Re: [OSM-talk] Cycleways and Access tags: Left, Right, Forward, Backward?

2012-05-18 Thread Tobias Knerr
18.05.2012 12:52 Martin Koppenhoefer:
> 2012/5/17  :
>> cycleway:right=* - 9190 occurrences
>> cycleway:left=* - 4329 occurrences
[...]
> maybe oneway-streets do more often have a cycleway only on one side of
> the road, and they are mostly drawn in the direction of traffic flow
> (although -1 is a valid alternative). Given that most of the ways in
> the db are in countries with traffic on the right, but GB is strong
> for mapping cycleways this could explain those numbers.

Yes, this is the most likely explanation. If you look at the taginfo
"combinations" for cycleway:right=*, you will notice that more than half
of these ways are oneway roads:
http://taginfo.openstreetmap.org:8001/keys/cycleway%3Aright#combinations

And almost all oneway roads are drawn accordint to traffic direction,
with the "-1" value at only 1.5%:
http://taginfo.openstreetmap.org:8001/keys/oneway#values

So a large share of the roads with direction-dependent cycleway tags do
actually _not_ have a 50% chance of being drawn either way, because they
are also oneways.

Tobias

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] Cycleways and Access tags: Left, Right, Forward, Backward?

2012-05-18 Thread Richard Mann
I think we should spend time making maps and not having silly arguments
that aren't going to be resolved because there are pros and cons both ways.

Rob - for your purposes - the wiki should describe simple versions of both
methods (ignore left and right), credit them both with having virtues, and
probably advise people not to go round deleting things if there's no
immediate need.

Some of this won't be resolved until there's been some tool development,
and as RichardF is wont to remind us, that doesn't happen by itself.

Richard

On Fri, May 18, 2012 at 2:20 PM, Richard Fairhurst wrote:

> Martin Koppenhoefer wrote:
> > simply draw cycleways with separate carriageways like any
> > other highway with its own way in OSM and you resolve
> > lots of issues, including distinct surfaces and restrictions.
>
> Yes. Absolutely that.
>
> Things like cycleway=track were a hack back in the day when we only had a
> few mappers and barely usable tools, and we needed to grow our coverage as
> fast as possible. That's not the case now. We can spend the time to map
> things properly (hippy), and we should.
>
> cheers
> Richard
>
>
>
> --
> View this message in context:
> http://gis.19327.n5.nabble.com/Cycleways-and-Access-tags-Left-Right-Forward-Backward-tp5709253p5709424.html
> Sent from the General Discussion mailing list archive at Nabble.com.
>
> ___
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk
>
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] Cycleways and Access tags: Left, Right, Forward, Backward?

2012-05-18 Thread Richard Fairhurst
Martin Koppenhoefer wrote:
> simply draw cycleways with separate carriageways like any 
> other highway with its own way in OSM and you resolve 
> lots of issues, including distinct surfaces and restrictions.

Yes. Absolutely that.

Things like cycleway=track were a hack back in the day when we only had a
few mappers and barely usable tools, and we needed to grow our coverage as
fast as possible. That's not the case now. We can spend the time to map
things properly (hippy), and we should.

cheers
Richard



--
View this message in context: 
http://gis.19327.n5.nabble.com/Cycleways-and-Access-tags-Left-Right-Forward-Backward-tp5709253p5709424.html
Sent from the General Discussion mailing list archive at Nabble.com.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] Cycleways and Access tags: Left, Right, Forward, Backward?

2012-05-18 Thread Martin Koppenhoefer
simply draw cycleways with separate carriageways like any other
highway with its own way in OSM and you resolve lots of issues,
including distinct surfaces and restrictions. You also get more stable
data which is not dependent on the direction of another way, and you
can be sure that most applications will be able to use your data due
to default tags.

cheers,
Martin

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] Cycleways and Access tags: Left, Right, Forward, Backward?

2012-05-18 Thread Richard Mann
Off-carriageway tracks tend to be bidirectional (they all are in the UK).
So no-one would bother to use bidirectional_track.

Richard


On Fri, May 18, 2012 at 1:56 PM,  wrote:

> Arg, thats still not right is it? Firstly it leaves 2 values for
> cycleway:left and also the opposite_track is not quite right because if it
> had been on the right hand side of the way (but the way was still
> orientated in the same direction) you would have to use
> cycleway:right=opposite_track to imply that the cycle track flows in the
> same direction of the traffic closest to it (i.e. the traffic on the right
> that is going in the opposite direction of how the way is drawn).
>
> Maybe it should be "The track may be cycled in the opposite direction of
> other traffic using the same side (i.e. right / left) of the road"
>
> Finally the talk page has a good suggestion to use
> cylceway:left=bidirectional_track but this has no uses according to tag
> info.
>
> Rob
>
>
>
> On , rob.j.nicker...@gmail.com wrote:
> > Oh, I think I get it now. So for example if you had a cycle route that
> runs parallel to a road (but not within the road carriageway), is on the
> left side and allows cycling in both directions it would be tagged as:
> >
> > * highway=*
> > * cycleway:left=track
> > * cycleway:left=opposite_track
> >
> > If this is correct that the wiki could do with the definition changing
> so that opposite_track reads "The track may be cycled in the opposite
> direction to the way", with an explanation of how a way is essentially an
> arrow.
> >
> > Regards,
> > Rob
> >
> > p.s. Sorry for the large number of emails on this. Am trying to get
> things clear in my mind!
> >
> >
> >
> > On , rob.j.nicker...@gmail.com wrote:
> > > Are there any cases of that?
> > >
> > > Wouldn't you assume that the cycle lane is on the same side as the
> flow of traffic (so forward would imply left in GB)? Surely the same
> argument can be made that cycleway:left tells you which side of the road it
> is on but doesn't tell you the direction of flow.
> > >
> > > Rob
> > >
> > >
> > >
> > > On , Ed Loach e...@loach.me.uk> wrote:
> > > > Imagine a two way road with a cycleway on one side. Neither forward
> or backward tell you to which side of a way the cycleway is – this would
> only tell you if it were in the same direction as the way, rather than on
> one side of it. So forward, backward, left, right all have their places. Ed
> > > >
> > > >
>
> ___
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk
>
>
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] Cycleways and Access tags: Left, Right, Forward, Backward?

2012-05-18 Thread rob . j . nickerson
Arg, thats still not right is it? Firstly it leaves 2 values for  
cycleway:left and also the opposite_track is not quite right because if it  
had been on the right hand side of the way (but the way was still  
orientated in the same direction) you would have to use  
cycleway:right=opposite_track to imply that the cycle track flows in the  
same direction of the traffic closest to it (ie the traffic on the right  
that is going in the opposite direction of how the way is drawn).


Maybe it should be "The track may be cycled in the opposite direction of  
other traffic using the same side (ie right / left) of the road"


Finally the talk page has a good suggestion to use  
cylceway:left=bidirectional_track but this has no uses according to tag  
info.


Rob



On , rob.j.nicker...@gmail.com wrote:
Oh, I think I get it now. So for example if you had a cycle route that  
runs parallel to a road (but not within the road carriageway), is on the  
left side and allows cycling in both directions it would be tagged as:



* highway=*
* cycleway:left=track
* cycleway:left=opposite_track


If this is correct that the wiki could do with the definition changing so  
that opposite_track reads "The track may be cycled in the opposite  
direction to the way", with an explanation of how a way is essentially an  
arrow.



Regards,
Rob


ps Sorry for the large number of emails on this. Am trying to get things  
clear in my mind!





On , rob.j.nicker...@gmail.com wrote:
> Are there any cases of that?
>
> Wouldn't you assume that the cycle lane is on the same side as the flow  
of traffic (so forward would imply left in GB)? Surely the same argument  
can be made that cycleway:left tells you which side of the road it is on  
but doesn't tell you the direction of flow.

>
> Rob
>
>
>
> On , Ed Loach e...@loach.me.uk> wrote:
> > Imagine a two way road with a cycleway on one side. Neither forward  
or backward tell you to which side of a way the cycleway is – this would  
only tell you if it were in the same direction as the way, rather than on  
one side of it. So forward, backward, left, right all have their places.  
Ed

> >
> >
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] Cycleways and Access tags: Left, Right, Forward, Backward?

2012-05-18 Thread rob . j . nickerson
Oh, I think I get it now. So for example if you had a cycle route that runs  
parallel to a road (but not within the road carriageway), is on the left  
side and allows cycling in both directions it would be tagged as:


* highway=*
* cycleway:left=track
* cycleway:left=opposite_track

If this is correct that the wiki could do with the definition changing so  
that opposite_track reads "The track may be cycled in the opposite  
direction to the way", with an explanation of how a way is essentially an  
arrow.


Regards,
Rob

ps Sorry for the large number of emails on this. Am trying to get things  
clear in my mind!




On , rob.j.nicker...@gmail.com wrote:

Are there any cases of that?


Wouldn't you assume that the cycle lane is on the same side as the flow  
of traffic (so forward would imply left in GB)? Surely the same argument  
can be made that cycleway:left tells you which side of the road it is on  
but doesn't tell you the direction of flow.



Rob





On , Ed Loach e...@loach.me.uk> wrote:
> Imagine a two way road with a cycleway on one side. Neither forward or  
backward tell you to which side of a way the cycleway is – this would  
only tell you if it were in the same direction as the way, rather than on  
one side of it. So forward, backward, left, right all have their places.  
Ed

>
>
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] OpenSeaMap

2012-05-18 Thread Gregory
OpenStreetMap can be hard for new people to understand.

OpenStreetMap is not a company. It is a project, a community/group, and a
database. It is run and supported by volunteers that contribute to the map
data or to code that creates and/or uses the data.
There is a long list of companies, organisations, and other groups that are
not "officially" part of OpenStreetMap but they use or contribute to the
project.
OpenStreetMap database is free for anyone to use for their own purposes (so
long as they give credit, and share improvements in the same way), they do
not have to ask or tell anyone that they are using the data.


So Open SEA Map is a group (I don't think it is a company) that uses
OpenStreetMap data to create a map of sea mark icons. It also has an editor
to add data back to OpenStreetMap.
Some people involved with OpenSeaMap also spend a lot of their time
involved with OpenStreetMap. Some of them read messages on this mailing
list.


The phone app is made by a company. That company uses the OpenSeaMap images
and puts them on top of Google Maps for you to see in your phone. I don't
know if they are closely involved with OpenStreetMap or OpenSeaMap, but
they don't need to be. We happily let them take the data we make and they
do not need to pay anyone or tell anyone they exist.


There is also the OpenStreetMap Foundation. That's a not-for-profit
organisation that looks after some servers for the database and the
openstreetmap.org website, wiki, and a mailing list so that people can talk
about the project. What the foundation says is perhaps what is "official"
with OpenStreetMap, but most of the time it lets people do what they like
without trying to give orders or be aware of everything.


On 17 May 2012 21:14, Peter Wendorff  wrote:

> Frans,
> I fear, you still don't understand many things around the OSM universe.
> No, OpenSeaMap is not part of the OSM project.
> Yes, OpenSeaMap is part of the OSM ecosystem, that is build of very few
> core components/services and tons of tools around, that are independent of
> OSM.
>
> Core services of osm are
> - the osm.org website
> - the mapnik rendering on the osm website
> - the api to use by any (independent!) editor
> - one out of several hostings of Potlatch as one possible editor for osm
> data.
> - a forum
> - mailing lists
>
> "Third party" components are
> - all editor software projects, including Potlatch, Potlatch2 and JOSM
> - most other map renderings, including openseamap
> - as far as I know ALL mobile apps using osm data in any way you imagine
>
> Some of these projects are nearby the osm project, but none is "officially
> part of OSM" or whatever you would like to call it.
>
> If any of these projects wants to have a mailinglist in the
> openstreetmap.org namespace, they may ask for one, and if it's a serious
> project in the OSM universe, I think, it's possible to get one; but e.g.
> the JOSM delevopment is done "under" openstreetmap.de, therefore it's
> josm.openstreetmap.de, even for the international audience.
> And it's similar for many other sub-projects like OpenSeaMap (if you're
> right).
>
> regards
> Peter
>
> Am 17.05.2012 19:41, schrieb Frans Thamura:
>
>  strange why the openseamap if it is part of osm, the mailing list is
>> not using @openstreetmap mailing list
>>
>> OpenSeaMap was created in 2009 in response to a great need for
>> freely-accessible seafaring maps. OpenSeaMap's goal is to add nautical
>> and tourism information that would interest sailors OSM, and to
>> present it in a pleasing way. This includes beacons, buoys and other
>> seamarks, port information, repair shops, ship supplies and much more,
>> but also shops, restaurants and places of interest. OpenSeaMap is part
>> of OpenStreetMap and uses its database.
>>
>>
>> the newsgroups sound "old"
>>
>> F
>>
>> __**_
>> talk mailing list
>> talk@openstreetmap.org
>> http://lists.openstreetmap.**org/listinfo/talk
>>
>>
>
> __**_
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.**org/listinfo/talk
>



-- 
Gregory
o...@livingwithdragons.com
http://www.livingwithdragons.com
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] Cycleways and Access tags: Left, Right, Forward, Backward?

2012-05-18 Thread rob . j . nickerson

Are there any cases of that?

Wouldn't you assume that the cycle lane is on the same side as the flow of  
traffic (so forward would imply left in GB)? Surely the same argument can  
be made that cycleway:left tells you which side of the road it is on but  
doesn't tell you the direction of flow.


Rob



On , Ed Loach  wrote:
Imagine a two way road with a cycleway on one side. Neither forward or  
backward tell you to which side of a way the cycleway is – this would  
only tell you if it were in the same direction as the way, rather than on  
one side of it. So forward, backward, left, right all have their places.  
Ed



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] Cycleways and Access tags: Left, Right, Forward, Backward?

2012-05-18 Thread Richard Mann
I don't know when bicycle:backward=yes appeared - I've always used
oneway:bicycle=no
(and taginfo puts it as 131 to 4831 uses, so I'm not the only one)

On Fri, May 18, 2012 at 11:59 AM,  wrote:

> It has been pointed out that some of these may be due to one-way roads
> (and as there are more countries that drive on the right, this would cause
> a slight bias), however in these cases the :right suffix is not always
> needed.
>
> I can have a look for incorrect examples, but irrespective of whether I
> find any, I believe that the wiki page needs updating to better explain
> right/left. I am happy to have a go at doing this however as the lane
> enthusiasts and the 'access' page uses forward / backward instead I
> wondered whether there was any intention or hope to switch at some point in
> the future to bring consistency.
>
> Cheers,
> Rob
>
> p.s. I'm not saying we should change, just wanted to open a discussion to
> allow others to voice their opinion before I add more details to the wiki
> page. :-)
>
>
>
> On , Steve Bennett  wrote:
> > On Fri, May 18, 2012 at 1:02 AM,  rob.j.nicker...@gmail.com> wrote:
> >
> > > My concern with right / left is that some may think "ok we drive on the
> >
> > > right side of the road so it must be cycleway:right" (similarly left
> for
> >
> > > countries such as the UK that drive on the left side of the road). A
> quick
> >
> > > look on TagInfo reveals:
> >
> > >
> >
> > > cycleway:right=* - 9190 occurrences
> >
> > > cycleway:left=* - 4329 occurrences
> >
> > >
> >
> > > A way has a 50/50% chance of being drawn in either direction so (unless
> >
> > > people are reversing ways as they prefer right to left) then you would
> >
> > > expect the split to be closer.
> >
> >
> >
> > Ok, so, you have a hypothesis that people are using the tag incorrectly.
> >
> >
> >
> > 1) Could you find some examples to see if this true
> >
> > 2) What do you think should be done, if so?
> >
> >
> >
> > Steve
> >
> >
> ___
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk
>
>
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] Cycleways and Access tags: Left, Right, Forward, Backward?

2012-05-18 Thread Ed Loach
Imagine a two way road with a cycleway on one side. Neither forward
or backward tell you to which side of a way the cycleway is – this
would only tell you if it were in the same direction as the way,
rather than on one side of it. So forward, backward, left, right all
have their places.

 

Ed

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] Cycleways and Access tags: Left, Right, Forward, Backward?

2012-05-18 Thread rob . j . nickerson
It has been pointed out that some of these may be due to one-way roads (and  
as there are more countries that drive on the right, this would cause a  
slight bias), however in these cases the :right suffix is not always needed.


I can have a look for incorrect examples, but irrespective of whether I  
find any, I believe that the wiki page needs updating to better explain  
right/left. I am happy to have a go at doing this however as the lane  
enthusiasts and the 'access' page uses forward / backward instead I  
wondered whether there was any intention or hope to switch at some point in  
the future to bring consistency.


Cheers,
Rob

ps I'm not saying we should change, just wanted to open a discussion to  
allow others to voice their opinion before I add more details to the wiki  
page. :-)




On , Steve Bennett  wrote:

On Fri, May 18, 2012 at 1:02 AM, rob.j.nicker...@gmail.com> wrote:



> My concern with right / left is that some may think "ok we drive on the



> right side of the road so it must be cycleway:right" (similarly left for


> countries such as the UK that drive on the left side of the road). A  
quick



> look on TagInfo reveals:



>



> cycleway:right=* - 9190 occurrences



> cycleway:left=* - 4329 occurrences



>



> A way has a 50/50% chance of being drawn in either direction so (unless



> people are reversing ways as they prefer right to left) then you would



> expect the split to be closer.





Ok, so, you have a hypothesis that people are using the tag incorrectly.





1) Could you find some examples to see if this true



2) What do you think should be done, if so?





Steve



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] Cycleways and Access tags: Left, Right, Forward, Backward?

2012-05-18 Thread Martin Koppenhoefer
2012/5/17  :
> cycleway:right=* - 9190 occurrences
> cycleway:left=* - 4329 occurrences
>
> A way has a 50/50% chance of being drawn in either direction so (unless
> people are reversing ways as they prefer right to left) then you would
> expect the split to be closer.


maybe oneway-streets do more often have a cycleway only on one side of
the road, and they are mostly drawn in the direction of traffic flow
(although -1 is a valid alternative). Given that most of the ways in
the db are in countries with traffic on the right, but GB is strong
for mapping cycleways this could explain those numbers.

To investigate further you should also look on countryspecific
preferences to use tags vs. explicit geometry.

cheers,
Martin

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk