to now is whether your vehicle, which can be > 1, 2, 4 ... meters wide will be able to drive along road in question. actually (in Germany, but I doubt that this differs a lot from other countries) a vehicle can not be more than 2,55m wide and 4 meters high. Otherwise it is not generally allowed to drive on roads (it must drive in exceptional convoys accompagnied by policecars). cheers, Martin ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Sun, Aug 2, 2009 at 11:38 PM, Blaž Lorger wrote: > To my knowledge there is no such thing as usual highway width. There are > certain standards for width of newly built roads, but those usually increase > over time, which means you will be forced to periodically reevaluate *ALL* > "narrow" tagging. +1 > Having actual road width is always more useful than having just some > subjective estimate whether road is too narrow or not. Besides rendering you > can use it to improve routing based on actual vehicle width/size. +1. > some subjective factor is inevitable, but at least it should be kept as low as > possible. +1 Tagging width=* is more faithful to what actually exists on the ground, which is always the better long term approach. And the meaning is much clearer than narrow=*, especially for those who only skim the wiki. Also, width=* interacts nicely with lanes=* - from these two tags you can see the width of the entire way, and also calculate the width of each lane. Whereas with narrow=*, it's not quite as clear whether this refers to narrow lanes or a narrow way... ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
And by the way, the Key:width wiki page is horrible and could do with a rework after this discussion. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Sun, Aug 2, 2009 at 11:14 PM, Roy Wallace wrote: > On Sun, Aug 2, 2009 at 11:38 PM, Blaž Lorger wrote: >> To my knowledge there is no such thing as usual highway width. There are >> certain standards for width of newly built roads, but those usually increase >> over time, which means you will be forced to periodically reevaluate *ALL* >> "narrow" tagging. > > +1 I'm not sure that the width of what we consider unclassified roads will double in the next century. >> Having actual road width is always more useful than having just some >> subjective estimate whether road is too narrow or not. Besides rendering you >> can use it to improve routing based on actual vehicle width/size. > > +1. >> some subjective factor is inevitable, but at least it should be kept as low >> as >> possible. > > +1 > > Tagging width=* is more faithful to what actually exists on the > ground, which is always the better long term approach. And the meaning > is much clearer than narrow=*, especially for those who only skim the > wiki. I never mentionned narrow=* but narrow=yes, where did you see narrow=* ? > > Also, width=* interacts nicely with lanes=* - from these two tags you > can see the width of the entire way, and also calculate the width of > each lane. Whereas with narrow=*, it's not quite as clear whether this > refers to narrow lanes or a narrow way... > Why don't you think width is for a lane ? oh, ok it is documented on the wiki. Again, width is not less subjective because it is always estimated (deprecating est_width just hides this point), it is missing in most of the highways, it is changing continuously along the roads and a width of 6 meters does not say if an hgv can pass or not, it will never replace the access restriction tags. Pieren ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
2009/8/3 Pieren : > Again, width is not less subjective because it is always estimated > (deprecating est_width just hides this point), actually according to the German ML there is some guys around with laser-distos to measure the real width. > it is missing in most > of the highways, it is changing continuously along the roads and a > width of 6 meters does not say if an hgv can pass or not, it will > never replace the access restriction tags. well, not replace, it is a different datum. And it is quite useful. Certainly it would be even more useful, if there was a definition how to measure (inside road marking, complete with pavement, does the lateral paved area outside the road marking count, etc.). cheers, Martin ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Mon, Aug 3, 2009 at 8:08 AM, Pieren wrote: > I'm not sure that the width of what we consider unclassified roads > will double in the next century. Nevertheless, anything referring to "what we consider" is more variable across time and people than the length of a metre. > I never mentionned narrow=* but narrow=yes, where did you see narrow=* ? I just meant using narrow as a tag, sorry, didn't realise narrow=* had a special meaning. > Again, width is not less subjective because it is always estimated > (deprecating est_width just hides this point), Precision is not synonymous with objectivity. This is important. Width is less subjective because the length of a metre is well-defined. If someone says "I think a metre is this long", and holds out their hands, they can be proven correct or incorrect. If someone says "I think this street is more narrow that what I would consider usual", they cannot be proven correct or incorrect. That is what it means when someone says width in metres is less subjective than a concept of narrowness and of "usual" width. > it is missing in most > of the highways This does not mean it is not a good tag. > it is changing continuously along the roads So? So does the number of lanes, but that doesn't mean lanes is not a good tag. A way can be split where necessary (obviously a trade-off is necessary between precision of width value and number of splits, which would be same in the case of the use of narrow=yes). > a width of 6 meters does not say if an hgv can pass or not, it will > never replace the access restriction tags. So? No one is suggesting it should. narrow=yes has the same issue, but it is even less clear. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Mon, Aug 3, 2009 at 8:18 AM, Martin Koppenhoefer wrote: > Certainly it would be even more useful, if there was a definition how > to measure (inside road marking, complete with pavement, does the > lateral paved area outside the road marking count, etc.). I think this is very important, and probably the biggest issue with a width tag. I would suggest: Tag the width of the surface on which users of the way are expected to travel. For paved ways (roads, cycleways, footpaths, etc), this would normally be between the parallel edges of the paved area (i.e. not including road shoulder, etc). For roads with line marking, users of the way are expected to travel between the lines, so area outside the road marking would not count toward the value of the width tag. For unpaved ways, the definition does not change - "the surface on which users of the way are expected to travel". ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
2009/8/3 Roy Wallace : > On Mon, Aug 3, 2009 at 8:18 AM, Martin > Koppenhoefer wrote: >> Certainly it would be even more useful, if there was a definition how >> to measure (inside road marking, complete with pavement, does the >> lateral paved area outside the road marking count, etc.). > > I think this is very important, and probably the biggest issue with a > width tag. I would suggest: > > Tag the width of the surface on which users of the way are expected to travel. I agree and would like to add: "and that is not constricted in the full usable height" Sorry for my English, feel free to put it better, I try to explain: it is not about the height but the surface must be available in the full height, if there are obstacles protruding into the way, this width does not count. For plants I'm less sure here, as they tend to grow (yes, really) and after a while are cut though. So maybe it will only be about solid obstacles (say incl. trees) but not bushes and the like. > For paved ways (roads, cycleways, footpaths, etc), this would normally > be between the parallel edges of the paved area (i.e. not including > road shoulder, etc). For roads with line marking, users of the way are > expected to travel between the lines, so area outside the road marking > would not count toward the value of the width tag. well, why not outside the lines? If you really have to know the width of the road (transport or similar, or you want to calculate the sealed area), you won't care about lines. Otherwise you won't need the width tag, because as I pointed out in another post: all usual vehicles (in Germany and probably Europe) must be inside 2,55 width and 4,00 m height. Otherwise they can not travel without beeing accompanied by policecars and other expensive stuff (like special permits, ...). > For unpaved ways, the definition does not change - "the surface on > which users of the way are expected to travel". yes, in this case the tag will be highly subjective. Besides that unpaved ways tend to change continuously their width (be it along them as in different seasons), there will also be need of interpretation about the limits. Still a very useful tag, and good to distinguish a 30 cm footway from a 1,50 m one. If someone else sees them as 50 cm and 1,65 m, it still remains usefull. cheers, Martin ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Mon, Aug 3, 2009 at 9:39 AM, Martin Koppenhoefer wrote: >> Tag the width of the surface on which users of the way are expected to >> travel. > I agree and would like to add: "and that is not constricted in the > full usable height" I think the maxheight tag should be used here. There is no need to complicate the definition of width. If there is a large obstacle, then the width under that obstacle would not be included if and only if "users of the way are NOT expected to travel" under that obstacle. >> For paved ways (roads, cycleways, footpaths, etc), this would normally >> be between the parallel edges of the paved area (i.e. not including >> road shoulder, etc). For roads with line marking, users of the way are >> expected to travel between the lines, so area outside the road marking >> would not count toward the value of the width tag. > > well, why not outside the lines? If you really have to know the width > of the road (transport or similar, or you want to calculate the sealed > area), you won't care about lines. Because users are not expected to travel outside the lines. It also removes the need to consider the quality of the road outside the lines, e.g. if there's gravel next to a paved road, does that count? What about a drop-off? etc., etc. The lines are there for a reason, and that is to mark the width of the road that is designated as suitable for driving on. I think that's the most suitable width to tag. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
2009/8/3 Roy Wallace : > On Mon, Aug 3, 2009 at 9:39 AM, Martin > Koppenhoefer wrote: >>> Tag the width of the surface on which users of the way are expected to >>> travel. >> I agree and would like to add: "and that is not constricted in the >> full usable height" > > I think the maxheight tag should be used here. no. I tried to explain, but I was aware that it might be not understandable. I am not talking about height. It's about width. But the width is only then available, if there are no obstacles above. In your definition you were defining the width by the width of the "surface on which users of the way are expected to travel.". This should include that above this surface (I would suggest up to 4,4 meters or up to maxheight where available) there are no obstacles, because otherwise literally it is not complete. The technical correct term in German is "Lichtraumbreite" (my dictionaries don't know it in English, maybe someone else here can help us). > There is no need to > complicate the definition of width. If there is a large obstacle, then > the width under that obstacle would not be included if and only if > "users of the way are NOT expected to travel" under that obstacle. it's IMHO not about complication but about completeness. And it doesn't matter if the obstacle is large or small, it matters if is removable or not. >> well, why not outside the lines? If you really have to know the width >> of the road (transport or similar, or you want to calculate the sealed >> area), you won't care about lines. > > Because users are not expected to travel outside the lines. It also > removes the need to consider the quality of the road outside the > lines, e.g. if there's gravel next to a paved road, does that count? well, it might be interesting to know under certain conditions about this as well, but I agree that this gravel should be put into other tags (e.g. "shoulder", "shoulder:width", "shoulder:surface"). But why not put the width from line to shoulder, still paved, into the width-tag? You are not expected to use this, but you can do. > What about a drop-off? etc., etc. The lines are there for a reason, > and that is to mark the width of the road that is designated as > suitable for driving on. I think that's the most suitable width to > tag. actually I would consider the lines part of the lanes, not of the road. So I would see the width between the inner border of the lines as lanes:width (gets more complicated with different widths of the lanes, but this is a general problem in OSM: currently can't model lanes as they are). This results in a hierarchical model: 1. entire road-construction, consisting of 2. paved road, shoulders, beam barrier, separators, bed, 3. the paved roads furthermore consistentent of different lanes where each level can have it's own tags for e.g. width, surface, maxspeed, maxheight, maxweight, access restrictions, etc. which would be inherited to the sublevels if there was not the same tag overriding it. cheers, Martin ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
2009/8/3 Martin Koppenhoefer : The technical correct > term in German is "Lichtraumbreite" (my dictionaries don't know it in > English, maybe someone else here can help us). now I found it, maybe it's "clearance width" Martin ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Monday 03 August 2009 12:18:14 Emilie Laffray wrote: > > 4. At the end it is always up to the individual mapper to decide what is > > narrow. While 1 meter is 1 meter. > > Yes, 1 meter is 1 meter. That's why using an approximation is actually > worse than using a relative factor. Using a precise number is going to > introduce errors that can be quite bad in the end. I strongly oppose any > tags that is using a measurement that WILL NOT be accurate. Height is fine > because it is defined. Relative parameters are more flexible. We generally agree that there needs to be a way to assure quality of entered data. Using two distinct tags for measured and estimated values (width and est_width) is one possible approach. But in my opinion it is not the right approach: a. You must be careful when entering the data to use appropriate tag. It is easier to use tag width in place of est_width. So most people will probably use width instead of est_width. In such dual tag approach using tag width for estimated width and measured_width for exact width is likely to give better results, since people that put more effort in gathering data are more likely to put more effort in entering such data in database. b. But dual tag approach is also problematic for software that uses data. Either software has to use exact value with fallback to estimated value, or more likely, software will only use one value and ignore the other. This in turn will likely cause that mappers will use tag supported by software regardless if their data is accurate or not. What I'm proposing is to add additional quality assurance tags. Absence of such tags would mean that there is no way to know how accurate data is. But presence of such tags would give reasonable assurance on data quality. I see need for two such tags: measurement method and date of data acquisition. So, when road width is just roughly estimated mapper would add only tag width. When road width is actually measured, tagging would look like this: width=6.5 width:method=tape width:date=2009-07-23 When width is measured on aerial photography method could be aerial and date would be date when photography was taken, ... This approach can be used with all values that require measurement. It gives a way to quickly gather rough and inaccurate data with a way to identify inaccurate data, so it can be improved upon. > > > 5. Should definition of "default road width" ever change. All narrow=* > > tagging > > will be completely useless and will have to be reevaluated from scratch. > > Actually it will be useless before that due to subjective nature of value > > assigned to tag. > > Roads don't change width every day. Any change in the road will likely mean > that it has been redone and therefore would probably need to be retagged if > you want to be fully accurate in the first place. > > > 6. You will actually require large number of values for "narrow" to even > > approach granularity offered by one simple tag "width". Either you will > > have to > > have narrow=no|foot|bicycle|motorcycle|car|suv|lgv|hgv|... Vale yes could > > not > > be used, since it does not specify how narrow the road is or it could be > > equivalent for narrow=car. > > I disagree. It is a relative parameter not to the vehicles but to the roads > tag used. But in all honesty, it will only be applied on the smaller subset > of the roads to be properly meaningful. > > > 7. You must prepare clear enough instructions how to select value for > > narrow, > > to reduce subjective factor to minimum. > > I believe that the subjective factor is no worse and actually better than > using approximate width. It is something that is relative. > > > 8. You must get renderers to support it. > > Like everything else, like width.. > > > On the other hand, if you use tag width, which is already established tag > > if I > > may add, you have to accomplish following: > > 1. You must get renderers to support it. > > 2. Prepare some guidelines how to estimate road width. Of course using > > measuring equipment is always preferred but less realistic. > > 3. Deprecate tag est_width and always store width data in tag width. Add > > additional tag that would state accuracy of width data. This is really > > not necessary, but will be easier to use by the software and probably > > also easier > > to map. > > This is ridiculous. I don't see how you can give guidelines to estimate > road width. One of the problem with that is that in some areas the roads > width will be fluctuating. In residential areas, it is common for streets > to be changing. > Deprecating est_width is a bit ridiculous. You are just saying in the end > that width is estimated and therefore cannot be relied upon. Numbers and > units are not something to be toyed with. The estimated factor clearly says > it all. It is estimated. If you change this, width will not have a proper > meaning as you will be mixing two qualities (estimated very broadly, and
Re: [OSM-talk] tagging roads
2009/8/3 Blaž Lorger : > On Monday 03 August 2009 12:18:14 Emilie Laffray wrote: > We generally agree that there needs to be a way to assure quality of entered > data. Using two distinct tags for measured and estimated values (width and > est_width) is one possible approach. But in my opinion it is not the right > approach: > > a. You must be careful when entering the data to use appropriate tag. It is > easier to use tag width in place of est_width. So most people will probably > use width instead of est_width. this might be true, but most probably people won't tag any width when just mapping superficiously. I agree that it would have been better to have width for the estimated width and something like "precise_width" or "measured_width" for measured values. > In such dual tag approach using tag width for > estimated width and measured_width for exact width is likely to give better > results, since people that put more effort in gathering data are more likely > to > put more effort in entering such data in database. +1 > b. But dual tag approach is also problematic for software that uses data. > Either software has to use exact value with fallback to estimated value, or > more likely, software will only use one value and ignore the other. This in > turn will likely cause that mappers will use tag supported by software > regardless if their data is accurate or not. maybe you should be more confident in the others ;-). I don't think that most of our width-tags is useless because unprecise or badly estimated. > What I'm proposing is to add additional quality assurance tags. Absence of > such tags would mean that there is no way to know how accurate data is. But > presence of such tags would give reasonable assurance on data quality. I see > need for two such tags: measurement method and date of data acquisition. > So, when road width is just roughly estimated mapper would add only tag width. > When road width is actually measured, tagging would look like this: > width=6.5 > width:method=tape > width:date=2009-07-23 > When width is measured on aerial photography method could be aerial and date > would be date when photography was taken, ... actually there is already some people using tags like source and note to add this kind of information. But they are few I guess. > This approach can be used with all values that require measurement. It gives a > way to quickly gather rough and inaccurate data with a way to identify > inaccurate data, so it can be improved upon. IMHO good idea. Did you check if there are already proposals to tag this kind of datum? Of course also tags like can never make sure, that they don't end up at some other way (someone copies the tags to another way, some roads are combined, etc.), but they promise to have information about the source. >> > 5. Should definition of "default road width" ever change. All narrow=* >> > tagging >> > will be completely useless and will have to be reevaluated from scratch. >> > Actually it will be useless before that due to subjective nature of value >> > assigned to tag. yes I agree, "narrow" does not seem to be a appropriate tag to tag the width. cheers, Martin ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Mon, Aug 3, 2009 at 10:23 PM, Martin Koppenhoefer wrote: > >>> Tag the width of the surface on which users of the way are expected to > >>> travel. > >> I agree and would like to add: "and that is not constricted in the > >> full usable height" > > > > I think the maxheight tag should be used here. > > This should include that above this surface (I would suggest up to 4,4 > meters or up to maxheight where available) there are no obstacles, > because otherwise literally it is not complete. Introducing 4,4 meters is arbitrary - don't like this idea. Maxheight is already established for this purpose. You are right, though, that if width=y and maxheight=x, it means there is a width of y metres with at least x metres of clearance above the surface of the way. > it's IMHO not about complication but about completeness. And it > doesn't matter if the obstacle is large or small, it matters if is > removable or not. Width tag and maxheight tag gives completeness. You don't need to mix the meaning of the two tags together. > But why not put the width from line to shoulder, still paved, into the > width-tag? You are not expected to use this, but you can do. So you think the width should be defined by what "you can" use? That is not good, because it depends who "you" are (e.g. car, bike, pedestrian, in a wheelchair). I believe it is against the law in Australia, for example, to drive outside the line markings. So "you CAN'T" do this. > actually I would consider the lines part of the lanes, not of the > road. So I would see the width between the inner border of the lines > as lanes:width (gets more complicated with different widths of the > lanes, but this is a general problem in OSM: currently can't model > lanes as they are). Well, lines are part of the lanes, and lanes form the road. So the sum of the widths of the lanes should be about equal to the width of the road. > This results in a hierarchical model: > 1. entire road-construction, consisting of > 2. paved road, shoulders, beam barrier, separators, bed, > 3. the paved roads furthermore consistentent of > different lanes > > where each level can have it's own tags for e.g. width, surface, > maxspeed, maxheight, maxweight, access restrictions, etc. which would > be inherited to the sublevels if there was not the same tag overriding > it. Sounds good to me. But if you want to keep it simple (which is usually the case), I still stand by my definition of width of a way: "the width of the surface on which users of the way are expected to travel". ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Tue, Aug 4, 2009 at 7:06 AM, Martin Koppenhoefer wrote: >> What I'm proposing is to add additional quality assurance tags. Absence of >> such tags would mean that there is no way to know how accurate data is. But >> presence of such tags would give reasonable assurance on data quality. I see >> need for two such tags: measurement method and date of data acquisition. >> So, when road width is just roughly estimated mapper would add only tag >> width. >> When road width is actually measured, tagging would look like this: >> width=6.5 >> width:method=tape >> width:date=2009-07-23 >> When width is measured on aerial photography method could be aerial and date >> would be date when photography was taken, ... > > actually there is already some people using tags like source and note > to add this kind of information. But they are few I guess. +1 - use source:width, not width:method. It's established and documented already. As for date, I would think the date of the changeset, which is recorded automatically, should be sufficient. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Tue, Aug 4, 2009 at 3:00 AM, Blaž Lorger wrote: > On Monday 03 August 2009 12:18:14 Emilie Laffray wrote: >> Yes, 1 meter is 1 meter. That's why using an approximation is actually >> worse than using a relative factor. Using a precise number is going to >> introduce errors that can be quite bad in the end. I strongly oppose any >> tags that is using a measurement that WILL NOT be accurate. Height is fine >> because it is defined. Relative parameters are more flexible. This makes no sense. It's not good to be "flexible" in this situation. Please read http://wiki.openstreetmap.org/wiki/Verifiability - particularly the example on height=* Note that verifiability is not the same as precision. Use source:width=approx, or something like that to indicate how precise your measurement is. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Tue, Aug 4, 2009 at 10:30 AM, Martin Koppenhoefer wrote: > yes, you're right, 4,40 m was indeed wrong. In the EU it is 4,50 m. > That's the general maxheight (the clearance streets must have), > resulting from 4,00 maxheight for the vehicle plus 50 cm clearance. > This might differ on other continents. This could be the default, so > we don't have to post a maxheight on all streets that don't have > signs. Just in case the clearance is below 4,50 there will be a > maxheight-sign. In my opinion, this has nothing to do with width=*. But you're free to disagree, of course. > I want to make it clear in the width-definition which height must be > available. Otherwise there will be confusion in some cases. An example might help. > In Germany it depends. If you are a car, you must not use [ the width from > line to shoulder], > if you are a bike or pedestrian and outside town, you should use it if there > is no cycleway (or footway for pedestrians). If you are planning a > special transport, you will be interested in this data. If you drive a > car, you won't need this data, because you can be sure that you will > fit on a street. ... > my proposal would result in width=lanes+marginal strip. Marginal strip > is not where you are expected to travel but it is a elemental part of > the road. For sidewalks I'm unsure. maybe it's better to have a > width:total where they are included and in normal simple "width" they > aren't. In my opinion, "marginal strip" and "elemental part of the road" is a little tricky to define for all kinds of ways. And width:total seems strange to me at first glance. Why isn't width = width:total? There we have it, my definition and your definition. The floor's open... ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
2009/8/4 Roy Wallace : > In my opinion, "marginal strip" and "elemental part of the road" is a > little tricky to define for all kinds of ways. yes, that's surely for streets only, small ways won't have this to be tagged. > And width:total seems > strange to me at first glance. Why isn't width = width:total? well, I'd separate width (for width of the "drivable" street) from a total width that could include sidewalks/pavements, shoulders, parkinglots, etc. I wouldn't put pavements into the same width as the lanes for driving. cheers, Martin ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
Interesting - I've measured the widths of most of the main roads in Oxford, mostly at quiet times of day (easy enough with a wheely device - I wouldn't recommend tape). I do kerb-kerb. My inclination would be to put widths on nodes, since they are measured at points, but that might not be too helpful for renderers. But I don't think I really want to break a way every time I do a measurement (I did one particular stretch of road every 10m). Richard ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
2009/8/5 Richard Mann : > Interesting - I've measured the widths of most of the main roads in Oxford, > mostly at quiet times of day (easy enough with a wheely device - I wouldn't > recommend tape). I do kerb-kerb. yes, that seems reasonable in urban context. Do you do the same if there is parking lots along the way? In this case I would probably measure where there aren't to indicate the width of the way (because otherwise - I was thinking of putting the tags to the way - you really would have to split the way every 10 meters). I don't like the idea of putting the width to nodes that much, as nodes tend to get moved - but maybe with more width attached to them, this would change and people get more cautious. What would you measure out of town? > My inclination would be to put widths on nodes, since they are measured at > points, but that might not be too helpful for renderers. But I don't think I > really want to break a way every time I do a measurement (I did one > particular stretch of road every 10m). did you find a lot of differences every 10m? I thought that most streets remain there width (for the "driving zone"). cheers, Martin ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
On Wed, Aug 5, 2009 at 9:47 AM, Richard Mann wrote: > My inclination would be to put widths on nodes, since they are measured at > points, but that might not be too helpful for renderers. But I don't think I > really want to break a way every time I do a measurement (I did one > particular stretch of road every 10m). No, I would mark width on a way - just use your judgement as to when the way needs to be split. Apply the width to a section of way, and the width should describe roughly the narrowest width on that section. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
I wouldn't include parking bays if the kerb is built out around them. Generally I'd measure the running carriageway, but include any central islands. The road I measured every 10m had widths varying between 7.7m and 9.3m over about 50m, with no change in lane markings. Richard On Wed, Aug 5, 2009 at 1:12 AM, Martin Koppenhoefer wrote: > 2009/8/5 Richard Mann : > > Interesting - I've measured the widths of most of the main roads in > Oxford, > > mostly at quiet times of day (easy enough with a wheely device - I > wouldn't > > recommend tape). I do kerb-kerb. > > yes, that seems reasonable in urban context. Do you do the same if > there is parking lots along the way? In this case I would probably > measure where there aren't to indicate the width of the way (because > otherwise - I was thinking of putting the tags to the way - you really > would have to split the way every 10 meters). I don't like the idea of > putting the width to nodes that much, as nodes tend to get moved - but > maybe with more width attached to them, this would change and people > get more cautious. What would you measure out of town? > > > My inclination would be to put widths on nodes, since they are measured > at > > points, but that might not be too helpful for renderers. But I don't > think I > > really want to break a way every time I do a measurement (I did one > > particular stretch of road every 10m). > > did you find a lot of differences every 10m? I thought that most > streets remain there width (for the "driving zone"). > > cheers, > Martin > ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] tagging roads
2009/8/5 Richard Mann : > I wouldn't include parking bays if the kerb is built out around them. > Generally I'd measure the running carriageway, but include any central > islands. +1 > The road I measured every 10m had widths varying between 7.7m and 9.3m over > about 50m, with no change in lane markings. that's a good way to go if you really, really want to be accurate, but if you're short on time you would map this as 7.7m and for many cases this would be sufficient. If you go into micromapping I would consider mapping the road as an area (additionally), just like we're already doing for squares. cheers, Martin ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk