Re: [Talk-us] suburban superblocks that nobody wants to survey
What about using county or state-wide parcel data for address points? Centroid of each real property lot. There's many problems with doing this for a whole state; NJ has many cases of one house sitting on multiple lots (old subdivisions of 25'x60', later built as 75'x60' or so without consolidation of the lots) but for newer subdivisions this would work fine. NJ data is public domain, don't know about other states and counties. I could dump our address points into OSM format if there's interest. John Reiser Campus GIS Specialist Rowan University http://gis.rowan.edu/ Date: Fri, 16 Mar 2012 21:34:35 +0700 From: Kate Chapman k...@maploser.com To: Rich ric...@nakts.net Cc: Nathan Edgars II nerou...@gmail.com, talk-us@openstreetmap.org Subject: Re: [Talk-us] suburban superblocks that nobody wants to survey Message-ID: cagn7morbrn76bltxqy9eytng9d3pzrokwb5mzavn8ro7-xj...@mail.gmail.com Content-Type: text/plain; charset=ISO-8859-1 What about actually doing the entry while you are in the car? My first job after college was to map every house in a county. (Wayne County, Indiana for those that are curious) The set-up was to use GPS with a computer and to map the addresses as you went. We were already starting with the county's master address list, but I don't see how this would be different. The only entire suburban neighborhood I mapped was my own. Can't say it was that much fun, but I'd just get another string of addresses everyday on the way to work and enter them that night. I didn't end up with that much data at any one time, but it did take a long time to complete. -Kate ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] suburban superblocks that nobody wants to survey
On 3/17/2012 10:20 AM, Reiser, John J. wrote: What about using county or state-wide parcel data for address points? Centroid of each real property lot. There's many problems with doing this for a whole state; NJ has many cases of one house sitting on multiple lots (old subdivisions of 25'x60', later built as 75'x60' or so without consolidation of the lots) but for newer subdivisions this would work fine. NJ data is public domain, don't know about other states and counties. Where I map, the address point data is indeed public domain (when it's available), but it takes ages for the public sites to be updated and nobody's terribly interested in fixing that given the budget situation they're dealing with. I did an import of address point data for four counties in Arkansas, but I couldn't really get consensus that it was what the community wanted so I didn't continue beyond that. -Nathan ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] suburban superblocks that nobody wants to survey
On 03/16/12 01:10, Nathan Edgars II wrote: On 3/15/2012 6:43 PM, Nathan Mills wrote: On 3/15/2012 4:59 PM, Nathan Edgars II wrote: lots of driving and all you get is street names, since everything else is single-family houses. And address points How does this work? Do you stop at every house and write down the address? my workflow for the most efficient data collection : 2 persons in a car. one takes photos and instructs the other on the route. later photos are correlated to the gpx based on their timestamps - and i have location-identified photos of housenumbers, shops, pubs, restaurants, water hydrants and lots and lots of other features. there is only one problem with this approach - i can collect lots more data than i can process, sleeping factored in ;) -- Rich ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] suburban superblocks that nobody wants to survey
What about actually doing the entry while you are in the car? My first job after college was to map every house in a county. (Wayne County, Indiana for those that are curious) The set-up was to use GPS with a computer and to map the addresses as you went. We were already starting with the county's master address list, but I don't see how this would be different. The only entire suburban neighborhood I mapped was my own. Can't say it was that much fun, but I'd just get another string of addresses everyday on the way to work and enter them that night. I didn't end up with that much data at any one time, but it did take a long time to complete. -Kate On Fri, Mar 16, 2012 at 3:10 PM, Rich ric...@nakts.net wrote: On 03/16/12 01:10, Nathan Edgars II wrote: On 3/15/2012 6:43 PM, Nathan Mills wrote: On 3/15/2012 4:59 PM, Nathan Edgars II wrote: lots of driving and all you get is street names, since everything else is single-family houses. And address points How does this work? Do you stop at every house and write down the address? my workflow for the most efficient data collection : 2 persons in a car. one takes photos and instructs the other on the route. later photos are correlated to the gpx based on their timestamps - and i have location-identified photos of housenumbers, shops, pubs, restaurants, water hydrants and lots and lots of other features. there is only one problem with this approach - i can collect lots more data than i can process, sleeping factored in ;) -- Rich ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
[Talk-us] suburban superblocks that nobody wants to survey
On 3/14/2012 21:18:57 -0400 Nathan Edgars II wrote: Depending on the state or local government, you may be able to verify names against an official dataset. Otherwise subdivision plats work for the endless suburban superblocks that nobody wants to survey. In the interest of figuring out how to attract more people to participate in OSM, I'd like to see some more discussion of this. Is it generally true that people who work on OSM don't like to map subdivisions? And, if so, why? Because these are home to so many people in the US, it raises a question about the viability of strategies that suggest people start in OSM by mapping their own neighborhoods. I admit that I prefer not to, but because of where I live and work, my activity space doesn't take me into them very often. I did start out in OSM by mapping my neighborhood surroundings. I have mapped some subdivisions (ways and land-uses but not individual houses) and don't find them boring or onerous (for onerous, splitting streets into dual carriageways is at the top of my list). Ed Hillsman ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] suburban superblocks that nobody wants to survey
On 3/15/2012 8:52 AM, Hillsman, Edward wrote: In the interest of figuring out how to attract more people to participate in OSM, I'd like to see some more discussion of this. Is it generally true that people who work on OSM don't like to map subdivisions? And, if so, why? Because these are home to so many people in the US, it raises a question about the viability of strategies that suggest people start in OSM by mapping their own neighborhoods. I don't know anything about this specifically. It's interesting that not a single person in those 120 subdivisions was interested in mapping their own subdivision. I have done some onsite surveys of smaller subdivisions (100-400 homes), and can set this up with a camera, video cam, and bike to collect quite a lot of information in a single visit, and the end result is streets with lanes, speed limits, one ways, and house numbers. In this area, since no one else is participating[1], it's just a practical matter to create the base new subdivision information from TIGER since the local governments don't freely give this information. The only followup surveys are quick to clarify obvious errors in the TIGER data. The subdivision plat idea is new to me, but I'm not sure where I'd find them. [1] It is notable that likely because of the Apple publicity spike, a single new mapper popped up and added streets in his neighborhood and did a quality job. If this was indeed because of the lure of the 'blank page', our license removal exercise will create many more blank pages to test this theory with. ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] suburban superblocks that nobody wants to survey
Hi, On Thu, Mar 15, 2012 at 7:52 AM, Hillsman, Edward hills...@cutr.usf.eduwrote: In the interest of figuring out how to attract more people to participate in OSM, I'd like to see some more discussion of this. Is it generally true that people who work on OSM don't like to map subdivisions? And, if so, why? Because these are home to so many people in the US, it raises a question about the viability of strategies that suggest people start in OSM by mapping their own neighborhoods. It would be straightforward enough to do some analysis on this, and I concur that it would be interesting, if only there was some geographical demarcation file of these types of neighborhoods. Would it be possible to define that? Apart from that, I can see why these areas seem to have low priority. I am going to make a lot of assumptions here, because that's all I can do without having really investigated. Subdivisions make for boring mapping. It is repetitive and there seems to be no end to them. OSM contributors invest their free time in the project and want to get something out of it: fun, a sense of accomplishment. So they spend their time on fun mapping projects, not boring ones. Considering that the US is horribly understaffed with contributors, subdivisions are much less likely to be tended to, unless one of those sparse mappers has some personal incentive. I see two ways to break this: 1) attract more mappers. 2) make boring projects fun. I see both tasks as core to growing OpenStreetMap in the US, and they are related. -- martijn van exel geospatial omnivore 1109 1st ave #2 salt lake city, ut 84103 801-550-5815 http://oegeo.wordpress.com ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] suburban superblocks that nobody wants to survey
On Thu, Mar 15, 2012 at 11:35 AM, Martijn van Exel m...@rtijn.org wrote: On Thu, Mar 15, 2012 at 7:52 AM, Hillsman, Edward hills...@cutr.usf.edu wrote: In the interest of figuring out how to attract more people to participate in OSM, I'd like to see some more discussion of this. Is it generally true that people who work on OSM don't like to map subdivisions? [ ... ] Apart from that, I can see why these areas seem to have low priority. [ ... ] Subdivisions make for boring mapping. [ ... ] I see two ways to break this: 1) attract more mappers. I don't accept that mapping a subdivision is any more-boring or less-rewarding than other forms of mapping. The caveat is as long as it is your subdivision. Sure, they might be row upon row of similar construction, but at least they are row upon of similar construction full of your friends and neighbours. That motivates the mapper to put it on the map and to get it right. That new, local mapper is most likely to update the map when one field is converted to a cricket pitch, and when the local convenience store has a name change a few years down the road. You can't beat that as a mapper from a distance. So, yes. Attract more mappers, in more places. ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] suburban superblocks that nobody wants to survey
At 2012-03-15 06:27, Mike N wrote: On 3/15/2012 8:52 AM, Hillsman, Edward wrote: In the interest of figuring out how to attract more people to participate in OSM, I'd like to see some more discussion of this. Is it generally true that people who work on OSM don't like to map subdivisions? And, if so, why? Because these are home to so many people in the US, it raises a question about the viability of strategies that suggest people start in OSM by mapping their own neighborhoods. I don't know anything about this specifically. It's interesting that not a single person in those 120 subdivisions was interested in mapping their own subdivision. Assuming we're talking about the US, not really surprising. I've mapped hundreds of subdivisions in southern Cal. In particular, dozens of them were not on the map at all, having been developed after TIGER's 2005 source date. Some were not even on Google Maps, so you'd think someone out there other than me would have wanted to map them. I have done some onsite surveys of smaller subdivisions (100-400 homes), and can set this up with a camera, video cam, and bike to collect quite a lot of information in a single visit, and the end result is streets with lanes, speed limits, one ways, and house numbers. Yup - me too, with a car, GPS, and a digital camera. In this area, since no one else is participating[1], it's just a practical matter to create the base new subdivision information from TIGER since the local governments don't freely give this information. The only followup surveys are quick to clarify obvious errors in the TIGER data. The subdivision plat idea is new to me, but I'm not sure where I'd find them. I've recently done this when I see an area that really is untouched. I first make sure that all the ways are the original TIGER ways (tiger:cfcc=* ((version:1 user:DaveHansenTiger) | (version:2 user:balrog-kun))), remove them, then convert and transplant in new TIGER2011 data, connecting it to existing ways at the borders. BTW, many (most IME) county governments have at least some data available for free. Assessor's maps are generally more available, though keep in mind that they are less authoritative on naming than tract/parcel maps because the assessor's role is more related to the land parcels than the streets between them. Tract/parcel maps, records of surveys, roadbooks, etc. are generally available from the planning and/or public works departments. While they are usually filed with the county recorder, that avenue is usually not free. All it takes is a little digging. If you run into a fee-required situation, don't be afraid to ask for a waiver, describing OSM and your need to use them as a reference. That's worked for me. -- Alan Mintz alan_mintz+...@earthlink.net ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] suburban superblocks that nobody wants to survey
On 3/15/2012 8:52 AM, Hillsman, Edward wrote: On 3/14/2012 21:18:57 -0400 Nathan Edgars II wrote: Depending on the state or local government, you may be able to verify names against an official dataset. Otherwise subdivision plats work for the endless suburban superblocks that nobody wants to survey. In the interest of figuring out how to attract more people to participate in OSM, I'd like to see some more discussion of this. Is it generally true that people who work on OSM don't like to map subdivisions? And, if so, why? Because these are home to so many people in the US, it raises a question about the viability of strategies that suggest people start in OSM by mapping their own neighborhoods. I was talking specifically about driving through these neighborhoods to get street names. Obviously if you live there it's different, or if you live nearby and exercise by bike. But driving through subdivisions far from home has a high cost-to-benefit ratio - lots of driving and all you get is street names, since everything else is single-family houses. ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] suburban superblocks that nobody wants to survey
On 3/15/2012 4:59 PM, Nathan Edgars II wrote: lots of driving and all you get is street names, since everything else is single-family houses. And address points, amenities, water features, gates, and whatever else might be around, maybe a bridge or a stream or something. Oh, and don't forget the streets themselves if aerial imagery is outdated. It's not terribly exciting, but it's not terribly hard, either. A few hundred house subdivision only takes a couple of hours at most to map and input. If all I wanted was street names, I'd ask the planning commission. Besides, there's some amount of fun that comes from beating Google. ;) -Nathan ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] suburban superblocks that nobody wants to survey
On 3/15/2012 6:43 PM, Nathan Mills wrote: On 3/15/2012 4:59 PM, Nathan Edgars II wrote: lots of driving and all you get is street names, since everything else is single-family houses. And address points How does this work? Do you stop at every house and write down the address? ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] suburban superblocks that nobody wants to survey
On 3/15/2012 6:10 PM, Nathan Edgars II wrote: How does this work? Do you stop at every house and write down the address? I mount a camera on the windshield and use JOSM's image plugin to place them on the track, then it's just a matter of looking at the images and identifying addresses, then estimating the position for the address point based on the picture and the GPS track. I first tried to do this with my cell phone camera, but the aperture control and dynamic range were poor enough that I could only get about half the addresses. It turns out even a cheap PS works a lot better. Obviously the points aren't going to be perfectly placed on the entrance to the house, but it's as accurate as anything else those of us without pro equipment can really do. My original tactic was to write down the addresses at the end of each block and use interpolation ways. In a lot of subdivisions around here that works really well because the city forces them to address (mostly) correctly. In other cities where the addressing is less regular it doesn't work so well, in that not all addresses in the range actually exist. The problem with this method becomes apparent when you find that the subdivision you've mapped is too new to have aerial coverage. Too much guessing for me. The really sad thing is when you get to a new subdivision before there are any houses. No address points at all in that case. ;) -Nathan ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] suburban superblocks that nobody wants to survey
On 3/15/12 7:10 PM, Nathan Edgars II wrote: On 3/15/2012 6:43 PM, Nathan Mills wrote: On 3/15/2012 4:59 PM, Nathan Edgars II wrote: lots of driving and all you get is street names, since everything else is single-family houses. And address points How does this work? Do you stop at every house and write down the address? you might get the endpoints for each segment and a couple of in between and build interpolations. richard ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
Re: [Talk-us] suburban superblocks that nobody wants to survey
On Thu, Mar 15, 2012 at 7:10 PM, Nathan Edgars II nerou...@gmail.comwrote: On 3/15/2012 6:43 PM, Nathan Mills wrote: On 3/15/2012 4:59 PM, Nathan Edgars II wrote: lots of driving and all you get is street names, since everything else is single-family houses. And address points How does this work? Do you stop at every house and write down the address? I've done audio recordings on my phone as I drive slowly or can be biking or walking. :) Best with hands free. Yell out the house numbers (and other features), one side of the street at a time (optionally with gps tracks, but USGS imagery is good) I have only so much time to drive around though and takes time to postprocess and upload. It's nice to see a bunch of new mappers in my area recently. Cheers, Katie __**_ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.**org/listinfo/talk-ushttp://lists.openstreetmap.org/listinfo/talk-us -- Katie Filbert President, Wikimedia District of Columbia http://wikimediadc.org filbe...@gmail.com @filbertkm / @wikimediadc ___ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us
