Recent diary entries
This weekend I have been out mapping a number of recently built bike trails in Toronto, Ontario using a GPS device. These were funded through a government infrastructure stimulus program. These were built along hydro corridors (high voltage power lines) and abandoned railways. There are also a few sections where a paved bike path runs along the side of a road, along Ellesmere Road and Morningside Avenue, with bike traffic signals at major intersections (like the existing bike path on Lake Shore Blvd near the Exhibition). Some sections of this trail system are still under construction, and are not finished yet.
At major road crossings bicycle traffic signals have been installed like the one shown in the Toronto Star article. These have not been turned on yet.
Just noticed this map of the new Bixi bike sharing stations in Toronto using OSM data. It has several other cities as well. It shows in real time the number of bikes parked at each station.
Today the City of Toronto has opened a one-block segment of Dufferin Street between Queen Street West and Peel Street that eliminates the "Dufferin Jog". Dufferin Street was formerly interrupted here by the railway line (which runs at a 45-degree angle to the road) and drivers had to take a time-consuming detour via Queen, Gladstone and Peel. The city has built a north-south underpass under the railway, which was a difficult task because the railway had to remain open at all times. (This railway will be used for the controversial Pearson Airport rail link which will open in 2015.)
It seems like another user beat me to it marking the underpass as open. Now I am simply tidying up the underpass and adding a few minor details. Of course, expect it to be many months before Google Maps is updated.
Google Maps has accidentally renamed Wellesley Street in Toronto "Lourdes Lane". This is blatantly wrong and if this were done in OSM would be corrected very fast.
Currently importing missing data from CanVec 030M11,12,13,14 in the Greater Toronto Area.
A quick question: why do certain tiles have housenumbers while others do not? Specifically, 030M14 is missing housenumbers while 030M11,12,13 have them.
Just stumbled on a bike routing application for Toronto using OSM data:
I noticed that as of a few days ago, the keepright validation tool is now available in Canada. This is a good tool for cleaning up after the GeoBase import, as ways from the GeoBase dataset need to be connected to the existing data, and this tool makes it much easier to find areas where this has not been done yet. Check it out at keepright.ipax.at.
Had an interesting adventure yesterday when we were lost in Gatineau Park in Quebec, near Ottawa. We were trying to drive to the Mackenzie King estate which is located within the park but the main road to it (Gatineau Parkway) was closed except to bikes and pedestrians on Sunday between 6am-11am; we were trying to get there just before 11am when the Mackenzie King estate opened. It turns out the only way to get there before 11am is to take various back roads. We got lost for a while (OSM was still incomplete there and the commercial maps were very unclear), but I eventually figured my way using some official park maps. We parked in parking lot P7 (to the northeast of Lac Kingsmere), which was clearly intended for hikers in the park, not visitors to the King Estate; the official maps were deliberately confusing as to how to get from it to the King Estate, because you were supposed to park at two pay parking lots (south of Lac Kingsmere), but they were only accessible from Gatineau Parkway which was closed before 11am.
Lesson learned: to get to the Mackenzie estate, take exit 12 off Autoroute 5, turn left on Chemin Old Chelsea, left again on Chemin Kingsmere, and continue to parking lot P7 (which is free). Then walk south on Chemin Swamp and west on Chemin Barnes until you get to the estate. Ignore all signs to the Mackenzie Estate because they direct you to the two pay parking lots.
Note to self: There is still quite a lot of mapping to be done here. In particular, GeoBase import is still needed.
I have been manually adding road names to previously traced data from the Statistics Canada 2009 road names dataset, which OpenStreetMap has got permission to use. I am currently adding road names to the wonderful 1970's-era "new town" of Bramalea, Ontario, Canada (part of Brampton now) where all the roads in a given area start with a given letter e.g. Addington, Albion, Alexandria, Aloma, Ascot, etc. See Wikipedia.
If you want to help, download the "road network" data from here. The license is compatible with OSM. Select the year (2009), the language (English), the format (several available, shapefile works best in OpenJump), and the area (Ontario). Then download a program called OpenJump, load the file, and start adding road names. Note that this data hasn't been imported directly into OpenStreetMap as the roads in that dataset don't quite line up with reality, but the road names are all correct and can be added with the assistance of Yahoo imagery, GPS traces or Geobase data.
Added Ed Mirvish Way, formerly Duncan Street, which was officially renamed today around noon by the city in honour of the late Ed Mirvish, whose company owns several theatres in the area. Unfortunately the city's GIS system seems to have beat me to it.
The weather was better today (+5 degrees C) so I went mapping again around Eglinton & Victoria Park today. I hadn't been doing much mapping in January because of the cold. Unfortunately reading week is coming soon which means essays, essays and more essays due right after so I won't be able to map as much as I like.
(Finally) mapped and uploaded a big new subdivision in Kingston, Canada. It is still under construction, so I will have to come back in a year or so to map new roads. This one is by far the largest. As usual, the mapping companies got it completely wrong. Navteq has a slightly outdated but otherwise correct map. TeleAtlas, on the other hand, has downright wrong data; its map appears to be based on an old plan of the subdivision. Not only does it show roads that have yet to be built, but it shows a quite different layout from the parts that have been built. It is utterly and completely wrong! (I reported it to them via the online form, but I fully expect that it will take them about 5-10 years to fix, by which time OSM will be fully complete.)
See http://geo.topf.org/comparison/index.html?mt0=googlemap&mt1=tah&lon=-76.5552449&lat=44.2630579&z=15 and feel free to laugh.
Purchased new GPS (Garmin eTrex Legend HCx) last Wednesday. The accuracy is quite good - usually about 5m after obtaining a good fix with WAAS on (it's off by default). Have been using it in conjunction with Yahoo photos to map parts of Toronto - it's much faster to use it plus a voice recorder than it was to use pencil and paper and it allows me to map areas that are difficult to make out on the sat photo. Haven't mapped any areas with no sat photos - yet - but undoubtably I will find some new subdivisions soon enough.
Mapped (almost) all the buildings in University of Toronto (where I'm going to school right now) and nearby Ryerson University, tracing their outlines using Yahoo imagery. I missed a few buildings because they were missing signs or were under construction. I added the addresses of most of the buildings (shown on the signs) so if someone gets a geocoder running then it can use them. They should show up on Mapnik tomorrow.
Sorry about that. The email is dave (ATSYMBOL) sr71 (DOT) net. See the dev mailing list for more info.
The TIGER import has restarted again. If you want data for New York City, just email Dave Hansen, the person who is doing the imports, at , and ask him to move New York up the queue.