Tim, I have plenty of free time and I am interested in this import. I am about to complete a pre-processing tool that seems to “orthogonalize” building footprints pretty well using FME (safe software). I plan to present/discuss its functionalities next week on this list (vertex filtering, ensuring right angles, sorting building according to processing results, etc.). I have not examined how to break up building blocks into single units yet but I am interested to include it in the pre-processing tool if it is possible.
Daniel From: Tim Elrick [mailto:o...@elrick.de] Sent: Saturday, March 02, 2019 19:58 To: talk-ca@openstreetmap.org Subject: Re: [Talk-ca] Microsoft has released its building outlines for Canada Hi Steve, As for Montreal: We will create an import plan on the wiki as soon as we have expanded the discussion about the Montreal import from our local face-to-face group to the Montreal OSM list and agreed on importing. Before we do this, we wanted to test the feasibility of the pre-processing first, as it involves quite some postgis coding to break up the building blocks into single buildings. Only thereafter, we will suggest an import (or not), depending on the feasibility of extracting single buildings. Otherwise we will follow the hand-drawn approach as usual (and as it is done on a daily basis at the moment by a couple of OSMappers). The Microsoft data set might still be useful for remote areas. Let's explore this altogether. Cheers, Tim On 2019-03-02 19:17, OSM Volunteer stevea wrote: On Mar 2, 2019, at 3:47 PM, John Whelan <jwhelan0...@gmail.com><mailto:jwhelan0...@gmail.com> wrote: Two years ago a group of Toronto mappers submitted the City of Toronto Open Data license to the LWG to see if it was acceptable. I assume they meant to import things such as building outlines. I also assumed as I think others did that this meant Toronto mappers were happy to import the City of Toronto's data especially as it was discussed on talk-ca first. Historical info is appreciated for context, however, the LWG found Canada-wide city-by-city submissions for ODbL-compliance burdensome, given LWG's limited bandwidth. Assuming about events in the past is unhelpful, first because it is assuming (seldom helpful) and second, these events are in the past. How Toronto imported (building) data can't really help us first understand and second improve from what we learn until we know what we learned. That isn't presented here, but it could be. More recently Nate who currently lives in Toronto feels that this should be discussed once more in Toronto to work out what is desired etc. I agree with Nate. Perhaps first in Toronto, perhaps wider in talk-ca. "Once more" seems limiting, though it's possible it could suffice. Tim I think is organising Montreal open data import. Please consider adding this (and links to user: wiki or Talk pages) to the active Import wiki. Generate communication using our media! I note that Nate and Tim have different ideas about what should be imported. One is happy with bay windows and I think the other feels they should be removed. More discussion often yields consensus, especially as it "goes wide" (or as wide as is practical). We also have Pierre who is unhappy because the imported building outlines available have too many corners that are not right angles. More discussion often yields consensus. The local Ottawa mappers are content with their Open Data import and find the data quality acceptable even though Pierre has expressed reservations about it. More discussion often yields consensus. Wide area (large cities, province-wide, nationwide) imports are not easy to achieve consensus but can often reach something approaching one as data are entered, not liked, improved, liked better, et cetera. These are often an interactive, iterative process. Someone in Manitoba? mentioned there were no building outlines released for Manitoba? I apologise if I have the province name wrong. It is spelled correctly. I am not Canadian and I know that; it isn't hard to spell-check Manitoba. So we have a mixture of expectations which is only to be expected in a large group. More discussion often yields consensus. It might be part "mixture of expectations" but I'm sure that everyone will agree that "high quality data entering OSM" is expected. What can be difficult is "what do we mean by high quality?" (in addition to establishing and communicating clear goals for the importation of the data). Microsoft's Open Data provides another source of Open Data which might meet Pierre's data quality expectations. They may meet Nate's. All provinces and Territories now have Open Data building outlines available. OK, thanks for the clarification that a "union" of these datasets (Stats Canada-produced building data + Microsoft-produced building data) provide an "all provinces and Territories dataset." That truly is helpful as it makes it clear that "if Set A doesn't have your province's or Territory's building data, Set B will." Northwest Territories, Nunavut, and Yukon have populations of around 35,000 people. Realistically I don't think they have a group of local OSM mappers. Please don't "write them off" so easily. Not only does it seem "not nice," it may not be true. A better approach may be to actively develop community there, difficult as that might seem. I believe there is usually Internet available there in the villages (sometimes via clever and state-of-the-art methodologies) and it may be as simple as "shaking the trees" of the right people, then "they'll take it from there." Essentially the problem now we no longer have a Canada wide consensus on what is acceptable More discussion often yields consensus. ...appears to be how do you identify local mappers across Canada and how far away can a "local" mapper be to be considered local since it is the local mappers who make the decision about what is acceptable and I think it is they who have to drive the import process. There are no such "hard and fast" rules as this. I think you're on the right track that "hinterland" (I do not mean that disparagingly, rather more like "far away from others") OSM volunteers "drive the import process," so I again encourage you and others to "better develop" this community and let them, teach them, encourage them to "do what they will." I'm sure if a local group would like to contact me we can find resources to assist them if required. Excellent that you volunteer to be a "point person." While it's important that people "step up" like that, wider, open communication (here in talk-ca, the wiki page, the wiki's Discussion tab / talk page...) is also to be encouraged. In short, you can't cast the net too wide, so be broad in the reach to do so. De-centralize while developing both breadth and width (the whole country, at province-wide and city/vllage-wide levels) as well as "experts/consultants who are available" to answer questions and provide directional guidance and technical assistance. Good luck! SteveA California _______________________________________________ Talk-ca mailing list Talk-ca@openstreetmap.org<mailto:Talk-ca@openstreetmap.org> https://lists.openstreetmap.org/listinfo/talk-ca
_______________________________________________ Talk-ca mailing list Talk-ca@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-ca