Hi Jason, On Mon, 2016-03-21 at 20:42 -0400, Jason Remillard wrote: > Hi Roman, > > The city of Boston building data set for buildings has address. > > http://bostonopendata.boston.opendata.arcgis.com/datasets/492746f09dd > e475285b01ae7fc95950e_1 Interesting, so that's where the tax parcel basemap buildings are coming from...
> It seems like they have already figured out what address goes on what > building. Should this data set be used rather than the parcel data > set? We are not using the parcel data set, instead, we are using a service build specifically to point to street addresses - http://bostonopendata .boston.opendata.arcgis.com/datasets/b6bffcace320448d96bb84eabb8a075f_0 (Live Street Address Management (SAM) Addresses), which is, well, live and appears to have changed every time I fetched the geojson feed. > The licensing link says the following, it is kind of weird. > > "The City of Boston recognizes the value and benefit gained by > sharing GIS data. ... The GIS data is updated > and modified on a regular basis and users are encouraged to report > any errors to the City." At some point I found a notice that since the data was paid for by the taxpayers, it is freely available to the public. I hope I'll meet someone from MassGIS this week to clarify the statement above. > I suggest that you move forward with the building splitting, since it > is a manual process, it can proceed like a normal mapping activity. Yup, that's what I am doing now. > While the buildings are being cut up, we can work on address import > and make sure it is good shape. Before I would be comfortable with > the import, I would like to see some sample OSM files we can load > into JOSM to look over. I have been regenerating the .osc files daily (new batch in an hour) when new data is available or the code is optimized. All the files are linked to from the wiki page - https://wiki.openstreetmap.org/wiki/Impo rt/Boston_Street_Address_Management_%28SAM%29_Import#OSM_Data_Files. I'd like to point out that only -unique.osc are considered for import, the -fixmes.osc are for further investigation and additional processing. As per the "Buildings" dataset you mentioned above, that does look like a good source, but I did a spot-check on various buildings I've surveyed: * Object ID: 91094 - no house number (should be 32-36 John A Andrew Street) * Object 90372 - 110 McBride (should be 108-110) - and generally building ranges are missing, so there are e.g. 5 "378 Riverway"s since they are located on the same parcel. The geometry information, however, looks very yummy. It's not directly usable since the various building parts are provided as separate objects probably linked to by the BUILDING_I. Now, if we at some point decide to replace the Boston geometry information, that would be a great dataset. But I feel this is far more involved than adding the house numbers for basic navigation needs which is what I am after. And even if we do, there are newer shapefiles based on Orthoimagery/LIDAR which has the newer building shapes, and geometry is completely different, buildings are not split and have no addresses. -- Sincerely, Roman
signature.asc
Description: This is a digitally signed message part
_______________________________________________ Talk-us mailing list Talk-us@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-us
