Hi all Diagnosing what seemed like superfluous navigation pop-up, eg "Turn left on A1", while driving on the A1, I discover quite a few junctions are surrounded by a closed way with tags [highway=unclassified, area=yes] https://www.openstreetmap.org/way/239371846 https://www.openstreetmap.org/way/239371836 https://www.openstreetmap.org/way/239754207 and these "way"s, with the default style (and all other styles I've ever looked at) will generated a road that probably shares many routing points with the real roads in the junction. These seem to be used for documentation of changes and are possibly being create automatically by some OSM editing tool. I haven't found any references to this being something that is acceptable or is wrong, so I don't know if these should be deleted from OSM and/or ignored by the mkgmap style. Thoughts anyone - Ticker
On Sat, Sep 12, 2020 at 05:27:03PM +0100, Ticker Berkin wrote:
Hi all
Diagnosing what seemed like superfluous navigation pop-up, eg "Turn left on A1", while driving on the A1, I discover quite a few junctions are surrounded by a closed way with tags [highway=unclassified, area=yes]
https://www.openstreetmap.org/way/239371846 https://www.openstreetmap.org/way/239371836 https://www.openstreetmap.org/way/239754207
I haven't found any references to this being something that is acceptable or is wrong, so I don't know if these should be deleted from OSM and/or ignored by the mkgmap style.
They look clearly wrong to me, but then I don't have any local knowledge. ael
Yes, look like mapping errors (mapping for the renderer?). Maybe area:highway=* was meant. I would not change the style for them. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von ael <witwall3@disroot.org> Gesendet: Samstag, 12. September 2020 18:52 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] highway=unclassified & area=yes On Sat, Sep 12, 2020 at 05:27:03PM +0100, Ticker Berkin wrote:
Hi all
Diagnosing what seemed like superfluous navigation pop-up, eg "Turn left on A1", while driving on the A1, I discover quite a few junctions are surrounded by a closed way with tags [highway=unclassified, area=yes]
https://www.openstreetmap.org/way/239371846 https://www.openstreetmap.org/way/239371836 https://www.openstreetmap.org/way/239754207
I haven't found any references to this being something that is acceptable or is wrong, so I don't know if these should be deleted from OSM and/or ignored by the mkgmap style.
They look clearly wrong to me, but then I don't have any local knowledge. ael _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Those mappers might have used this new proposal for highway=junction incorrectly. https://wiki.openstreetmap.org/wiki/Proposed_features/highway%3Djunction I'm not sure about their motives or reasoning but clearly a changeset comment or an email to those people is in order. Dave On Sun, Sep 13, 2020 at 12:02 AM Gerd Petermann < gpetermann_muenchen@hotmail.com> wrote:
Yes, look like mapping errors (mapping for the renderer?). Maybe area:highway=* was meant. I would not change the style for them.
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von ael <witwall3@disroot.org> Gesendet: Samstag, 12. September 2020 18:52 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] highway=unclassified & area=yes
On Sat, Sep 12, 2020 at 05:27:03PM +0100, Ticker Berkin wrote:
Hi all
Diagnosing what seemed like superfluous navigation pop-up, eg "Turn left on A1", while driving on the A1, I discover quite a few junctions are surrounded by a closed way with tags [highway=unclassified, area=yes]
https://www.openstreetmap.org/way/239371846 https://www.openstreetmap.org/way/239371836 https://www.openstreetmap.org/way/239754207
I haven't found any references to this being something that is acceptable or is wrong, so I don't know if these should be deleted from OSM and/or ignored by the mkgmap style.
They look clearly wrong to me, but then I don't have any local knowledge.
ael
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
-- Dave Swarthout Homer, Alaska Chiang Mai, Thailand Travel Blog at http://dswarthout.blogspot.com
Hi all Checking the GBR, I detect 1865 of these highway areas. They are done in a consistent and systematic way by a number of mappers. I don't think it is a mapping error or an attempt to define the extent of a junction as per the proposed tagging highway=junction. Rather it is some option on the editing tool or some guidelines they are following that says to do this as a way of documenting the changes they have made within the junction area. I'll ask some of the mappers why they are doing it. These constructs can cause invalid routes to be calculated as well as confusing/irrelevant direction pop-ups and I now ignore them in my style. Ticker
Hi Ticker just for me a hint, how can i ignore this in a style ? Am 13.09.2020 um 09:59 schrieb Ticker Berkin:
Hi all
Checking the GBR, I detect 1865 of these highway areas. They are done in a consistent and systematic way by a number of mappers.
I don't think it is a mapping error or an attempt to define the extent of a junction as per the proposed tagging highway=junction. Rather it is some option on the editing tool or some guidelines they are following that says to do this as a way of documenting the changes they have made within the junction area.
I'll ask some of the mappers why they are doing it.
These constructs can cause invalid routes to be calculated as well as confusing/irrelevant direction pop-ups and I now ignore them in my style.
Ticker
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
-- ##################################################### Viele Grüße und 73 de Manfred Haiduk, DD8KQ e-mail mhaiduk@t-online.de dd8kq@gmx.de #####################################################
Hi In styles / default / lines, around line 181, the change would be: highway=unclassified [0x06 road_class=0 road_speed=3 resolution 21] to: highway=unclassified & area!=yes [0x06 road_class=0 road_speed=3 resolution 21] If you style has a mop-up for unhandled highways, make sure that this isn't triggered Ticker On Sun, 2020-09-13 at 11:09 +0200, DD8KQ wrote:
Hi Ticker
just for me a hint, how can i ignore this in a style ?
In my style I have the following: (highway=motorway | highway=trunk | highway=primary | highway=secondary | highway=tertiary | highway=motorway_link | highway=trunk_link | highway=primary_link | highway=secondary_link | highway=tertiary_link | highway=residential | highway=unclassified | highway=track | highway=bridleway | highway=cycleway | highway=footway | highway=path) & area=yes {delete highway} # delete unwanted areas I leave highway=service and highway=pedestrian as these are valid but the others are not. Regards, Mike -----Original Message----- From: DD8KQ [mailto:dd8kq@gmx.de] Sent: 13 September 2020 10:09 To: Development list for mkgmap <mkgmap-dev@lists.mkgmap.org.uk>; Ticker Berkin <rwb-mkgmap@jagit.co.uk> Subject: Re: [mkgmap-dev] highway=unclassified & area=yes Hi Ticker just for me a hint, how can i ignore this in a style ? Am 13.09.2020 um 09:59 schrieb Ticker Berkin:
Hi all
Checking the GBR, I detect 1865 of these highway areas. They are done in a consistent and systematic way by a number of mappers.
I don't think it is a mapping error or an attempt to define the extent of a junction as per the proposed tagging highway=junction. Rather it is some option on the editing tool or some guidelines they are following that says to do this as a way of documenting the changes they have made within the junction area.
I'll ask some of the mappers why they are doing it.
These constructs can cause invalid routes to be calculated as well as confusing/irrelevant direction pop-ups and I now ignore them in my style.
Ticker
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
-- ##################################################### Viele Grüße und 73 de Manfred Haiduk, DD8KQ e-mail mhaiduk@t-online.de dd8kq@gmx.de #####################################################
Hi I have made a test in the region nearby and got lots of ways with area=yes changed your rule a little bit ;-) ... & area=yes {echo '${highway}'} Most of this ways didn't have a virtual way between each node across the area, so they are possibly lost for routing and create lot of islands, when deleting highway=* I think this problem can't be solved with rules, only in the map data. Maybe creating virtual way (highway=virtual) in the area is a solution, then the area can be used without routing problems Bernd Am Sonntag, 13. September 2020, 18:18:40 CEST schrieb Mike Baggaley:
In my style I have the following:
(highway=motorway | highway=trunk | highway=primary | highway=secondary | highway=tertiary | highway=motorway_link | highway=trunk_link | highway=primary_link | highway=secondary_link | highway=tertiary_link | highway=residential | highway=unclassified | highway=track | highway=bridleway | highway=cycleway | highway=footway | highway=path) & area=yes {delete highway} # delete unwanted areas
I leave highway=service and highway=pedestrian as these are valid but the others are not.
Hi Bernd Once you start looking you see all sorts of mapping errors - I've been finding a lot of roads mapped as long closed ways. What seems like a common fix to this is to create a correctly joined and tagged linear way and leave the area untouched. I don't understand what you mean by creating highway=virtual; if editing OSM you might just as well create a correct road. If defining style rules just respect or ignore the highway type in conjunction with area=yes. I thought about doing what Mike Baggaley does and dropping all but pedestrian/service but I felt this might lose some important roads so I only drop "unclassified" because this seems to be the type chosen for an arbitary commented area. I think some mappers thought "unclassified" was not a real road and so wouldn't be used for routing, rather than the formal OSM definition as a "minor public road". Ticker On Mon, 2020-09-14 at 10:37 +0200, Bernd Weigelt wrote:
Hi
I have made a test in the region nearby and got lots of ways with area=yes
changed your rule a little bit ;-)
... & area=yes {echo '${highway}'}
Most of this ways didn't have a virtual way between each node across the area, so they are possibly lost for routing and create lot of islands, when deleting highway=*
I think this problem can't be solved with rules, only in the map data.
Maybe creating virtual way (highway=virtual) in the area is a solution, then the area can be used without routing problems
Bernd
Am Sonntag, 13. September 2020, 18:18:40 CEST schrieb Mike Baggaley:
In my style I have the following:
(highway=motorway | highway=trunk | highway=primary | highway=secondary | highway=tertiary | highway=motorway_link | highway=trunk_link | highway=primary_link | highway=secondary_link | highway=tertiary_link | highway=residential | highway=unclassified | highway=track | highway=bridleway | highway=cycleway | highway=footway | highway=path) & area=yes {delete highway} # delete unwanted areas
I leave highway=service and highway=pedestrian as these are valid but the others are not.
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
On 12/09/2020 18:02, Gerd Petermann wrote:
Yes, look like mapping errors (mapping for the renderer?). Maybe area:highway=* was meant. I would not change the style for them.
I'm not local to these examples but they're clearly mapping for the renderer. Looking at https://www.openstreetmap.org/way/239754207/history , unclassified makes no sense as two tertiary roads join the A1 here, not unclassified. I _am_ familiar with the ones added by a prolific mapper in Lincolnshire (see e.g. https://www.openstreetmap.org/way/378343668/history ) and examples like that are clearly wrong - "area:highway" describes the situation there, and should be used instead. Ticker Berkin said elsewhere in the thread "They are done in a consistent and systematic way by a number of mappers" - I'd agree that the small number of mappers doing this in the UK (maybe 2 or 3?) are consistent and systematic, but unfortunately this mapping is in my experience consistently wrong*. Contacting the original mappers is an excellent idea, but (at least in Lincs, as http://resultmaps.neis-one.org/osm-discussion-comments?uid=125259 shows) may not get a reply. Best Regards, Andy * That's not to say that "highway=unclassified; area=yes" can't ever be correct - it can; but these examples aren't valid examples of it.
I've found quite a few proper roads mapped as closed ways with [highway=unclassified, area=yes], but in the cases I've looked at so far, there has also been a correct unclosed way to represent the road. I can't think of any method using style rules to detect the case when there isn't this additional road, but my preference is to ignore these areas and so avoid messing up junctions on major roads at the expense of maybe not having a route over unclassified roads. Ticker
Well, if these areas are legitimate OSM objects, that's one thing. But if they're some mapper's idea of a way to customize the map for his or her (or a company's) particular use then I think they should be removed. Especially if they're causing routing problems. What are they? What purpose do they serve? If those questions cannot be answered then I say delete them. Dave On Sun, Sep 13, 2020 at 6:21 PM Ticker Berkin <rwb-mkgmap@jagit.co.uk> wrote:
I've found quite a few proper roads mapped as closed ways with [highway=unclassified, area=yes], but in the cases I've looked at so far, there has also been a correct unclosed way to represent the road.
I can't think of any method using style rules to detect the case when there isn't this additional road, but my preference is to ignore these areas and so avoid messing up junctions on major roads at the expense of maybe not having a route over unclassified roads.
Ticker
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
-- Dave Swarthout Homer, Alaska Chiang Mai, Thailand Travel Blog at http://dswarthout.blogspot.com
participants (8)
-
ael -
Andy Townsend -
Bernd Weigelt -
Dave Swarthout -
DD8KQ -
Gerd Petermann -
Mike Baggaley -
Ticker Berkin