Recent diary entries
Hi OSM contrubutors, here you can see the result in the software (because on the classic map it isn't visible with classic layer)
We need to use the key lanes = x (x is the number of way in the both direction) in this case: lanes=4. We indicate the number of lane in the both directions: lanes:backward=2 lanes:forward=2 We need to use the lane on the left road if we want to turn left placement:forward=left_of:1 This is the mark sign indicator on the road turn:lanes:forward=left;through|through;right
Of course all can improve it again (my work, my contribution)
P.S.: Don't forget, it's for to do a "map" and if we want somebody use it, it's must be a little bit "pleasant"...). But in the reality, it isn’t a simple map...
This is an example of traffic calming = island, we need only to draw one way and not two with several tag one it.
On this picture (the link below), you can see the lanes (numbers of lanes forward/backward and turn lanes) the result in the software and the mark on the real road (it isn't a trunk and motorway), so in the middle it' isn't a "guard_rail" so we can considered that like a "traffic_calming = island" (the road in the bottom is a bus way and we can't turn on it probably we can use a relation with "no turn" on it)
Before changeset: http://imgur.com/d8bPCdY
After changeset http://imgur.com/aM64YF4
I have uploaded the entirety of the file processed_processed_fix_v7.osm as described on the wiki page https://wiki.openstreetmap.org/wiki/Import/South_Australian_Waterbodies
I will now begin some minor manual editing to deal with locations where recent user edits overlap with the imported data.
Yesterday marked three years from my first OSM changeset. Not long after starting mapping I set myself the objective of adding enough data to be able to make a good walking map of the Southern part of the Peak District National Park in three years. The main objectives were to survey and add all the remaining public rights of way (PROW) (footpaths & bridleways) and also to add all the field boundaries. When I started there was reasonable PROW coverage in some areas and virtually no field boundaries.
To put this objective into perspective there are, according to the National Park figures, 3510 PROWs in the park covering 1867 miless and 5440 mile of dry stone walls within the park. There are also good number of fences and hedges so the lenght of field boundaries to be mapped is probably half as much again.
I’m afraid to say I have failed to achieve the objectives in the three years I set myself.
The above gives a representation of what has been mapped to date. I was hoping to have achieved single figures for the PROWs but it stands at around 20. In retrospect it was a mistake to have a target for achieving the objective in the middle of the UK winter. The weather and short daylight hours have certainly been against me. It is also proving quite time consuming to pick off the last few footpaths. Parking is often problematic so you end up walking several times the length of the footpath just to survey it and get the gps trace. Survey is also not that efficient when trying to cover new footpaths. I've walked almost 1500 miles in the last 3 years but probably only mapped a quarter of this in new footpaths.
Field boundary coverage it quite good along all the footpaths that have been mapped. There are still a few empty patches. Some of this is moorland without field boundaries. Other areas don’t have any footpaths so they have been the last on the list to map.
After initial experimentation with note taking, voice mapping I became an avid photo mapper. I use to take a few composed pictures but this was time consuming and you could almost guarantee you hadn’t taken a picture of something when it came to editing. I now don’t stop to take pictures but just shoot several picture is an arc at regular intervals. I seem to have amassed 378,099 pictures to date! They have certainly made editing in JOSM much easier and hopefully more accurate and detailed. Editing has generally taken up more time than surveying.
With luck I will complete the above objectives for the Southern Peak District National Park in the new few months and is should be possible to produce, with the addition of contours, an alternative walking map for this area. It is then on to the Northern Peak District. Although it covers a larger area, large parts are moorland so the actual number of field boundaires is probably less. I was hoping to do this in another two years but logistically it is more difficult and surveying will have to be limited to the summer months.
The Peak District is a good place to go walking so hopefully this work will prove useful.
I have commenced the upload process for the import described on this wiki page:
Wakacyjna plaża zapełniona turystami, w tym pięknymi kobietami w seksownych strojach kąpielowych to najlepsze, co co mnie może spotkać w tym właśnie okresie.
Edinburgh's map is looking very full. I started doing building editing when I moved here, but now there's not much to add, so I've started to add some building:levels tags in areas I walk through regularly.
An academic project, Mapping Edinburgh's Social History has been doing a lot of work enhancing Edinburgh map, in particular adding addresses that will allow geocoding without postcodes, which only take you back as far as 1971. One of the MESH mappers, eric_, must be hawkishly watching the map, saw a recent batch of my edits and suggested that I start adding in roof:levels and building:material "but no pressure!"
While i'm into doing this if it has value to others and i'm surveying anyway, I have my doubts about building interpretation in a city full of architectural idiosyncracies such as Edinburgh. One is, what happens when the top storey of a building is embedded within a roof, like this?
You can see from the side of the building that the roof storey is not an add-on or an afterthought, that the side facade is genuinely five storeys high, so i'm really not sure whether to model this as building:levels=5 or building:levels=4 and roof:levels=1
The advantage of using the levels is that they're easy to observe and record without any special surveying equipment (and yes i have thought about trying to use ultrasound with an arduino to take measurements of building heights in metres, but that wouldn't give that much more accuracy value than building:levels alone.
Then of course in a mixed-style European city we often get scenarios like this, where an older high-ceilinged building is built next to a modern, low-budget low-ceilinged building: this shows adjacent buildings as part of the Summerhall complex:
Any thoughts from other local mappers would be appreciated. What i want to get out of this personally are some simple, impressionistic "haptic maps" using a 3D printer or maybe plaster cast into a mould. I'm not personally worried about precision, more curious to see what drops out of the existing tools, so building:levels are generally enough for me (and well supported by amazing tools like osm2world and osmbuildings.org.
The new routing feature on openstreetmap.org will make it a lot easier to find certain types of errors in the map. I tried it out and found errors that I would have overlooked otherwise.
Autoroute 40 is a limited-access divided highway in Quebec. Crossing gaps in the median strip are used for circulation during roadworks. These gaps are currently tagged as service roads on OSM.
What route should you take if you miss your highway exit and you want to go back in the opposite direction? If you are a boring law-abiding citizen, you will have to keep driving until you can take the next exit and then re-enter the highway in the other direction. OSM's routing algorithm, however, relying on less-than-perfect road access data, worries not about following such pointless traditions. Its free mind comes up with a more daring, if illegal, route that involves a hard 180-degree left turn through a crossing gap:
Clever, but a little too dangerous for my taste. Adding access=no to those service roads turns OSM back into a less audacious, but safer route planner.
natural=wood and landuse=forest completely camouflage streets (especially highway=trunk) in low zoom levels. Do you see the three rings of highway=trunk around Moscow in the attached screenshot? I don't.
What do you think? Please leave your comments here:
Correcting and adding roads where I traveld on the way from the Netherlands to South Africa have been completed.
The local Missing Maps organisers are keeping up the pace with the mapathons happening in Edinburgh every couple of weeks at the moment.
This time the word got out to the OSM community with plenty of notice, and the mapathon was duly mobbed by the Usual Suspects. I enjoy the pub meets as much as some, so it was a good chance to catch up with chrisfl, drnoble, fozy81 and eisa, while stevefaeembra was busily OCD mapping up a storm again, and we even flushed out Bob Kerr from his OSM semi-retirement, full of curiosity and helpfulness as ever.
This was my first attempt to ever use the HOT tasking manager in anger and I remain unconvinced about the validity of the task. The interpretation of aerial imagery remains delicate and uncertain, and the currency of date in the imagery is unknowable. How useful this is to displaced communities seeking shelter, who can tell? Are tracks to farm buildings really missing on the map? Is tagging for the renderer appropriate at the level of detail where a road may be a service road, residential, or a track, according to the iD pre-sets?
Meanwhile, many of the tasks marked as open seem almost complete; as a casual mapper, not familiar with the HOT standards of quality, i would have a hard time marking most squares as "Done" and it appears others have similar existential problems. One wants to create useful work, not busywork, for those validating the maps on the ground. As I'm mildly stricken with a cold and indulging in a "duvet day" today, I may keep going with some of the random tasks on the grid.
I'm partially converted to the iD editor, though the usual gripes came up in the general discussion - it remains far too easy for a new user to start deleting objects without an explicit check or warning or an obvious means of rowing back. We hear that RichardF has a relevant patch in, and deserves cake, let's hope it gets accepted and released soon.
Thanks again to the redoubtable Margaux Meslé for so much organising and communications work, and for really making extra effort this time to reach out to the local OSM mapper community, it can only be of benefit to all, and i look forward to future mapathons.
I've been updating OSM for a few years now based on GPS tracks from my hikes in forested parks and adjacent public greenspace in the U.S. State of Georgia. Those updates and a lot more trails created by others in OSM are now incorporated into My Trails for Garmin (free download for PC and Mac).
It is a transparent map, so it sits on top of whatever topo you use on your GPS; my favorite free one is Georgia Topo:
Just over 600 miles of trails added. I'm being conservative with this first update by only adding trails in the 64 parks and adjacent public greenspace areas I've edited in OSM. Of those, I consider 17 areas to be complete (and together, represent over half the mileage of the update). They are: * Amicalola Falls State Park * Arabia Mountain Nature Preserve * Charlie Elliott Wildlife Center * Fort Yargo State Park * Harbins Park * Hard Labor Creek State Park * High Falls State Park * Kennesaw Mountain National Battlefield Park * Little Mulberry Park * Miners Creek Park * Mistletoe State Park * Piedmont Park * Red Top Mountain State Park * Skidaway Island State Park * South Rockdale Community Park * Stone Mountain Park * Sweetwater Creek State Park
Periodic updates will be made.
A quick and simple way to visualize the usage of data derived from Mapillary street view photos is to use overpass turbo.
Check out http://overpass-turbo.eu/s/7JX for more details.
After almost 3 months of work, development reached version 1.0! Also I have changed its name from Taglocator to OpenPoiMap which is maybe more descriptive for the non-mapper.
And it is now available under its own domain name: http://openpoimap.org
Compared to the first version many improvements have taken place (thanks to the community) and it is fully described in a wiki.
The tool is handy for mappers, to check their own work and see what is missing from the map in the areas they know, but it is at least as useful for the layman mapuser who just wants to know the nearest Busstop, Bookstore, Library, Restaurant, Pub, Cinema, Hotel, Giftstore, Optician, Monumental Tree, Picnic, Windmill, Museum, Viewpoint, Parking, Bicycle rental. . . (not necessarily in that order!)
But if you want to check which Bar has no Telephone number, that's also possible!
Many improvements are yet to come, and development continues. Give it a try!
This morning I was a guest in Durham University's Geography department and I gave a workshop to students. It was great to see people become interested and I hope it will inspire them to continue making improvements to OpenStreetMap in their free time. I've written a entry about this on my main blog http://www.livingwithdragons.com/2015/02/learning-to-map
Do send a message to my account through the OpenStreetMap website if you have questions.
As of now, OSM now has routing on their website as standard: you can plan a simple journey from point A to point B, by car, foot or bicycle. GraphHopper, MapQuest and OSRM are the current routing information suppliers for this module.
I've already made a route to show you how it all works. I'm not sure when the "via" feature will come in.
P.S. at time of posting, Amaroussi's current exhaustion level is OVER 9,000 !!!
I participated in the first Missing Maps event in the Netherlands on 2015-02-14 and worked on an area in the C.A.R. [Here[(http://www.openstreetmap.org/user/Polyglot/diary/34394) are more details.
After participating in the Missing Maps in Antwerpen, I went to the one in The Netherlands yesterday. Big thanks to Philip for making this possible for me!
The people from MSF/AZG/DWB/Red Cross had 2 Tasks in mind for us.
A though one in Africa where we had to add natural wood and rivers, as those are the (breeding) habitat for tse-tse flies, which cause sleeping disease. The beginners using iD had to abandon that one, as it's impractical to zoom in far enough to work and still see the bigger picture. With JOSM one can draw a general contour of the wood at low zoom, then refine it by using the Improve Way Accuracy mode zoomed in a lot further. It's still a daunting task to get it right though. What can be considered wood/forest? How does one recognise wetlands?
The other task was in Haiti. It was not easy either. A densely populated area and MSF wants to know how many buildings there are. (It's explained better in the description...)
It peaked my interest for 2 reasons:
- They used a drone to create superior aerial imagery
- The Red Cross drove around with a smartphone used as a dashcam. The pictures were uploaded to Mapillary. Press that little play button and enjoy the ride.
This allows to go and have a look around and see what's actually there. The width and state of the road, the state of completion of the buildings and so on.
Coverage is more limited than what can be seen from above, but it's possible to read the name of a school and then it's clear as well that it is a school.
The great thing about Mapillary is that the barrier of entry to contribute yourself is a lot lower than say, fly a drone to create excellent imagery...
All you need is a smartphone with a descent camera and GPS. Bonus points if it has a electronic compass, but that information can also be gotten from the GPX track. (electronic bread crumbs trail of where you've been).
I've been doing this for Geofort, somewhat surprised that I was the first one to do it. We've also covered a good part of the way over to Geofort from Belgium. Interrupted between Brasschaat and Breda because of a drizzle.
Only It's so Funny did the same. I had expected to see all the roads to the geofort from all directions. Now I feel bad for not having mentioned it on the Meetup group beforehand. Mapillary still seems less known than I had expected. Hence this diary entry.
Let's give Google Streetview some competition and do them one better! There are so many places their survey cars can't reach! And hopefully one day, they decide to be even nicer than they claim to be and give us permission to look at their Streetview to improve Openstreetmap. One can always be optimistic :-)
My first set of edits and additions are conducted as part of the GIS class I am currently taking towards my BSc in Geography at Northern Illinois University. Some initial edits were done around the university campus in DeKalb IL as part of a class initiative to comprehensively digitize the area. As part of the group effort I have digitized sidewalks, crosswalks and traffic calming features. I followed this up by examining the level of detail in the area where I currently live, La Grange IL, and have added sidewalks, parking lots, building polygons and restaurants in the town, though much remains to be done. Finally, I started to digitize physical features (lakes, streams, forest and roads) in the area around the village in which I grew up in Mayo, Ireland.
I am new to this world and realized that I can contribute to OpenStreetMap-project, so excited. I will finally get all details on the maps from my region up to date as I can directly contribute. Awesome!