[OSM-talk] Too subjective & problematic Re: no-go-areas
This topic has come up before, and unfortunately when you think about it, there is no objective way to define a "no go area". It's all subjective. So it doesn't belong in OSM. People do live in many of these areas, so software that didn't route in/through these areas would be pretty bad for people who need to go there! Plus, "no go areas" are often correlated with where ethnic minorities or different classes live, which is not good. On 31.12.19 16:14, Martin Trautmann wrote: hi all, did you read about the Suisse tourist couple which was shot because they got lost in a Brasilian favela? NZZ (Neue Zürcher Zeitung) from Tuesday 31.12.2019. ("Schweizer Ehepaar bei Irrfahrt duch Favela in Brasilien angeschossen") Other examples are e.g. Mafia areas within Kosovo - or name your own home town no-go area. Is there any option to mark certain areas in order to bypass routing whenever possible? ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
On 2020-01-01 15:28, Rory McCann wrote: This topic has come up before, and unfortunately when you think about it, there is no objective way to define a "no go area". It's all subjective. So it doesn't belong in OSM. People do live in many of these areas, so software that didn't route in/through these areas would be pretty bad for people who need to go there! Plus, "no go areas" are often correlated with where ethnic minorities or different classes live, which is not good. I agree! A map cannot solve a lack of general awareness when visiting a new/unknown place. Going to the mountains to hike can also be dangerous if you are not well prepared. This is of course not marked on the map! ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
On 20-01-02 12:23, pangoSE wrote: > A map cannot solve a lack of general awareness when visiting a > new/unknown place. Going to the mountains to hike can also be dangerous > if you are not well prepared. This is of course not marked on the map! I agree that I don't know any non-subjective way how to identify such an area. But a good map is for people who do NOT know this area. People who know about neither need a map nor a warning. Schönen Gruß Martin signature.asc Description: OpenPGP digital signature ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
> On Jan 11, 2020, at 12:22 PM, Martin Trautmann via talk > > and > On 20-01-02 12:23, pangoSE wrote what they wrote. To be clear, the hazards I'm hazily identifying are naturally-occurring or are human-made real-life hazards that can cause you real harm if you approach them and are not careful to avoid them, not "stay out of that neighborhood" kinds of "hazards." Things like an area which is radioactive, has a "falling hazard" (such as a pit, though I think we have "adit" for mine shafts — and we do have natural=cliff, which I agree suffices for what it is) and other unusual hazards like places which have a propensity to be repeatedly struck by lightning (that's a weird one, and kind of controversial, I know). As before, I doubt "hazard" or "no-go" will get more traction than it has (here and now), I simply make that clarification. SteveA ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
On 11/01/2020 20:22, Martin Trautmann via talk wrote: But a good map is for people who do NOT know this area. People who know about neither need a map nor a warning. Which those with more accurate, regularly updated data, such a emergency services & governmental authorities, can provide by overlaying it onto OSM. DaveF ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
On Sat, Jan 11, 2020 at 2:25 PM Martin Trautmann via talk < talk@openstreetmap.org> wrote: > On 20-01-02 12:23, pangoSE wrote: > > > A map cannot solve a lack of general awareness when visiting a > > new/unknown place. Going to the mountains to hike can also be dangerous > > if you are not well prepared. This is of course not marked on the map! > > I agree that I don't know any non-subjective way how to identify such an > area. > OK, too subjective for OSM then. > But a good map is for people who do NOT know this area. > People who know about neither need a map nor a warning. > Not our problem, we map the objective. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
On Sat, 2020-01-11 at 21:22 +0100, Martin Trautmann via talk wrote: > On 20-01-02 12:23, pangoSE wrote: > > > A map cannot solve a lack of general awareness when visiting a > > new/unknown place. Going to the mountains to hike can also be > > dangerous > > if you are not well prepared. This is of course not marked on the > > map! > > I agree that I don't know any non-subjective way how to identify such > an > area. Well, one could rely on authority, e.g. if a national police authority designated certain areas as high risk. Not saying that OSM should do it, just that it doesn't differ that much from postcode areas, where a certain authority has designated a certain number to a certain area, although there is no ground-truth that could be independently verified. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
On Sun, Jan 12, 2020 at 1:47 AM Snusmumriken wrote: > On Sat, 2020-01-11 at 21:22 +0100, Martin Trautmann via talk wrote: > > On 20-01-02 12:23, pangoSE wrote: > > > > > A map cannot solve a lack of general awareness when visiting a > > > new/unknown place. Going to the mountains to hike can also be > > > dangerous > > > if you are not well prepared. This is of course not marked on the > > > map! > > > > I agree that I don't know any non-subjective way how to identify such > > an > > area. > > Well, one could rely on authority, e.g. if a national police authority > designated certain areas as high risk. > Yeah, that's not really going to work, either. Just look at Portland. Most arrests happen in poor, black neighborhoods, but you're most likely to get hurt or killed in a suburban white area. Besides, if you really want to go that route, just composite their data as a layer over OpenStreetMap in Leaflet. There's no reason whatsoever to include it in OpenStreetMap's database. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
On Sun, 2020-01-12 at 08:35 -0600, Paul Johnson wrote: > > > On Sun, Jan 12, 2020 at 1:47 AM Snusmumriken < > snusmumriken.map...@runbox.com> wrote: > > On Sat, 2020-01-11 at 21:22 +0100, Martin Trautmann via talk wrote: > > > On 20-01-02 12:23, pangoSE wrote: > > > > > > > A map cannot solve a lack of general awareness when visiting a > > > > new/unknown place. Going to the mountains to hike can also be > > > > dangerous > > > > if you are not well prepared. This is of course not marked on > > the > > > > map! > > > > > > I agree that I don't know any non-subjective way how to identify > > such > > > an > > > area. > > > > Well, one could rely on authority, e.g. if a national police > > authority > > designated certain areas as high risk. > > Yeah, that's not really going to work, either. Just look at > Portland. Most arrests happen in poor, black neighborhoods, but > you're most likely to get hurt or killed in a suburban white area. > Besides, if you really want to go that route, just composite their > data as a layer over OpenStreetMap in Leaflet. There's no reason > whatsoever to include it in OpenStreetMap's database. I understand that it would politically sensitive, but from a data-model point of view it doesn't really differ from postcode areas (under the assumption that there's an authority that designates some areas as high-risk areas) ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
12 Jan 2020, 18:39 by snusmumriken.map...@runbox.com: > On Sun, 2020-01-12 at 08:35 -0600, Paul Johnson wrote: > >> >> >> On Sun, Jan 12, 2020 at 1:47 AM Snusmumriken < >> snusmumriken.map...@runbox.com> wrote: >> > On Sat, 2020-01-11 at 21:22 +0100, Martin Trautmann via talk wrote: >> > > On 20-01-02 12:23, pangoSE wrote: >> > > >> > > > A map cannot solve a lack of general awareness when visiting a >> > > > new/unknown place. Going to the mountains to hike can also be >> > > > dangerous >> > > > if you are not well prepared. This is of course not marked on >> > the >> > > > map! >> > > >> > > I agree that I don't know any non-subjective way how to identify >> > such >> > > an >> > > area. >> > >> > Well, one could rely on authority, e.g. if a national police >> > authority >> > designated certain areas as high risk. >> >> Yeah, that's not really going to work, either. Just look at >> Portland. Most arrests happen in poor, black neighborhoods, but >> you're most likely to get hurt or killed in a suburban white area. >> Besides, if you really want to go that route, just composite their >> data as a layer over OpenStreetMap in Leaflet. There's no reason >> whatsoever to include it in OpenStreetMap's database. >> > > I understand that it would politically sensitive, but from a data-model > point of view it doesn't really differ from postcode areas (under the > assumption that there's an authority that designates some areas as > high-risk areas) > There is a single authority assigning postal codes. With high-risk areas you may have different organizations with competing opinions. Also, in general people are not disputing postal codes. In case of officially designed dangerous zones situation is going to be different.___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
12 Jan 2020, 18:46 by matkoni...@tutanota.com: > > > > 12 Jan 2020, 18:39 by snusmumriken.map...@runbox.com: > >> I understand that it would politically sensitive, but from a data-model >> point of view it doesn't really differ from postcode areas (under the >> assumption that there's an authority that designates some areas as >> high-risk areas) >> > There is a single authority assigning > postal codes. > in a given region > With high-risk areas you may have different > organizations with competing opinions. > > Also, in general people are not disputing postal codes. > > In case of officially designed dangerous zones > situation is going to be different. > ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
On Sun, 2020-01-12 at 18:46 +0100, Mateusz Konieczny wrote: > > I understand that it would politically sensitive, but from a data- > > model > > point of view it doesn't really differ from postcode areas (under > > the > > assumption that there's an authority that designates some areas as > > high-risk areas) > > There is a single authority assigning > postal codes. > > With high-risk areas you may have different > organizations with competing opinions. > > Also, in general people are not disputing postal codes. > > In case of officially designed dangerous zones > situation is going to be different. Well, wasn't that the assumption I put forward in my previous e-mail? In Sweden, Swedish police lists 60 areas as what they call Vulnerable areas, subdivided into three groups based upon severity. Here's a wikipage about it https://en.wikipedia.org/wiki/Vulnerable_area And here's an official report by the Swedish police (in Swedish) https://polisen.se/siteassets/dokument/ovriga_rapporter/kriminell-paverkan-i-lokalsamhallet.pdf In everyday speech these are often called no-go zones ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
So you suggest that we help routers to blacklist these areas, which means living there will become even worse as various services depending on our data stop delivering there? These no-go areas are very much based on estimates. Personally if we are marking hazards then my country can pretty much be marked as a hazard - for example the curent weather in January is strong winds and snow, meaning if you are outside in normal daily clothes you are not surviving more than a few hours. This idea is for a dataset that should be solved on a local level with apps using OSM data and their own hazard estimations. This does not belong on OSM. -- Stalfur 12. janúar 2020 kl. 19:35, skrifaði "Snusmumriken" : > On Sun, 2020-01-12 at 18:46 +0100, Mateusz Konieczny wrote: > >> I understand that it would politically sensitive, but from a data- >> model >> point of view it doesn't really differ from postcode areas (under >> the >> assumption that there's an authority that designates some areas as >> high-risk areas) >> >> There is a single authority assigning >> postal codes. >> >> With high-risk areas you may have different >> organizations with competing opinions. >> >> Also, in general people are not disputing postal codes. >> >> In case of officially designed dangerous zones >> situation is going to be different. > > Well, wasn't that the assumption I put forward in my previous e-mail? > > In Sweden, Swedish police lists 60 areas as what they call Vulnerable > areas, subdivided into three groups based upon severity. Here's a > wikipage about it https://en.wikipedia.org/wiki/Vulnerable_area > > And here's an official report by the Swedish police (in Swedish) > https://polisen.se/siteassets/dokument/ovriga_rapporter/kriminell-paverkan-i-lokalsamhallet.pdf > > In everyday speech these are often called no-go zones > > ___ > talk mailing list > talk@openstreetmap.org > https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
On Sun, 2020-01-12 at 21:00 +, Jóhannes Birgir Jensson wrote: > So you suggest that we help routers to blacklist these areas, Wow, you're putting words in my mouth that is the exact opposite of what I wrote a couple of e-mails back. That is really dishonest. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
sent from a phone > On 12. Jan 2020, at 08:47, Snusmumriken > wrote: > > Not saying that OSM should do it, just that it doesn't differ that much > from postcode areas, where a certain authority has designated a certain > number to a certain area, although there is no ground-truth that could > be independently verified of course there is ground truth that can be verified: ask people or businesses in the area about their postcode / address Cheers Martin ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
On Sun, 2020-01-12 at 22:53 +0100, Martin Koppenhoefer wrote: > On 12. Jan 2020, at 08:47, Snusmumriken < > > snusmumriken.map...@runbox.com> wrote: > > > > Not saying that OSM should do it, just that it doesn't differ that > > much > > from postcode areas, where a certain authority has designated a > > certain > > number to a certain area, although there is no ground-truth that > > could > > be independently verified > > of course there is ground truth that can be verified: ask people or > businesses in the area about their postcode / address I don't believe any postcode _area_ in OSM has ever been created that way. But I would love to learn of one. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
On Sun, Jan 12, 2020 at 11:49 AM Mateusz Konieczny wrote: > > > > 12 Jan 2020, 18:39 by snusmumriken.map...@runbox.com: > > On Sun, 2020-01-12 at 08:35 -0600, Paul Johnson wrote: > > > > On Sun, Jan 12, 2020 at 1:47 AM Snusmumriken < > snusmumriken.map...@runbox.com> wrote: > > On Sat, 2020-01-11 at 21:22 +0100, Martin Trautmann via talk wrote: > > > On 20-01-02 12:23, pangoSE wrote: > > > > > > > A map cannot solve a lack of general awareness when visiting a > > > > new/unknown place. Going to the mountains to hike can also be > > > > dangerous > > > > if you are not well prepared. This is of course not marked on > > the > > > > map! > > > > > > I agree that I don't know any non-subjective way how to identify > > such > > > an > > > area. > > > > Well, one could rely on authority, e.g. if a national police > > authority > > designated certain areas as high risk. > > Yeah, that's not really going to work, either. Just look at > Portland. Most arrests happen in poor, black neighborhoods, but > you're most likely to get hurt or killed in a suburban white area. > Besides, if you really want to go that route, just composite their > data as a layer over OpenStreetMap in Leaflet. There's no reason > whatsoever to include it in OpenStreetMap's database. > > > I understand that it would politically sensitive, but from a data-model > point of view it doesn't really differ from postcode areas (under the > assumption that there's an authority that designates some areas as > high-risk areas) > > There is a single authority assigning > postal codes. > Well, two in the US. > Also, in general people are not disputing postal codes. > The US Census Bureau and the Postal Service, but that's their problem to sort out; just putting city and state will still get your mail there and it's specific enough to wayfind. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
In the USA a postal code is not actually an area, but a set of addresses. Often they are all in one area, but sometimes the area is not clearly defined. This is partially why postal codes are usually just added to the POI directly in the USA. Trying to make a sensible set of areas or boundaries will not work for all USPS postal codes. Joseph Eisenberg ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
On Mon, Jan 13, 2020 at 6:34 PM Joseph Eisenberg wrote: > In the USA a postal code is not actually an area, but a set of > addresses. Often they are all in one area, but sometimes the area is > not clearly defined. This is partially why postal codes are usually > just added to the POI directly in the USA. Trying to make a sensible > set of areas or boundaries will not work for all USPS postal codes. > For mailing addresses, yes. For street addresses, maybe. The Census does have areas for their ZIP codes. Whether a particular area uses Census or Postal zips depends on whether or not they're served by the USPS directly (and there's decent chunks that don't). Edge case is super edge, though, and the de-facto way of dealing with this is to just use the ZIP observed on the address marker or ask someone who lives/works there what the ZIP is ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Too subjective & problematic Re: no-go-areas
I agree. In the USA, the five-digit postal code was introduced in 1963 and called a "ZIP code," for "Zone" (first digit), "Improvement" (second and third digits), "Plan" (fourth and fifth digits). In 1983, nine-digit codes were introduced by adding a hyphen after the five digits and four more digits ("Sector" sixth and seventh digits and "Segment" eighth and ninth digits), hence the newer designation "ZIP+4." More recently, the Coding Accuracy Support System (CASS) systems added two more digits to standardize the exact "delivery point" (where 10th and 11th digits are calculated by CASS software based on the number in the address) and a final, 12th digit as a checksum. Helpful to remember about ZIP codes (proposed since 1978) is that they are NOT geographic areas (as can be mapped in a map), but rather "routing algorithms" helpful to facilitate mail delivery. I believe it is widespread consensus in OSM that while there are places where adding boundary=census polygons is considered helpful data (especially in Alaska's Unorganized Borough, as they are a joint effort by both the US Department of Commerce's Census Bureau AND the state of Alaska and have become a de facto, though not necessarily de jure definition of administrative areas), usually, adding census data to OSM is not especially helpful. But there are much of these extant now. I also believe it is widespread consensus that OSM should not contain postal (ZIP) code data in the USA, as ZIP codes (strictly speaking) cannot always (or even usually?) be defined as geographic areas, they are rather better thought of as a "routing algorithm to help facilitate mail delivery." SteveA California > On Jan 13, 2020, at 4:34 PM, Joseph Eisenberg > wrote: > > In the USA a postal code is not actually an area, but a set of > addresses. Often they are all in one area, but sometimes the area is > not clearly defined. This is partially why postal codes are usually > just added to the POI directly in the USA. Trying to make a sensible > set of areas or boundaries will not work for all USPS postal codes. > > Joseph Eisenberg > > ___ > talk mailing list > talk@openstreetmap.org > https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk