Re: [OSRM-talk] Foot profile
On Tue, Apr 29, 2014 at 8:36 PM, Sarah Hoffmann wrote: > > We have adapted foot profiles on our Swiss installation[1], which seem > to work fairly ok. Code is here: > https://github.com/sosm/cbf-routing-profiles Thanks!! I'll give it a try ;-) Regards, EB ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Foot profile
On Tue, Apr 29, 2014 at 07:18:34PM +0200, Emmanuel Bégué wrote: > Thanks for a prompt reply, but how would poor data quality explain the > fact that two points that are very very near one another result in > such a different outcome? > > Where can I find more information in how to write profiles? > > And, in your experience, would a "car" profile that would basically > accept to take any road, and ignore road directions, be an acceptable > approximation for a foot profile? We have adapted foot profiles on our Swiss installation[1], which seem to work fairly ok. Code is here: https://github.com/sosm/cbf-routing-profiles (foot-city.lua for 'city walking', i.e. shorter route, foot-hiking.lua for 'hiking', i.e. less asphalt, quieter roads) Main gotcha: the profile misuses speed to get appropriate route preferences. The frontend therefore ignores the times it gets from OSRM and simply computes its own using the distance and a fixed speed. Cheers Sarah [1] http://routing.osm.ch > On Tue, Apr 29, 2014 at 6:47 PM, Dennis Luxen wrote: > > Salut Emmanuel, > > > > the foot profile is the least maintained. And foot data is among the most > > inconsistent tagged data in OSM. You routing data probably broke into many, > > many unconnected pieces. > > > > —Dennis > > > > > > Am 29.04.2014 um 18:29 schrieb Emmanuel Bégué : > > > >> Hello, > >> > >> Trying to use Project-OSRM for directions by foot, it seems some > >> points simply don't work, either as start or stop points, whereas > >> points that are very near, work fine (as well as some points that > >> shouldn't be reachable because for example they're in the water). > >> > >> For example the point 48.88368971897955,2.332395315170288 (north of > >> Paris), used as a start or an end point, always results in 207, > >> "Cannot find route between points". > >> > >> But if we use instead 48.88371088449246,2.332277297973633 (a few > >> meters away) then everything's fine; or if we use the offending point > >> with a car profile on Project-OSRM demo site: no problem. > >> > >> No problem either if we begin or end our journey in the middle of a > >> river: 48.85939286077621,2.331901788711548, so it's clearly not the > >> case that the destination point is somehow "unreachable" by foot. > >> > >> I have tried to set the offending point to the nearest node with > >> "locate" but that didn't help: > >>locate?48.88368971897955,2.332395315170288 => 48.883674,2.332385 > >> -- but that last point doesn't work any better. > >> > >> How can I investigate this? (How do we ask Project-OSRM to print more > >> elaborate error messages?) > >> > >> I'm using Project-OSRM version before 3.9, the stock "foot.lua" > >> profile and OSM data for France from Geofabrik. > >> > >> Thanks for any pointer. > >> > >> Regards, > >> EB > >> > >> ___ > >> OSRM-talk mailing list > >> OSRM-talk@openstreetmap.org > >> https://lists.openstreetmap.org/listinfo/osrm-talk > > > > > > ___ > > OSRM-talk mailing list > > OSRM-talk@openstreetmap.org > > https://lists.openstreetmap.org/listinfo/osrm-talk > > ___ > OSRM-talk mailing list > OSRM-talk@openstreetmap.org > https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Foot profile
Mitch, yes, it is quite easy to generate a shape file that contains the geometries of all ‚small‘ components that are badly connected. You would need to boot strap with tools enabled cmake .. -DWITH_TOOLS=1 It requires libgdal to build. Then you run make to build the thing and subsequently the osrm-components tool generates the shape file: ./osrm-components file.osrm file.osrm.restrictiins. It will dump a components.shp file that highlights the errors. Note that we do this already for the road network in car routing. To get an idea, go to the demo site at http://osrm.at, zoom to z14+ and select the ‚small components‘ layer in the layer switcher. It will display bad pieces of road in purple. —Dennis Am 29.04.2014 um 19:24 schrieb Mitchell Oliver : > Is there a way we could use the foot profile to generate a set of broken > routes we can feed into MapRoulette? > > Mitch Oliver > SENIOR DEVELOPER > > +1 (513) 319 3230 > mi...@roadtrippers.com > > Website Press Kit Twitter Facebook > > On Apr 29, 2014, at 1:18 PM, Emmanuel Bégué wrote: > >> Thanks for a prompt reply, but how would poor data quality explain the >> fact that two points that are very very near one another result in >> such a different outcome? >> >> Where can I find more information in how to write profiles? >> >> And, in your experience, would a "car" profile that would basically >> accept to take any road, and ignore road directions, be an acceptable >> approximation for a foot profile? >> >> Thanks, >> Regards, >> EB >> >> >> >> On Tue, Apr 29, 2014 at 6:47 PM, Dennis Luxen wrote: >>> Salut Emmanuel, >>> >>> the foot profile is the least maintained. And foot data is among the most >>> inconsistent tagged data in OSM. You routing data probably broke into many, >>> many unconnected pieces. >>> >>> —Dennis >>> >>> >>> Am 29.04.2014 um 18:29 schrieb Emmanuel Bégué : >>> Hello, Trying to use Project-OSRM for directions by foot, it seems some points simply don't work, either as start or stop points, whereas points that are very near, work fine (as well as some points that shouldn't be reachable because for example they're in the water). For example the point 48.88368971897955,2.332395315170288 (north of Paris), used as a start or an end point, always results in 207, "Cannot find route between points". But if we use instead 48.88371088449246,2.332277297973633 (a few meters away) then everything's fine; or if we use the offending point with a car profile on Project-OSRM demo site: no problem. No problem either if we begin or end our journey in the middle of a river: 48.85939286077621,2.331901788711548, so it's clearly not the case that the destination point is somehow "unreachable" by foot. I have tried to set the offending point to the nearest node with "locate" but that didn't help: locate?48.88368971897955,2.332395315170288 => 48.883674,2.332385 -- but that last point doesn't work any better. How can I investigate this? (How do we ask Project-OSRM to print more elaborate error messages?) I'm using Project-OSRM version before 3.9, the stock "foot.lua" profile and OSM data for France from Geofabrik. Thanks for any pointer. Regards, EB ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk >>> >>> >>> ___ >>> OSRM-talk mailing list >>> OSRM-talk@openstreetmap.org >>> https://lists.openstreetmap.org/listinfo/osrm-talk >> >> ___ >> OSRM-talk mailing list >> OSRM-talk@openstreetmap.org >> https://lists.openstreetmap.org/listinfo/osrm-talk > > ___ > OSRM-talk mailing list > OSRM-talk@openstreetmap.org > https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Foot profile
Is there a way we could use the foot profile to generate a set of broken routes we can feed into MapRoulette? Mitch Oliver SENIOR DEVELOPER +1 (513) 319 3230 mi...@roadtrippers.com Website Press Kit Twitter Facebook On Apr 29, 2014, at 1:18 PM, Emmanuel Bégué wrote: > Thanks for a prompt reply, but how would poor data quality explain the > fact that two points that are very very near one another result in > such a different outcome? > > Where can I find more information in how to write profiles? > > And, in your experience, would a "car" profile that would basically > accept to take any road, and ignore road directions, be an acceptable > approximation for a foot profile? > > Thanks, > Regards, > EB > > > > On Tue, Apr 29, 2014 at 6:47 PM, Dennis Luxen wrote: >> Salut Emmanuel, >> >> the foot profile is the least maintained. And foot data is among the most >> inconsistent tagged data in OSM. You routing data probably broke into many, >> many unconnected pieces. >> >> —Dennis >> >> >> Am 29.04.2014 um 18:29 schrieb Emmanuel Bégué : >> >>> Hello, >>> >>> Trying to use Project-OSRM for directions by foot, it seems some >>> points simply don't work, either as start or stop points, whereas >>> points that are very near, work fine (as well as some points that >>> shouldn't be reachable because for example they're in the water). >>> >>> For example the point 48.88368971897955,2.332395315170288 (north of >>> Paris), used as a start or an end point, always results in 207, >>> "Cannot find route between points". >>> >>> But if we use instead 48.88371088449246,2.332277297973633 (a few >>> meters away) then everything's fine; or if we use the offending point >>> with a car profile on Project-OSRM demo site: no problem. >>> >>> No problem either if we begin or end our journey in the middle of a >>> river: 48.85939286077621,2.331901788711548, so it's clearly not the >>> case that the destination point is somehow "unreachable" by foot. >>> >>> I have tried to set the offending point to the nearest node with >>> "locate" but that didn't help: >>> locate?48.88368971897955,2.332395315170288 => 48.883674,2.332385 >>> -- but that last point doesn't work any better. >>> >>> How can I investigate this? (How do we ask Project-OSRM to print more >>> elaborate error messages?) >>> >>> I'm using Project-OSRM version before 3.9, the stock "foot.lua" >>> profile and OSM data for France from Geofabrik. >>> >>> Thanks for any pointer. >>> >>> Regards, >>> EB >>> >>> ___ >>> OSRM-talk mailing list >>> OSRM-talk@openstreetmap.org >>> https://lists.openstreetmap.org/listinfo/osrm-talk >> >> >> ___ >> OSRM-talk mailing list >> OSRM-talk@openstreetmap.org >> https://lists.openstreetmap.org/listinfo/osrm-talk > > ___ > OSRM-talk mailing list > OSRM-talk@openstreetmap.org > https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Foot profile
may be the road network is disconnected, so that the algorthim find a node on a edge that as no connection to the whole network. This is a problem, may be a depth-search-first would help to detected this kind of errors. As far as I know OSRM has no plugin to check the network or? Regards Carsten Gesendet: Dienstag, 29. April 2014 um 12:18 Uhr Von: "Emmanuel Bégué" An: "Mailing list to discuss Project OSRM" Betreff: Re: [OSRM-talk] Foot profile Thanks for a prompt reply, but how would poor data quality explain the fact that two points that are very very near one another result in such a different outcome? Where can I find more information in how to write profiles? And, in your experience, would a "car" profile that would basically accept to take any road, and ignore road directions, be an acceptable approximation for a foot profile? Thanks, Regards, EB On Tue, Apr 29, 2014 at 6:47 PM, Dennis Luxen wrote: > Salut Emmanuel, > > the foot profile is the least maintained. And foot data is among the most inconsistent tagged data in OSM. You routing data probably broke into many, many unconnected pieces. > > —Dennis > > > Am 29.04.2014 um 18:29 schrieb Emmanuel Bégué : > >> Hello, >> >> Trying to use Project-OSRM for directions by foot, it seems some >> points simply don't work, either as start or stop points, whereas >> points that are very near, work fine (as well as some points that >> shouldn't be reachable because for example they're in the water). >> >> For example the point 48.88368971897955,2.332395315170288 (north of >> Paris), used as a start or an end point, always results in 207, >> "Cannot find route between points". >> >> But if we use instead 48.88371088449246,2.332277297973633 (a few >> meters away) then everything's fine; or if we use the offending point >> with a car profile on Project-OSRM demo site: no problem. >> >> No problem either if we begin or end our journey in the middle of a >> river: 48.85939286077621,2.331901788711548, so it's clearly not the >> case that the destination point is somehow "unreachable" by foot. >> >> I have tried to set the offending point to the nearest node with >> "locate" but that didn't help: >> locate?48.88368971897955,2.332395315170288 => 48.883674,2.332385 >> -- but that last point doesn't work any better. >> >> How can I investigate this? (How do we ask Project-OSRM to print more >> elaborate error messages?) >> >> I'm using Project-OSRM version before 3.9, the stock "foot.lua" >> profile and OSM data for France from Geofabrik. >> >> Thanks for any pointer. >> >> Regards, >> EB >> >> ___ >> OSRM-talk mailing list >> OSRM-talk@openstreetmap.org >> https://lists.openstreetmap.org/listinfo/osrm-talk > > > ___ > OSRM-talk mailing list > OSRM-talk@openstreetmap.org > https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Foot profile
Thanks for a prompt reply, but how would poor data quality explain the fact that two points that are very very near one another result in such a different outcome? Where can I find more information in how to write profiles? And, in your experience, would a "car" profile that would basically accept to take any road, and ignore road directions, be an acceptable approximation for a foot profile? Thanks, Regards, EB On Tue, Apr 29, 2014 at 6:47 PM, Dennis Luxen wrote: > Salut Emmanuel, > > the foot profile is the least maintained. And foot data is among the most > inconsistent tagged data in OSM. You routing data probably broke into many, > many unconnected pieces. > > —Dennis > > > Am 29.04.2014 um 18:29 schrieb Emmanuel Bégué : > >> Hello, >> >> Trying to use Project-OSRM for directions by foot, it seems some >> points simply don't work, either as start or stop points, whereas >> points that are very near, work fine (as well as some points that >> shouldn't be reachable because for example they're in the water). >> >> For example the point 48.88368971897955,2.332395315170288 (north of >> Paris), used as a start or an end point, always results in 207, >> "Cannot find route between points". >> >> But if we use instead 48.88371088449246,2.332277297973633 (a few >> meters away) then everything's fine; or if we use the offending point >> with a car profile on Project-OSRM demo site: no problem. >> >> No problem either if we begin or end our journey in the middle of a >> river: 48.85939286077621,2.331901788711548, so it's clearly not the >> case that the destination point is somehow "unreachable" by foot. >> >> I have tried to set the offending point to the nearest node with >> "locate" but that didn't help: >>locate?48.88368971897955,2.332395315170288 => 48.883674,2.332385 >> -- but that last point doesn't work any better. >> >> How can I investigate this? (How do we ask Project-OSRM to print more >> elaborate error messages?) >> >> I'm using Project-OSRM version before 3.9, the stock "foot.lua" >> profile and OSM data for France from Geofabrik. >> >> Thanks for any pointer. >> >> Regards, >> EB >> >> ___ >> OSRM-talk mailing list >> OSRM-talk@openstreetmap.org >> https://lists.openstreetmap.org/listinfo/osrm-talk > > > ___ > OSRM-talk mailing list > OSRM-talk@openstreetmap.org > https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Foot profile
Salut Emmanuel, the foot profile is the least maintained. And foot data is among the most inconsistent tagged data in OSM. You routing data probably broke into many, many unconnected pieces. —Dennis Am 29.04.2014 um 18:29 schrieb Emmanuel Bégué : > Hello, > > Trying to use Project-OSRM for directions by foot, it seems some > points simply don't work, either as start or stop points, whereas > points that are very near, work fine (as well as some points that > shouldn't be reachable because for example they're in the water). > > For example the point 48.88368971897955,2.332395315170288 (north of > Paris), used as a start or an end point, always results in 207, > "Cannot find route between points". > > But if we use instead 48.88371088449246,2.332277297973633 (a few > meters away) then everything's fine; or if we use the offending point > with a car profile on Project-OSRM demo site: no problem. > > No problem either if we begin or end our journey in the middle of a > river: 48.85939286077621,2.331901788711548, so it's clearly not the > case that the destination point is somehow "unreachable" by foot. > > I have tried to set the offending point to the nearest node with > "locate" but that didn't help: >locate?48.88368971897955,2.332395315170288 => 48.883674,2.332385 > -- but that last point doesn't work any better. > > How can I investigate this? (How do we ask Project-OSRM to print more > elaborate error messages?) > > I'm using Project-OSRM version before 3.9, the stock "foot.lua" > profile and OSM data for France from Geofabrik. > > Thanks for any pointer. > > Regards, > EB > > ___ > OSRM-talk mailing list > OSRM-talk@openstreetmap.org > https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
[OSRM-talk] Foot profile
Hello, Trying to use Project-OSRM for directions by foot, it seems some points simply don't work, either as start or stop points, whereas points that are very near, work fine (as well as some points that shouldn't be reachable because for example they're in the water). For example the point 48.88368971897955,2.332395315170288 (north of Paris), used as a start or an end point, always results in 207, "Cannot find route between points". But if we use instead 48.88371088449246,2.332277297973633 (a few meters away) then everything's fine; or if we use the offending point with a car profile on Project-OSRM demo site: no problem. No problem either if we begin or end our journey in the middle of a river: 48.85939286077621,2.331901788711548, so it's clearly not the case that the destination point is somehow "unreachable" by foot. I have tried to set the offending point to the nearest node with "locate" but that didn't help: locate?48.88368971897955,2.332395315170288 => 48.883674,2.332385 -- but that last point doesn't work any better. How can I investigate this? (How do we ask Project-OSRM to print more elaborate error messages?) I'm using Project-OSRM version before 3.9, the stock "foot.lua" profile and OSM data for France from Geofabrik. Thanks for any pointer. Regards, EB ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] OSRM v0.3.9 release announcement
> > Is that extract, prepare or routed memory improvement? Mostly osrm-routed. The demo site uses approx. 25 GB of RAM now for car routing on the planet. osrm-extract is about the same and osrm-prepare still benefits some. —Dennis ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] OSRM v0.3.9 release announcement
On Tue, Apr 29, 2014 at 05:06:30PM +0200, Dennis Luxen wrote: > Dear all, > > I am happy to announce the release of OSRM v0.3.9 [1], your favorite > OpenStreetMap based routing engine. This is again a pretty big release with > more than 170 commits (adding about 1500 LOCs). And it is the first release > in a series to bring major improvements in memory usage. Depending on the > size and location of the extract, you can expect 25--45% improvements. Is that extract, prepare or routed memory improvement? Flo -- Florian Lohoff f...@zz.de signature.asc Description: Digital signature ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
[OSRM-talk] OSRM v0.3.9 release announcement
Dear all, I am happy to announce the release of OSRM v0.3.9 [1], your favorite OpenStreetMap based routing engine. This is again a pretty big release with more than 170 commits (adding about 1500 LOCs). And it is the first release in a series to bring major improvements in memory usage. Depending on the size and location of the extract, you can expect 25--45% improvements. This is the shortened change log: - compress all road geometries (49% on the planet for cars) - make all error messages JSON - activated link time optimizations by default - more test scenarios (25+) - use forward decl's to cut compile times - ongoing effort to make code more legible - several lints fixed that were detected by facebooks flint - several lints fixed that were detected by Coveritys static analyzer - several lints fixed that were detected by OCLint - remove remnants of C-Style - add clang-format style file As always, we are dog-fooding our releases and the code of v0.3.9 is already running on our demo site [2] for a couple of days now. Big thanks to everyone who contributed to this important milestone release! --Dennis (on behalf of Team OSRM) PS: Please note that this is going to be the last release using C++03. We are finally switching to C++11 with the next v0.4.0 release, which may require newer compilers (clang 3.3+ and gcc 4.7+) on user machines. The bleeding edge develop branch is already going to switch in a few days. [1] https://github.com/DennisOSRM/Project-OSRM/releases/tag/v0.3.9 [2] http://map.project-osrm.org ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
