Re: [OSM-talk] Addressing SIG

[stevea]

Nicely answered, I appreciate that!
SteveA

> On Nov 8, 2019, at 4:02 PM, Martin Koppenhoefer  
> wrote:
> 
> 
> 
> sent from a phone
> 
>> On 9. Nov 2019, at 00:48, stevea  wrote:
>> 
>> I wouldn't say "all" addresses, as Facebook isn't "all" of us.  
> 
> 
> of course, I apologize if this came along like a campaign just with facebook, 
> it was just an example that facebook was mentioned, because they are our 
> biggest data user (Apple as well, but they don’t use OpenStreetMap data in 
> their most  important markets, AFAIK). Collecting _all_ addresses is a 
> project goal (it’s part of mapping the whole world), and was not referring 
> specifically to facebook, nor had I imagined a campaign just with facebook 
> (if they are interested in this anyway), making such an announcement could be 
> an occasion for the media in general for featuring OpenStreetMap. The people 
> vs. Big Tech etc.
> 
> Cheers Martin


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing SIG

[Martin Koppenhoefer]

sent from a phone

> On 9. Nov 2019, at 00:48, stevea  wrote:
> 
> I wouldn't say "all" addresses, as Facebook isn't "all" of us.  


of course, I apologize if this came along like a campaign just with facebook, 
it was just an example that facebook was mentioned, because they are our 
biggest data user (Apple as well, but they don’t use OpenStreetMap data in 
their most  important markets, AFAIK). Collecting _all_ addresses is a project 
goal (it’s part of mapping the whole world), and was not referring specifically 
to facebook, nor had I imagined a campaign just with facebook (if they are 
interested in this anyway), making such an announcement could be an occasion 
for the media in general for featuring OpenStreetMap. The people vs. Big Tech 
etc.

Cheers Martin 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing SIG

[stevea]

I wouldn't say "all" addresses, as Facebook isn't "all" of us.  Also, it's an 
ambition, a gleam in a collective eye, a vision, something ahead in the future 
as a goal.  There will be, rightly, many paths to get there, rather than a 
single one.  This is true of any major goal in OSM.

SteveA

> On Nov 8, 2019, at 3:44 PM, Martin Koppenhoefer  
> wrote:
> We could announce a campaign, “citizen mapping project collects all the 
> addresses in the world and provides them freely to everybody” or similar.


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing SIG

[Martin Koppenhoefer]

sent from a phone

> On 8. Nov 2019, at 13:59, Simon Poole  wrote:
> 
> Just imagine that we increase the number of new OSM contributors by an
> order of magnitude, to ~2'000'000 per year, and just as magically we get
> them to make the single edit they typically make to be adding an
> address


imagine every facebook user would put their home address, or every business 
owner their business address. To make this happen, the procedure must be dead 
simple, and not require you understand an editor, or use a complex website with 
bells and whistles, but so overwhelming to my grand aunt that she will close 
the page as soon as she finds the button to do so.

We could announce a campaign, “citizen mapping project collects all the 
addresses in the world and provides them freely to everybody” or similar.

Cheers Martin 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing SIG

[marc marc]

Simon Poole:
>> how can the new contributor wishing to add an address find this tool?

> That doesn't make the slightest difference, because the only people
> adding addresses in any meaningful way are those 1% of contributors.

it's the problem of the egg and the chicken :
if effective tools were accessible to 2% 3% 10% of contributors,
then the speed would be multiplied by 2, 3 or 10

A more effective contribution would be to consider 3 types
of location/contributor:

- those where the opendata is of quality: we should talk about import

- those where the opendata exists but where the import has not been 
accepted and/or not yet done and/or new address exist: the "normal" 
contributor (99%) should be able to easily access the opendata in
an editor and validate in one click if it is correct. or correct
the position easily. osmose opendata-fork allow that but it'sn't
a editor that newbies find when surfing on osm.org.
and only a very few location are listed for addr.

- those where the opendata does not exist: a drop-down list
to avoid having to rewrite the street name is a minimum.
It also avoid typo and letter case.

Thinking about the time spent on the addresses I added,
I think I could have been much more efficient for both those
added by survey and those from opendata.
so by spending the same amount of time on it, I could have added
a lot more. And it would have been much more motivating to do it
often.

Regards,
Marc
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing SIG

[Simon Poole]

Am 08.11.2019 um 13:19 schrieb marc marc:
> Hello,
>
> Simon Poole :
>> The issue with addresses is definitely not due to a lack of tools 
>> for OSM contributors. For example
>> https://regio-osm.de/hausnummerauswertung/anzeige_dynamisch.html?land=Schweiz&lon=8.71423&lat=47.05777&zoom=8&layers=B
> lack of tools not only mean "no tool exist",
> it also means "tool not found for the contributor"
>
> how can the new contributor wishing to add an address find this tool?
> there's no way he'll find it. it's an advanced tool for the 1%
> of the most motivated contributors of for newbie at a mapping party.
> the other clicks on edit or note. it is osm.org's ergonomics it-self 
> and/or the greeting message during registration that must be improved
> so that the new contributor can find the tool best suited to his 
> contribution

That doesn't make the slightest difference, because the only people
adding addresses in any meaningful way are those 1% of contributors.

Just imagine that we increase the number of new OSM contributors by an
order of magnitude, to ~2'000'000 per year, and just as magically we get
them to make the single edit they typically make to be adding an
address, instead of whatever they actually wanted to do. Even then, in
the as good as it gets fantasy scenario, it would only be 20% of the
current run rate. And that in turn is probably an order of magnitude or
so too low for Steve (aka 50 years or so to "complete" world wide coverage).

Simon


> Regards,
> Marc
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk



signature.asc
Description: OpenPGP digital signature
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing SIG

[marc marc]

Hello,

Simon Poole :
> The issue with addresses is definitely not due to a lack of tools 
> for OSM contributors. For example
> https://regio-osm.de/hausnummerauswertung/anzeige_dynamisch.html?land=Schweiz&lon=8.71423&lat=47.05777&zoom=8&layers=B

lack of tools not only mean "no tool exist",
it also means "tool not found for the contributor"

how can the new contributor wishing to add an address find this tool?
there's no way he'll find it. it's an advanced tool for the 1%
of the most motivated contributors of for newbie at a mapping party.
the other clicks on edit or note. it is osm.org's ergonomics it-self 
and/or the greeting message during registration that must be improved
so that the new contributor can find the tool best suited to his 
contribution

Regards,
Marc
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing SIG

[Simon Poole]

Just as a further data point for the discussion: we are currently adding
roughly 10'000'000 addresses per year relatively constant since 2013,
with some exceptions due to imports (mainly NL in 2014 I believe).




signature.asc
Description: OpenPGP digital signature
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing SIG

[Simon Poole]

The issue with addresses is definitely not due to a lack of tools for
OSM contributors. For example
https://regio-osm.de/hausnummerauswertung/anzeige_dynamisch.html?land=Schweiz&lon=8.71423&lat=47.05777&zoom=8&layers=B
which covers essentially all the analytics needed for comparison with
open data datasets and that since years (not mention the various address
QA layers available, again since years). On the data entry side there
are good tools both for surveying, import and conflation en masse.

What might be missing is simpler variant of
https://osmybiz.osm.ch/#/18/47.40514/8.40289 (I actually have the domain
addmyaddress.org stashed away somewhere for that), but while it would be
nice to provide a simple facility for people to check and potentially
add their address, it is clear that the targeted long tail is not going
to make a substantial difference in coverage.

So what it really boils down to is grunt work*time (and that is even
true for imports). In Central Europe we are well on the way to
acceptable coverage, given a couple of years more I suspect it will be
really good. Nearly everywhere else (special case the US, and apologies
to all the the exceptions to "nearly everywhere") we are missing
essential metadata that should come first, aka road names and
references, POIs, places and so on, essentially all the stuff that
building doodling and ML doesn't provide, but is essential to actually
having a usable map.

Simon

Am 07.11.2019 um 13:18 schrieb marc marc:
> Hello,
>
>> We've been "addressing the address topic" for more than 
>> 5 years in France with our BANO project.
> and despite the amount of opendata information available, 5 years later,
> there is still a lot of red (missing road name or mismatch between
> osm and opendata).
>
> I agree with the original author: there is a lack of a simple tool
> to contribute more effectively to addresses.
> for example a new contributor has no way to validate the name of a 
> street from the opendata. Osmose and BANO layers are good advanced tools 
> but are not adapted to this kind of beginner audience but also out of 
> their sight.
>
> there is also a lack of awareness that missing addresses are
> a lack of osm compared to some proprietary solutions.
>
> Regards,
> Marc
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk


signature.asc
Description: OpenPGP digital signature
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing SIG

[marc marc]

Hello,

> We've been "addressing the address topic" for more than 
> 5 years in France with our BANO project.

and despite the amount of opendata information available, 5 years later,
there is still a lot of red (missing road name or mismatch between
osm and opendata).

I agree with the original author: there is a lack of a simple tool
to contribute more effectively to addresses.
for example a new contributor has no way to validate the name of a 
street from the opendata. Osmose and BANO layers are good advanced tools 
but are not adapted to this kind of beginner audience but also out of 
their sight.

there is also a lack of awareness that missing addresses are
a lack of osm compared to some proprietary solutions.

Regards,
Marc
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing SIG

[Yves]

Wait,. .. when was the 'noname' layer gone?
Yves

Le 7 novembre 2019 08:06:27 GMT+01:00, Christian Quest 
 a écrit :
>We've been "addressing the address topic" for more than 5 years in
>France
>with our BANO project.
>
>Here is an overlay I created back then to show existing and missing
>address
>data in OSM compared to available OSM compatible sources.
>
>http://osm13.openstreetmap.fr/~cquest/leaflet/bano.html#16/48.7908/2.6542
>
>Green: the address is in OSM (and the named road too)
>Blue: address is missing but the road name exist in OSM
>Red: address missing and we found no road with that name nearby
>
>If you want to make missing data obvious, you should no dimm the
>shapes,
>but make them highly visible.
>The goal with the above rendering became "dégommer du rouge" (get rid
>of
>red).
>
>
>Le mar. 5 nov. 2019 à 19:43, Steve Coast  a écrit
>:
>
>> Hello
>>
>>
>>
>> Maps have three basic components: Display (does it look nice?),
>Routing
>> (Can I get from a to b?) and Geocoding (Where is this address?).
>>
>>
>>
>> OSM is extremely good at the first one, and pretty good at the second
>one.
>> But it’s pretty deficient in the third area: address data.
>>
>>
>>
>> The question is, how can we fix this? Addresses are a big, big
>problem in
>> terms of how much data we need to go collect. There are a few ways
>forward
>> with outside commercial or government data, but they tend to be
>difficult
>> because the data is patchy or licensed in ways that aren’t very
>compatible
>> with OSM.
>>
>>
>>
>> It seems like it would be a good idea to think about this from the
>bottom
>> up in a community way, and this doesn’t really exist in OSM right
>now. It
>> seems like we need better feedback loops to:
>>
>>
>>
>>1. Community can see where the address data is (and isn’t),
>because
>>it’s not very obvious today when using osm.org
>>2. Make the tools to add address data better so that it’s easier
>to
>>fix.
>>
>>
>>
>> To that end, here’s a tile server that highlights address data:
>>
>>
>>
>>
>>
>http://ec2-52-50-19-165.eu-west-1.compute.amazonaws.com/#10/39.7561/-104.9574
>>
>>
>>
>> It shows roads with address data normally and kind-of hides other
>roads,
>> to make it obvious that “something is wrong with this map”. We could
>have a
>> tag (maybe it exists already) that says “this road doesn’t have
>addresses”
>> and/or a tag that says “this road is complete”. (right now it’s just
>got
>> Colorado and Utah in it).
>>
>>
>>
>> When OSM started, the map looked very broken and incomplete because
>there
>> was missing data all over the place. This created a large incentive
>to go
>> fix the map. The idea with this tileserver is to do the same thing
>and make
>> the map look broken to create a large incentive to fix it. If we, one
>day,
>> switched the main osm.org site to using this rendering then it would
>> create an urgent need to find all the addresses in the places where
>they
>> exist. It could also be done on a temporary basis for a few weeks, or
>on a
>> per-country basis or some other slow introduction to see if it
>worked. It’s
>> just an idea.
>>
>>
>>
>> On the tools side, there’s much that can be done to make collecting
>and
>> entering addresses easier. I’ve been collecting UI/UX changes to
>tools
>> (e.g. iD or Go Map!) that would make addresses better:
>>
>>
>>
>> https://wiki.openstreetmap.org/wiki/Address_SIG
>>
>>
>>
>> It also seems worthwhile to create a group of people interested in
>> addressing in OSM (an address special interest group or working
>group) to
>> push these ideas forward so that we can “finish” OSM by getting all
>the
>> addresses done.
>>
>>
>>
>> What do you think?
>>
>>
>>
>> Best
>>
>>
>>
>> Steve
>>
>>
>> ___
>> talk mailing list
>> talk@openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/talk
>>
>
>
>-- 
>Christian Quest - OpenStreetMap France
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing SIG

[Mateusz Konieczny]

See https://wiki.openstreetmap.org/wiki/StreetComplete/Quests 

and https://wiki.openstreetmap.org/wiki/StreetComplete
for more info, including full list of available predefined tasks.

6 Nov 2019, 20:40 by talk@openstreetmap.org:

> For your usecase, Tom, perhaps Street-Complete would work for you if you 
> turned on all the building-related quests and turned off the other quests?
> https://play.google.com/store/apps/details?id=de.westnordost.streetcomplete&hl=en_US
>  
> 
>
> On Wed, Nov 6, 2019 at 11:33 AM Tom Russell <> tom.russ...@ouce.ox.ac.uk 
> > > wrote:
>
>>
>> Am Mi., 6. Nov. 2019 um 09:17 Uhr schrieb Oleksiy Muzalyev <>> 
>> oleksiy.muzal...@bluewin.ch >> >:
>>
>>
>>>
>>> On the main >>> osm.org >> building and select "Show address" or "Add a note here" . What if a new 
>>> type of a note is introduced, a structured address note?
>>>
>>>
>>
>>  
>>
>>
>> This is something I’ve been thinking about recently, with a slightly broader 
>> interest in building data more generally.
>>
>>
>>  
>>
>>
>> As part of an academic project (>> https://colouring.london/ 
>> >> ) looking at buildings in London in the UK, 
>> we’re thinking about how to collect various building data attributes. We’re 
>> not currently using OpenStreetMap data for our buildings, however I would be 
>> interested to look into ways of linking to, working with, or building on OSM 
>> in the future.
>>
>>
>>  
>>
>>
>> It might be interesting to create a lighter, more restricted user interface 
>> for editing the map, for example following the idea of an “Address” 
>> structured note, or to collect other data about buildings (number of 
>> storeys, commercial use). Or I could imagine a system that doesn’t edit OSM 
>> directly but creates a “review queue” of linked data which could feed into 
>> the main database as mappers work through it.
>>
>>
>>  
>>
>>
>> All early ideas - in any case, I’ll be interested to follow an Addressing 
>> SIG.
>>
>>
>>  
>>
>>
>> Best wishes,
>>
>>
>> Tom
>>
>> ___
>>  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] Addressing SIG

[Christian Quest]

We've been "addressing the address topic" for more than 5 years in France
with our BANO project.

Here is an overlay I created back then to show existing and missing address
data in OSM compared to available OSM compatible sources.

http://osm13.openstreetmap.fr/~cquest/leaflet/bano.html#16/48.7908/2.6542

Green: the address is in OSM (and the named road too)
Blue: address is missing but the road name exist in OSM
Red: address missing and we found no road with that name nearby

If you want to make missing data obvious, you should no dimm the shapes,
but make them highly visible.
The goal with the above rendering became "dégommer du rouge" (get rid of
red).


Le mar. 5 nov. 2019 à 19:43, Steve Coast  a écrit :

> Hello
>
>
>
> Maps have three basic components: Display (does it look nice?), Routing
> (Can I get from a to b?) and Geocoding (Where is this address?).
>
>
>
> OSM is extremely good at the first one, and pretty good at the second one.
> But it’s pretty deficient in the third area: address data.
>
>
>
> The question is, how can we fix this? Addresses are a big, big problem in
> terms of how much data we need to go collect. There are a few ways forward
> with outside commercial or government data, but they tend to be difficult
> because the data is patchy or licensed in ways that aren’t very compatible
> with OSM.
>
>
>
> It seems like it would be a good idea to think about this from the bottom
> up in a community way, and this doesn’t really exist in OSM right now. It
> seems like we need better feedback loops to:
>
>
>
>1. Community can see where the address data is (and isn’t), because
>it’s not very obvious today when using osm.org
>2. Make the tools to add address data better so that it’s easier to
>fix.
>
>
>
> To that end, here’s a tile server that highlights address data:
>
>
>
>
> http://ec2-52-50-19-165.eu-west-1.compute.amazonaws.com/#10/39.7561/-104.9574
>
>
>
> It shows roads with address data normally and kind-of hides other roads,
> to make it obvious that “something is wrong with this map”. We could have a
> tag (maybe it exists already) that says “this road doesn’t have addresses”
> and/or a tag that says “this road is complete”. (right now it’s just got
> Colorado and Utah in it).
>
>
>
> When OSM started, the map looked very broken and incomplete because there
> was missing data all over the place. This created a large incentive to go
> fix the map. The idea with this tileserver is to do the same thing and make
> the map look broken to create a large incentive to fix it. If we, one day,
> switched the main osm.org site to using this rendering then it would
> create an urgent need to find all the addresses in the places where they
> exist. It could also be done on a temporary basis for a few weeks, or on a
> per-country basis or some other slow introduction to see if it worked. It’s
> just an idea.
>
>
>
> On the tools side, there’s much that can be done to make collecting and
> entering addresses easier. I’ve been collecting UI/UX changes to tools
> (e.g. iD or Go Map!) that would make addresses better:
>
>
>
> https://wiki.openstreetmap.org/wiki/Address_SIG
>
>
>
> It also seems worthwhile to create a group of people interested in
> addressing in OSM (an address special interest group or working group) to
> push these ideas forward so that we can “finish” OSM by getting all the
> addresses done.
>
>
>
> What do you think?
>
>
>
> Best
>
>
>
> Steve
>
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk
>


-- 
Christian Quest - OpenStreetMap France
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing SIG

[Kathleen Lu via talk]

For your usecase, Tom, perhaps Street-Complete would work for you if you
turned on all the building-related quests and turned off the other quests?
https://play.google.com/store/apps/details?id=de.westnordost.streetcomplete&hl=en_US

On Wed, Nov 6, 2019 at 11:33 AM Tom Russell 
wrote:

> Am Mi., 6. Nov. 2019 um 09:17 Uhr schrieb Oleksiy Muzalyev <
> oleksiy.muzal...@bluewin.ch>:
>
> On the main osm.org site one can right-click on a building and select
> "Show address" or "Add a note here" . What if a new type of a note is
> introduced, a structured address note?
>
>
>
> This is something I’ve been thinking about recently, with a slightly
> broader interest in building data more generally.
>
>
>
> As part of an academic project (https://colouring.london/) looking at
> buildings in London in the UK, we’re thinking about how to collect various
> building data attributes. We’re not currently using OpenStreetMap data for
> our buildings, however I would be interested to look into ways of linking
> to, working with, or building on OSM in the future.
>
>
>
> It might be interesting to create a lighter, more restricted user
> interface for editing the map, for example following the idea of an
> “Address” structured note, or to collect other data about buildings (number
> of storeys, commercial use). Or I could imagine a system that doesn’t edit
> OSM directly but creates a “review queue” of linked data which could feed
> into the main database as mappers work through it.
>
>
>
> All early ideas - in any case, I’ll be interested to follow an Addressing
> SIG.
>
>
>
> Best wishes,
>
> Tom
> ___
> 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] Addressing SIG

[Tom Russell]

Am Mi., 6. Nov. 2019 um 09:17 Uhr schrieb Oleksiy Muzalyev 
mailto:oleksiy.muzal...@bluewin.ch>>:
On the main osm.org site one can right-click on a building and 
select "Show address" or "Add a note here" . What if a new type of a note is 
introduced, a structured address note?

This is something I’ve been thinking about recently, with a slightly broader 
interest in building data more generally.

As part of an academic project (https://colouring.london/) looking at buildings 
in London in the UK, we’re thinking about how to collect various building data 
attributes. We’re not currently using OpenStreetMap data for our buildings, 
however I would be interested to look into ways of linking to, working with, or 
building on OSM in the future.

It might be interesting to create a lighter, more restricted user interface for 
editing the map, for example following the idea of an “Address” structured 
note, or to collect other data about buildings (number of storeys, commercial 
use). Or I could imagine a system that doesn’t edit OSM directly but creates a 
“review queue” of linked data which could feed into the main database as 
mappers work through it.

All early ideas - in any case, I’ll be interested to follow an Addressing SIG.

Best wishes,
Tom
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing SIG

[Martin Koppenhoefer]

Am Mi., 6. Nov. 2019 um 09:17 Uhr schrieb Oleksiy Muzalyev <
oleksiy.muzal...@bluewin.ch>:

> On the main osm.org site one can right-click on a building and select
> "Show address" or "Add a note here" . What if a new type of a note is
> introduced, a structured address note?
>
> The structured address note entry form will consist of several input
> fields:
>
> Building number: ...
> Street (avenue): ...
> Post index (zip code, postal code): ...
> City (town, village): ...
> Region (State, Canton, Department): ...
> Country: ...
>
> It could also include a captcha to prevent mass automated entries.
>
> Usually people do know very well addresses of buildings in which they
> live, work, or which they visit. This way they will know that the OSM map
> is interested to map the correct postal address, since there is this
> readily available structured address entry form. So they can add an address
> without learning how to use a complex map editor.
>


I also believe it would be very benificial to call explicitly for address
contributions from unregistered users of the map. I am seeing a lot of
notes created by anonymous users, and providing a formalized way for
address entry would likely make people more frequently provide this kind of
information. If the form is structured, it could also be converted to
actual osm data more quickly by mappers (i.e. this assumes that those
anonymously contributed addresses would still have to be individually
reviewed by mappers).

In order to get good address coverage, provided we prefer local knowledge
over imported data, we must increase the contributor base. Notes have
proven to be able to provide useful information from people who are
reluctant to register.

Cheers
Martin
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing SIG

[Oleksiy Muzalyev]

On the main osm.org site one can right-click on a building and select 
"Show address" or "Add a note here" . What if a new type of a note is 
introduced, a structured address note?


The structured address note entry form will consist of several input fields:

Building number: ...
Street (avenue): ...
Post index (zip code, postal code): ...
City (town, village): ...
Region (State, Canton, Department): ...
Country: ...

It could also include a captcha to prevent mass automated entries.

Usually people do know very well addresses of buildings in which they 
live, work, or which they visit. This way they will know that the OSM 
map is interested to map the correct postal address, since there is this 
readily available structured address entry form. So they can add an 
address without learning how to use a complex map editor.


Best regards,
Oleksiy

On 11/5/19 19:37, Steve Coast wrote:


Hello

Maps have three basic components: Display (does it look nice?), 
Routing (Can I get from a to b?) and Geocoding (Where is this address?).


OSM is extremely good at the first one, and pretty good at the second 
one. But it’s pretty deficient in the third area: address data.


The question is, how can we fix this? Addresses are a big, big problem 
in terms of how much data we need to go collect. There are a few ways 
forward with outside commercial or government data, but they tend to 
be difficult because the data is patchy or licensed in ways that 
aren’t very compatible with OSM.


It seems like it would be a good idea to think about this from the 
bottom up in a community way, and this doesn’t really exist in OSM 
right now. It seems like we need better feedback loops to:


 1. Community can see where the address data is (and isn’t), because
it’s not very obvious today when using osm.org
 2. Make the tools to add address data better so that it’s easier to fix.

To that end, here’s a tile server that highlights address data:

http://ec2-52-50-19-165.eu-west-1.compute.amazonaws.com/#10/39.7561/-104.9574

It shows roads with address data normally and kind-of hides other 
roads, to make it obvious that “something is wrong with this map”. We 
could have a tag (maybe it exists already) that says “this road 
doesn’t have addresses” and/or a tag that says “this road is 
complete”. (right now it’s just got Colorado and Utah in it).


When OSM started, the map looked very broken and incomplete because 
there was missing data all over the place. This created a large 
incentive to go fix the map. The idea with this tileserver is to do 
the same thing and make the map look broken to create a large 
incentive to fix it. If we, one day, switched the main osm.org site to 
using this rendering then it would create an urgent need to find all 
the addresses in the places where they exist. It could also be done on 
a temporary basis for a few weeks, or on a per-country basis or some 
other slow introduction to see if it worked. It’s just an idea.


On the tools side, there’s much that can be done to make collecting 
and entering addresses easier. I’ve been collecting UI/UX changes to 
tools (e.g. iD or Go Map!) that would make addresses better:


https://wiki.openstreetmap.org/wiki/Address_SIG

It also seems worthwhile to create a group of people interested in 
addressing in OSM (an address special interest group or working group) 
to push these ideas forward so that we can “finish” OSM by getting all 
the addresses done.


What do you think?

Best

Steve


___
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] Addressing SIG

[James]

It's a pretty cool concept, but doesn't necessarily invoke ALL addresses
have been found, what happens if a few addresses are there? What happens if
someone adds 1 or 2 addresses?

Pretty good QA tool I'm guessing?
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing systems (Was: Paper/Article about stagnation in OSM)

[Steve Doerr]

On 03/08/2018 09:22, oleksiy.muzal...@bluewin.ch wrote:
I just tried to create the 3-words address for a building in Odessa, 
Ukraine. The system suggested "dressings.cookies.brothers". It would 
be close to impossible to transmit these three words over the 
telephone to a local taxi dispatcher.


Some people may just not know English words well enough. The same 
about 8 English letters. But 8 Cyrillic letters may work. If the UTF8 
encoding is used in a database then both Latin and Cyrillic letters 
could be used, and, perhaps, other scripts.


You can set the language to Russian. This, I believe, gives the address 
as лотерея.русый.замок


--
Steve

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing systems (Was: Paper/Article about stagnation in OSM)

[oleksiy.muzal...@bluewin.ch]

I understand what you mean and I share your view. I would like just to mention that the European civilization absorbed others' achievements massively.For example, corn ended for good famines in Europe. At the same time, it was the product of five thousand years selection effort by people of South America. Or Hindu-Arabic numerals, or coffee from Ethiopia, the list is very long.Perhaps, it is still possible to co-develop without interference. Perhaps, by creating the new open source address system, we get the feedback and improve our obsolete 18th century address system too.Best regards,OleksiySent from my Huawei Mobile___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing systems (Was: Paper/Article about stagnation in OSM)

[_ dikkeknodel]

To me this is just another way of colonisation, forcing (technical) systems 
upon other people who have no say in them. They are communicated as a means to 
help people, but are mainly to make a buck in the end.

Instead we at OSM should leave it up to the people themselves to choose how to 
describe their location in a way they feel like, and provide a means to do so 
based on that. Technology should be designed to support people in their 
preferred way of life, not to force them to a way of life because that’s what 
most easily is implemented in technology.

Cheers,
dikkeknodel
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing systems (Was: Paper/Article about stagnation in OSM)

[oleksiy.muzal...@bluewin.ch]

I just tried to create the 3-words address for a building in Odessa, Ukraine. The system suggested "dressings.cookies.brothers". It would be close to impossible  to transmit these three words over the telephone to a local taxi dispatcher.Some people may just not know English words well enough. The same about 8 English letters. But 8 Cyrillic letters may work. If the UTF8 encoding is used in a database then both Latin and Cyrillic letters could be used, and, perhaps, other scripts.The lattitude and longitude numbers will not reflect the apartment level, entrance lock code, name, and other information.The classic addresses are also somewhat outdated in this respect. For example, when I order a delivery, I must provide also the level and entrance code, otherwise a courrier will not be able to deliver an order.If the database is not big for 3-words, it would be even smaller for 8-letters approach. So, probably, it could fit on a mobile device, at least for one country or a region.Best regards,OleksiySent from my Huawei Mobile___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing System in OSM

[Josh Doe]

Start here:
http://wiki.openstreetmap.org/wiki/Address

Generally you just need the addr:housenumber and addr:street tags, but
if you're more specific with your question we can do a better job
answering.

Nominatim can also be useful for debugging purposes, go here:
http://open.mapquestapi.com/nominatim/v1/

Search for a place, then view details to see some useful information.

-Josh

On Tue, Jun 7, 2011 at 10:20 AM, Besfort Guri  wrote:
> Who can help me with Addressing System in OpenStreetMap, I need like a
> tutorial for that because I am trying to figure out some problems in Kosovo,
> but I need help to do that ?...
>
> --
> Regards
> Besfort Guri
> +377 44 49 88 91
> www.besiguri.wordpress.com
> http://besfortp.posterous.com/
>
>
>
> ___
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk
>
>

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing Question

[marcus.wolschon]

On Thu, 12 Nov 2009 14:29:14 -0500, Anthony  wrote:
> On Thu, Nov 12, 2009 at 9:14 AM, Andy Allan 
wrote:
>> It's a fairly well established convention that in OSM it's the
>> houses/plots, not the road centrelines, that are addressed.
> 
> But that doesn't always reflect reality.  The reality, at least in
> many parts of the world, is that the streets are given blocks of
> potential addresses, and the houses/plots/whatever are given actual
> addresses from those potential address blocks.

So your point being?
These blocks can be interpolation-ways next to the way
and if you like relations you can have both grouped
in an associatedStreet-relation.

>> I'd say it's better to approximate the gap between the road and the
>> houses
>> (10m?) than to just put it on the centreline due to that being easier.
> 
> First of all, how would you approximate the gap?  You mean by hand?

10m along the normal of the road.

> Secondly, what if the houses aren't yet there?  Tiger address data
> represents *potential* address blocks, not *actual* address blocks.
> There may or may not be any actual houses along those roads.

Then we have to assume it's there until a mapper who can actually look
for houses can correct this. That's the best we can do.

Marcus

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing Question

[Mike N.]

FYI - I applied the experimental script which creates address interpolation 
ways at -

http://svn.openstreetmap.org/applications/utils/import/tiger2osm/shape_to_osm-Tiger.py

The results are at 

http://cid-b17e2f1a4d519b13.skydrive.live.com/self.aspx/Public/tl%5E_2009%5E_45045%5E_addrInterpolation.zip

 (7 MB zipped, 88 MB unzipped).

  Runtime for my county was about 45 seconds.___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing Question

[andrzej zaborowski]

2009/11/12 Anthony :
> On Thu, Nov 12, 2009 at 9:14 AM, Andy Allan  wrote:
>> It's a fairly well established convention that in OSM it's the
>> houses/plots, not the road centrelines, that are addressed.
>
> But that doesn't always reflect reality.  The reality, at least in
> many parts of the world, is that the streets are given blocks of
> potential addresses, and the houses/plots/whatever are given actual
> addresses from those potential address blocks.

If we're going to go into detail, no type of interpolation reflects
reality, it's just interpolation.  It's there to help the mappers add
proper address / other information and in the meantime give users
approximate geocoding functionality.

If you look at TIGER a lot of it doesn't reflect reality, it just one step.

Cheers

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing Question

[Anthony]

On Thu, Nov 12, 2009 at 9:14 AM, Andy Allan  wrote:
> It's a fairly well established convention that in OSM it's the
> houses/plots, not the road centrelines, that are addressed.

But that doesn't always reflect reality.  The reality, at least in
many parts of the world, is that the streets are given blocks of
potential addresses, and the houses/plots/whatever are given actual
addresses from those potential address blocks.

> I'd say it's better to approximate the gap between the road and the houses
> (10m?) than to just put it on the centreline due to that being easier.

First of all, how would you approximate the gap?  You mean by hand?

Secondly, what if the houses aren't yet there?  Tiger address data
represents *potential* address blocks, not *actual* address blocks.
There may or may not be any actual houses along those roads.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing Question

[Shalabh]

Which somehow reminds of the AND data imported to OSM. I am not sure whether
AND and TIGER had anything to do with each other but most of the highways
from the AND data in India are straight lines, often a couple of hundred
metres off the actual road. I have been deleting old tracks and adding new
tracks. Just to give you an example, refer the below frame.

http://www.openstreetmap.org/?lat=32.1488&lon=76.4172&zoom=13&layers=B000FTF

This road, a curvy mountain road was a straight line even at the highest
zoom level with no waypoints whatsoever till 2-3 days ago. I have deleted 3
such roads in the last week and replaced with the new ones. And I know there
are thousands more to go. I doubt if roads like these actually add any
value, either from a mapping or routing point of view to OSM.

Regards,
Shalabh

On Thu, Nov 12, 2009 at 10:47 PM, Andy Robinson (blackadder-lists) <
ajrli...@googlemail.com> wrote:

> Andy Allan [mailto:gravityst...@gmail.com] wrote:
> >Sent: 12 November 2009 2:15 PM
> >To: Andy Robinson (blackadder-lists)
> >Cc: Ian Dees; OSM Talk; talk...@openstreetmap.org
> >Subject: Re: [OSM-talk] Addressing Question
> >
> >> Ian Dees wrote:
> >>>Hi everyone,
> >>>
> >>>I'm looking at some donated street centerline data that has addressing
> >data
> >>>in the form of "Right/Left From Addr" and "Right/Left To Addr" on each
> >>>street centerline. Is there an accepted way of applying these tags to
> the
> >>>road ways? It doesn't really make very much sense to create and store a
> >>>separate way just for the addressing information.
> >
> >It's a fairly well established convention that in OSM it's the
> >houses/plots, not the road centrelines, that are addressed. I'd say
> >it's better to approximate the gap between the road and the houses
> >(10m?) than to just put it on the centreline due to that being easier.
> >This has precedent already - a couple of areas in the US has Karlsruhe
> >schema addressing converted from what is clearly centreline data to
> >spread the addresses out on either side:
> >
> >http://www.openstreetmap.org/?lat=33.676517&lon=-
> >84.012017&zoom=18&layers=B000FTF
> >
> >Your point about 3 time the number of ways becomes increasing relevant
> >with the decreasing quality of the data you are importing - 3 times
> >the headache of fixing dreadful road geometries would be too much!
> >
> >On Thu, Nov 12, 2009 at 10:42 AM, Andy Robinson (blackadder-lists)
> > wrote:
> >> I'd let someone else work out if they can transcribe the data to another
> >> format once its in OSM, should that be desirable
> >
> >I disagree there. It's much better to put the effort in during the
> >initial import, than to import things badly and try to fix it up
> >later. We've been working on lots of post-import fixups in the last 6
> >months and it's much harder than everyone assumes. The 4 months to
> >remove TIGER node tags is a case in point - it took less time than
> >that to import them!
>
> You make a good point and I certainly wouldn't want to see data imported
> that was either difficult to rework or didn't make logical sense. If its
> been done before to offset left/right data automagically I too would vote
> for that as a preferred import method.
>
> Cheers Andy R
>
>
>
> ___
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk
>
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing Question

[Andy Robinson (blackadder-lists)]

Andy Allan [mailto:gravityst...@gmail.com] wrote:
>Sent: 12 November 2009 2:15 PM
>To: Andy Robinson (blackadder-lists)
>Cc: Ian Dees; OSM Talk; talk...@openstreetmap.org
>Subject: Re: [OSM-talk] Addressing Question
>
>> Ian Dees wrote:
>>>Hi everyone,
>>>
>>>I'm looking at some donated street centerline data that has addressing
>data
>>>in the form of "Right/Left From Addr" and "Right/Left To Addr" on each
>>>street centerline. Is there an accepted way of applying these tags to the
>>>road ways? It doesn't really make very much sense to create and store a
>>>separate way just for the addressing information.
>
>It's a fairly well established convention that in OSM it's the
>houses/plots, not the road centrelines, that are addressed. I'd say
>it's better to approximate the gap between the road and the houses
>(10m?) than to just put it on the centreline due to that being easier.
>This has precedent already - a couple of areas in the US has Karlsruhe
>schema addressing converted from what is clearly centreline data to
>spread the addresses out on either side:
>
>http://www.openstreetmap.org/?lat=33.676517&lon=-
>84.012017&zoom=18&layers=B000FTF
>
>Your point about 3 time the number of ways becomes increasing relevant
>with the decreasing quality of the data you are importing - 3 times
>the headache of fixing dreadful road geometries would be too much!
>
>On Thu, Nov 12, 2009 at 10:42 AM, Andy Robinson (blackadder-lists)
> wrote:
>> I'd let someone else work out if they can transcribe the data to another
>> format once its in OSM, should that be desirable
>
>I disagree there. It's much better to put the effort in during the
>initial import, than to import things badly and try to fix it up
>later. We've been working on lots of post-import fixups in the last 6
>months and it's much harder than everyone assumes. The 4 months to
>remove TIGER node tags is a case in point - it took less time than
>that to import them!

You make a good point and I certainly wouldn't want to see data imported
that was either difficult to rework or didn't make logical sense. If its
been done before to offset left/right data automagically I too would vote
for that as a preferred import method.

Cheers Andy R



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing Question

[Apollinaris Schoell]

On 12 Nov 2009, at 6:14 , Andy Allan wrote:

> I disagree there. It's much better to put the effort in during the
> initial import, than to import things badly and try to fix it up
> later. We've been working on lots of post-import fixups in the last 6
> months and it's much harder than everyone assumes. The 4 months to
> remove TIGER node tags is a case in point - it took less time than
> that to import them!
>

+10
Give everyone a chance to work in a constructive way and don't expect  
others to clean the mess bad import left behind.
No wonder there are only few motivated mappers in US. In Canada they  
do a much better job in integrating the community and don't import  
every shape file blindly just because it's available.



> Cheers,
> Andy


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing Question

[Andy Allan]

> Ian Dees wrote:
>>Hi everyone,
>>
>>I'm looking at some donated street centerline data that has addressing data
>>in the form of "Right/Left From Addr" and "Right/Left To Addr" on each
>>street centerline. Is there an accepted way of applying these tags to the
>>road ways? It doesn't really make very much sense to create and store a
>>separate way just for the addressing information.

It's a fairly well established convention that in OSM it's the
houses/plots, not the road centrelines, that are addressed. I'd say
it's better to approximate the gap between the road and the houses
(10m?) than to just put it on the centreline due to that being easier.
This has precedent already - a couple of areas in the US has Karlsruhe
schema addressing converted from what is clearly centreline data to
spread the addresses out on either side:

http://www.openstreetmap.org/?lat=33.676517&lon=-84.012017&zoom=18&layers=B000FTF

Your point about 3 time the number of ways becomes increasing relevant
with the decreasing quality of the data you are importing - 3 times
the headache of fixing dreadful road geometries would be too much!

On Thu, Nov 12, 2009 at 10:42 AM, Andy Robinson (blackadder-lists)
 wrote:
> I'd let someone else work out if they can transcribe the data to another
> format once its in OSM, should that be desirable

I disagree there. It's much better to put the effort in during the
initial import, than to import things badly and try to fix it up
later. We've been working on lots of post-import fixups in the last 6
months and it's much harder than everyone assumes. The 4 months to
remove TIGER node tags is a case in point - it took less time than
that to import them!

Cheers,
Andy

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing Question

[Andy Robinson (blackadder-lists)]

Ian Dees wrote:
>Sent: 12 November 2009 1:13 AM
>To: OSM Talk; talk...@openstreetmap.org Openstreetmap
>Subject: [OSM-talk] Addressing Question
>
>Hi everyone,
>
>I'm looking at some donated street centerline data that has addressing data
>in the form of "Right/Left From Addr" and "Right/Left To Addr" on each
>street centerline. Is there an accepted way of applying these tags to the
>road ways? It doesn't really make very much sense to create and store a
>separate way just for the addressing information.
>
>...but I might have arrived too late in the argument to say that :-)...

It doesn't have to conform to any existing scheme. Work out the best way to
import it so that it's of use and that's enough. If it can't be shoehorned
into the Karlsruhe schema that isn't a problem. There will be many places
where one-size-fits-all won't work.

I'd let someone else work out if they can transcribe the data to another
format once its in OSM, should that be desirable

Cheers

Andy



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing Question

[Marcus Wolschon]

On Thu, Nov 12, 2009 at 2:13 AM, Ian Dees  wrote:
> Hi everyone,
>
> I'm looking at some donated street centerline data that has addressing data
> in the form of "Right/Left From Addr" and "Right/Left To Addr" on each
> street centerline.

Can you identify the location of the Addr for the "From/To"?
If so it would be easy to calculate interpolation-ways right
and left of the streets using the Karlsruhe Schema.

> Is there an accepted way of applying these tags to the
> road ways?

No, only in applying it to houses next to the street and special
ways connecting next to the street for interpolating whole
streets of house-numbers at once.

> It doesn't really make very much sense to create and store a
> separate way just for the addressing information.

I disagree here because of the hundrets of special cases
that absolutely must be handled to be correct that come
from the fact that houses are not usually build in the middle
of a road.

> ...but I might have arrived too late in the argument to say that :-)...

Yes, you are. ;)

PS:
I`m not reading the talk-US but as you crossposted there, so do I.
Please let me know of yet another addressing-schema comes
that is in actual usage to make sure my address-search does
work on all the planet.

Marcus

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing Question

[andrzej zaborowski]

2009/11/12 Ian Dees :
> What does the math look like to handle intersections? Curvy roads?

I admit that data wasn't complex, all segments were straight and all
nodes were treated as intersections.  Do you have parametric (e.g.
bezier) curves or just lots of short segments making soft turns?  (I'm
optimistic we can figure something out in both cases)

Regards

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing Question

[Ian Dees]

On Wed, Nov 11, 2009 at 9:47 PM, andrzej zaborowski wrote:

>
> That's a pretty pessimistic view.
>
>
Sorry, I am pretty grumpy today. The area I'm looking at actually has quite
a few mappers already, so I imagine this data would probably get updated
quickly.


>
> For the record an import I've done had only this left/right - to/from
> housenumber information, too, so I have an ugly python script here
> ready to throw at this kind of data (after adapting to whatever the
> input format is) and I would be happy to do the processing on my PC if
> you decide to go this way.  The whole toolchain should still behave
> reasonably for data size of TIGER (though obviously I didn't have that
> much data)
>
>
What does the math look like to handle intersections? Curvy roads?
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing Question

[andrzej zaborowski]

2009/11/12 Ian Dees :
> On Wed, Nov 11, 2009 at 7:55 PM, andrzej zaborowski 
> wrote:
>> The hope is that local mappers there will be slowly improving imported
>> data until there are separate points for every address I think?  Then
>> I'd recommend just adding those separate ways and making it easier for
>> the mappers to build on this data, instead of making it harder for
>> nearly everyone dealing with OSM data by adding a whole different
>> addressing scheme.
>
> No, I doubt local mappers will improve the data. I sent this mail because

That's a pretty pessimistic view.

> almost all of the data I've seen available for import in the US (all the way
> from individual municipalities to TIGER's shapefiles) has this
> left/right-from/to scheme for addressing information.
>
> If the expectation is that we will always be following the Karlsruhe schema
> (with separate ways on each side of the road centerline), then importing
> this addressing data will be next to impossible*.

This would mean we are excluding some features, or some tagging
schemes, in some parts of the world due to disk space or processing
time but not in other parts of the world.  I'm sure we will never have
uniform tagging and uniform data quality everywhere but I still want
to aim at it.

For the record an import I've done had only this left/right - to/from
housenumber information, too, so I have an ugly python script here
ready to throw at this kind of data (after adapting to whatever the
input format is) and I would be happy to do the processing on my PC if
you decide to go this way.  The whole toolchain should still behave
reasonably for data size of TIGER (though obviously I didn't have that
much data)

Cheers

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing Question

[Ian Dees]

On Wed, Nov 11, 2009 at 7:55 PM, andrzej zaborowski wrote:

> Hiya,
>
> 2009/11/12 Ian Dees :
> > I'm looking at some donated street centerline data that has addressing
> data
> > in the form of "Right/Left From Addr" and "Right/Left To Addr" on each
> > street centerline. Is there an accepted way of applying these tags to the
> > road ways? It doesn't really make very much sense to create and store a
> > separate way just for the addressing information.
>
> The hope is that local mappers there will be slowly improving imported
> data until there are separate points for every address I think?  Then
> I'd recommend just adding those separate ways and making it easier for
> the mappers to build on this data, instead of making it harder for
> nearly everyone dealing with OSM data by adding a whole different
> addressing scheme.
>

No, I doubt local mappers will improve the data. I sent this mail because
almost all of the data I've seen available for import in the US (all the way
from individual municipalities to TIGER's shapefiles) has this
left/right-from/to scheme for addressing information.

If the expectation is that we will always be following the Karlsruhe schema
(with separate ways on each side of the road centerline), then importing
this addressing data will be next to impossible*.

-Ian

* Ok, not "impossible", but the import size would triple and the CPU time to
compute the new addressing-only ways might make it hard for the "regular
mapper" to do.
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Addressing Question

[andrzej zaborowski]

Hiya,

2009/11/12 Ian Dees :
> I'm looking at some donated street centerline data that has addressing data
> in the form of "Right/Left From Addr" and "Right/Left To Addr" on each
> street centerline. Is there an accepted way of applying these tags to the
> road ways? It doesn't really make very much sense to create and store a
> separate way just for the addressing information.

The hope is that local mappers there will be slowly improving imported
data until there are separate points for every address I think?  Then
I'd recommend just adding those separate ways and making it easier for
the mappers to build on this data, instead of making it harder for
nearly everyone dealing with OSM data by adding a whole different
addressing scheme.

Cheers

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Matias D'Ambrosio]

On Sunday 14 December 2008 17:20:52 Simon Ward wrote:
> On Fri, Dec 12, 2008 at 01:17:09PM +, Dave Stubbs wrote:
> > It's what happens when you hit reply to all. Get used to it -- it's
> > gonna happen to you a lot.
>
> It’s a pity more MUAs don’t have reply to list or take note of the
> List-Post header. ☹
>
 Indeed, Thunderbird 3 supports it, apparently, which should help with a lot 
of OSS-centric lists (other OSS MTUs I know already support it).

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Simon Ward]

On Fri, Dec 12, 2008 at 01:17:09PM +, Dave Stubbs wrote:
> It's what happens when you hit reply to all. Get used to it -- it's
> gonna happen to you a lot.

It’s a pity more MUAs don’t have reply to list or take note of the
List-Post header. ☹

Simon
-- 
A complex system that works is invariably found to have evolved from a
simple system that works.—John Gall


signature.asc
Description: Digital signature
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Matias D'Ambrosio]

On Friday 12 December 2008 17:34:39 Colin McGregor wrote:
> The best arrangement I have seen for street numbers is in Chicago, IL.
> Streets laid out on a grid pattern. On the north/south and east/west
> streets there is a number -> distance link. If you go from say 1 North
> Michigan to 801 North Michigan you will have travelled ~ 1 mile (or if
> you go from 1 to 501 you will have travelled ~ 1 kilometer).
>
> Toronto, Ontario things are almost as good. Almost all north/south
> street are have the lowest number at the most southern part of the
> street (Toronto being unable to go any further south due to Lake
> Ontario). On east/west streets the lowest number is almost always the
> point closest to Yonge St. (a major north/south street). What Toronto
> doesn't have but Chicago does is the link between street numbers and
> distance...
 Ah, both sound very similar to our system. In Argentina, if you want to go to 
the main city square, you just have to follow the numbers to 0, it's not the 
fastest way, but if you're lost... Here, Avenida Colon 2100 would be 2.1km 
from the street's starting point (which is the main square in my city).

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Matias D'Ambrosio]

On Friday 12 December 2008 11:17:09 Dave Stubbs wrote:
> OMG! Off-thread-topic shocker! What is this thread about BTW? I seem
> to have lost track.
 Street number craziness. By now it has degenerated into "look at this neat 
grid" and "look at this crazy swirling random number maze", still useful 
since we get to know wildly different areas (which comes in handy later, 
knowledge always does).

> > PS: I'm subscribed to this list, no need to CC me.
>
> It's what happens when you hit reply to all. Get used to it -- it's
> gonna happen to you a lot.
 I know :) And cars don't brake so I can cross the street when they are 
supposed to, I still yell at them and do my best to modify their incorrect 
conduct.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Colin McGregor]

On 12/12/08, Matias D'Ambrosio  wrote:
> On Friday 12 December 2008 10:23:37 Ed Loach wrote:
>> I just spotted the following (video) news report on the BBC website:
>> http://news.bbc.co.uk/1/hi/uk/7778886.stm
>> I don't believe the featured street has been mapped yet
>> http://www.openstreetmap.org/?lat=52.37471&lon=-2.3144&zoom=15
>> but it looks like to have addressing tags on it each house would
>> need to be added manually.

>  The messiness of UK and other European countries when it comes to numbering
> is old news, is there anything else on that page? I can't watch the video
> (flash video is a broken idea). Obviously, if there is no logical assignment
> of numbers, or one that is broken too often, mapping each number makes
> sense,
> this thread is not about that.

The above video offers what I hope is an extreme example of street
number craziness in the UK (a street with multiple number "2"s,
etc..). I gather there is similar messiness in parts of Asia, like
street numbers assigned based on the age of the building...

The best arrangement I have seen for street numbers is in Chicago, IL.
Streets laid out on a grid pattern. On the north/south and east/west
streets there is a number -> distance link. If you go from say 1 North
Michigan to 801 North Michigan you will have travelled ~ 1 mile (or if
you go from 1 to 501 you will have travelled ~ 1 kilometer).

Toronto, Ontario things are almost as good. Almost all north/south
street are have the lowest number at the most southern part of the
street (Toronto being unable to go any further south due to Lake
Ontario). On east/west streets the lowest number is almost always the
point closest to Yonge St. (a major north/south street). What Toronto
doesn't have but Chicago does is the link between street numbers and
distance...

> PS: I'm subscribed to this list, no need to CC me.

Colin McGregor

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Dave Stubbs]

2008/12/12 Matias D'Ambrosio :
> On Friday 12 December 2008 10:23:37 Ed Loach wrote:
>> I just spotted the following (video) news report on the BBC website:
>> http://news.bbc.co.uk/1/hi/uk/7778886.stm
>> I don't believe the featured street has been mapped yet
>> http://www.openstreetmap.org/?lat=52.37471&lon=-2.3144&zoom=15
>> but it looks like to have addressing tags on it each house would
>> need to be added manually.
>  The messiness of UK and other European countries when it comes to numbering
> is old news, is there anything else on that page? I can't watch the video
> (flash video is a broken idea). Obviously, if there is no logical assignment
> of numbers, or one that is broken too often, mapping each number makes sense,
> this thread is not about that.


OMG! Off-thread-topic shocker! What is this thread about BTW? I seem
to have lost track.


>
> PS: I'm subscribed to this list, no need to CC me.

It's what happens when you hit reply to all. Get used to it -- it's
gonna happen to you a lot.

Dave

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Matias D'Ambrosio]

On Friday 12 December 2008 10:23:37 Ed Loach wrote:
> I just spotted the following (video) news report on the BBC website:
> http://news.bbc.co.uk/1/hi/uk/7778886.stm
> I don't believe the featured street has been mapped yet
> http://www.openstreetmap.org/?lat=52.37471&lon=-2.3144&zoom=15
> but it looks like to have addressing tags on it each house would
> need to be added manually.
 The messiness of UK and other European countries when it comes to numbering 
is old news, is there anything else on that page? I can't watch the video 
(flash video is a broken idea). Obviously, if there is no logical assignment 
of numbers, or one that is broken too often, mapping each number makes sense, 
this thread is not about that.

PS: I'm subscribed to this list, no need to CC me.


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Ed Loach]

I just spotted the following (video) news report on the BBC website:
http://news.bbc.co.uk/1/hi/uk/7778886.stm
I don't believe the featured street has been mapped yet
http://www.openstreetmap.org/?lat=52.37471&lon=-2.3144&zoom=15
but it looks like to have addressing tags on it each house would
need to be added manually.

Ed




___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Matias D'Ambrosio]

On Thursday 11 December 2008 07:05:58 you wrote:
> On a more serious note about addressing, the Karlsruhe Schema would
> still fit your scenario. You can interpolate along each block, and
> assume that there is no guarantee that every building number exists.
> Then for extra detail, you can mark every house in. I don't see any
> reason why it wouldn't work.
 It would work, it's just a lot of unneeded work. Sort of like using assembler 
to get a directory listing when you can just use 'ls'. It's hard enough to 
get mappers (I did 90% of the work in that area on my own), and the karlsruhe 
schema isn't a good fit with the use of the data, either, since people use it 
as a grid with a proper numbering scheme. So it's more work to input the data 
and to process it later. There are people already (myself included) working 
on rendering and routing taking advantage of numbers (without which routing 
is not useful here).

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Shaun McDonald]

On 10 Dec 2008, at 23:04, Matias D'Ambrosio wrote:

> On Wednesday 10 December 2008 15:12:46 Shaun McDonald wrote:
>>
>> South Bridge in Edinburgh has the numbers run up one side, then back
>> down the other.
> I just looked at Edinburgh and all I can say is, I'm sorry! Looks  
> like a mess
> to map, excellent work :)

I'm sure the people who have been mapping Edinburgh and similar such  
cities will be happy to hear that.

>
>
>> I find it very strange that you city blocks are so consistently 100
>> metres by 100 metres. In my experience they are some random size.
>> Consistent block sizes and grid road layouts are just so strange.  
>> It's
>> trunk and branch cul-de-sacs that are the foray of the planners.
> I guess I just want you to hate me a little, here is my city, we often
> complain about the poor quality of the grid:
>
> http://www.openstreetmap.org/?lat=-38.7142&lon=-62.2544&zoom=13&layers=0B00FTF
>

I suppose I'm that used to wiggly roads that it would be strange to do  
some grid based city.

> Pay no attention to the weird mix of primary/secondary/tertiary,  
> we're soon
> having a meeting to improve that.

On a more serious note about addressing, the Karlsruhe Schema would  
still fit your scenario. You can interpolate along each block, and  
assume that there is no guarantee that every building number exists.  
Then for extra detail, you can mark every house in. I don't see any  
reason why it wouldn't work.

Shaun


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Matias D'Ambrosio]

On Wednesday 10 December 2008 15:12:46 Shaun McDonald wrote:
>
> South Bridge in Edinburgh has the numbers run up one side, then back
> down the other.
 I just looked at Edinburgh and all I can say is, I'm sorry! Looks like a mess 
to map, excellent work :)

> I find it very strange that you city blocks are so consistently 100
> metres by 100 metres. In my experience they are some random size.
> Consistent block sizes and grid road layouts are just so strange. It's
> trunk and branch cul-de-sacs that are the foray of the planners.
 I guess I just want you to hate me a little, here is my city, we often 
complain about the poor quality of the grid:

http://www.openstreetmap.org/?lat=-38.7142&lon=-62.2544&zoom=13&layers=0B00FTF

 Pay no attention to the weird mix of primary/secondary/tertiary, we're soon 
having a meeting to improve that.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Matias D'Ambrosio]

On Wednesday 10 December 2008 15:12:46 you wrote:
> South Bridge in Edinburgh has the numbers run up one side, then back
> down the other. I have come across some residential streets that are
> numbered all the way around. It's very rare around here for houses to
> be numbered based on the end of the road. What happens when you extend
> that end of the road?
 No idea, around here they are numbered from the start :) If you go past the 
start of a street, another street starts, so it can always be extended at the 
end.

> I find it very strange that you city blocks are so consistently 100
> metres by 100 metres. In my experience they are some random size.
> Consistent block sizes and grid road layouts are just so strange. It's
> trunk and branch cul-de-sacs that are the foray of the planners.
 Oh, no, that's the ideal, only a few cities are close to perfect. There are 
certainly forks, curves and cul de sacs, but they are the exception, not the 
rule. I have seen areas of Italy for example that don't seem to have a single 
stretch of road that's straight for more than 50 metres.


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Shaun McDonald]

On 10 Dec 2008, at 16:41, Matias D'Ambrosio wrote:


On Wednesday 10 December 2008 12:04:13 you wrote:

In your country, perhaps. In my country, that's EXACTLY what
the address is.
The address is the street and distance from the street's
starting point, in
metres.


What happens if two houses are built facing each other on opposite
sides of the street? Wouldn't they get assigned the same number? Or
does someone arbitrarily decide which is closer and increase or
decrease one of the numbers by 1?
Hmm, strange country you live in if houses are 1 metre wide, houses  
here are
usually 10 metres wide. And each side of the street has even or odd  
numbers,
although unusual (I can't think of any such street at this moment)  
it's

possible to have them mixed.


South Bridge in Edinburgh has the numbers run up one side, then back  
down the other. I have come across some residential streets that are  
numbered all the way around. It's very rare around here for houses to  
be numbered based on the end of the road. What happens when you extend  
that end of the road?



Each house has an assigned number, which I think
can be chosen by the owner (I don't know exactly when it's chosen, and
whether it can be changed, or how easily), that number is within the  
range

the house occupies.
What's also important, is the usage. It is very common to say "Alem  
Avenue
1200" to refer to a corner or the block face that has the range  
1200-1299.
Using "Alem Avenue and San Juan Street" is unusual unless both are  
well known
streets. Each city block is 100 metres by 100 metres, so counting  
blocks is
helpful if there are no signs on the corners and houses don't have  
numbers on
the front (or they are not easily visible from a car), which is not  
uncommon,

in fact I used this method of finding a house just last week. Mighty
convenient, let me tell you :)


I find it very strange that you city blocks are so consistently 100  
metres by 100 metres. In my experience they are some random size.  
Consistent block sizes and grid road layouts are just so strange. It's  
trunk and branch cul-de-sacs that are the foray of the planners.


Shaun



smime.p7s
Description: S/MIME cryptographic signature
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Matias D'Ambrosio]

On Wednesday 10 December 2008 12:04:13 you wrote:
> >  In your country, perhaps. In my country, that's EXACTLY what
> > the address is.
> > The address is the street and distance from the street's
> > starting point, in
> > metres.
>
> What happens if two houses are built facing each other on opposite
> sides of the street? Wouldn't they get assigned the same number? Or
> does someone arbitrarily decide which is closer and increase or
> decrease one of the numbers by 1?
 Hmm, strange country you live in if houses are 1 metre wide, houses here are 
usually 10 metres wide. And each side of the street has even or odd numbers, 
although unusual (I can't think of any such street at this moment) it's 
possible to have them mixed. Each house has an assigned number, which I think 
can be chosen by the owner (I don't know exactly when it's chosen, and 
whether it can be changed, or how easily), that number is within the range 
the house occupies.
 What's also important, is the usage. It is very common to say "Alem Avenue 
1200" to refer to a corner or the block face that has the range 1200-1299. 
Using "Alem Avenue and San Juan Street" is unusual unless both are well known 
streets. Each city block is 100 metres by 100 metres, so counting blocks is 
helpful if there are no signs on the corners and houses don't have numbers on 
the front (or they are not easily visible from a car), which is not uncommon, 
in fact I used this method of finding a house just last week. Mighty 
convenient, let me tell you :)

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[D Tucny]

2008/12/10 Ed Loach <[EMAIL PROTECTED]>

> >  In your country, perhaps. In my country, that's EXACTLY what
> > the address is.
> > The address is the street and distance from the street's
> > starting point, in
> > metres.
>
> What happens if two houses are built facing each other on opposite
> sides of the street? Wouldn't they get assigned the same number? Or
> does someone arbitrarily decide which is closer and increase or
> decrease one of the numbers by 1?
>

With such a structured system, do you think the planning codes would allow
this? I guess they have to be offset by at least 1m...

d
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Ed Loach]

>  In your country, perhaps. In my country, that's EXACTLY what
> the address is.
> The address is the street and distance from the street's
> starting point, in
> metres.

What happens if two houses are built facing each other on opposite
sides of the street? Wouldn't they get assigned the same number? Or
does someone arbitrarily decide which is closer and increase or
decrease one of the numbers by 1?

Ed



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Matias D'Ambrosio]

On Tuesday 09 December 2008 11:37:13 Frederik Ramm wrote:
> Hi,
>
> Matt Amos wrote:
> > sure, editor support isn't 100% yet, but why re-create a poor-man's
> > relations with name-based references, when we already have "proper"
> > relations?
>
> In my eyes an address is not a relation. It comes close, but a house
> *can* have an address that has nothing to do with the road that passes
> it.
 In your country, perhaps. In my country, that's EXACTLY what the address is. 
The address is the street and distance from the street's starting point, in 
metres.

> The address "31 So-and-So Street" does not mean that this is the 
> 31st house on So-and-So Street, it doesn't even necessarily mean that
> the entrance is via So-and-So Street or that it is in the vicinity of
> 30, 32, or 33...
 In my country, which has a very sane and predictable scheme, there is usually 
a central square which is the centre of the city, from which the main streets 
start, and from these start other streets (depending on terrain and proper 
planning it might end up as a perfect grid, cf. La Plata, Argentina). 31 
would in fact be before 32 and after 30.
 There may be other countries using the same/similar system. I believe the USA 
is one? And probably a few other countries in South America. I'm guessing 
being 'new' countries helps keeping things sane, we're still not overcome by 
entropy :)

> I view an address as an individual attribute of a 
> certain property that is often similar to addresses of neighbouring
> properties, but need not be.
 Cultural issue, no doubt. One map, two systems. Kidding :D
 I do think we need more than one system, and people have to stop imposing 
*their* view/system on other countries.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[SteveC]

On 8 Dec 2008, at 23:51, Jochen Topf wrote:

> On Mon, Dec 08, 2008 at 03:37:00PM -0800, SteveC wrote:
>> On 8 Dec 2008, at 13:41, Frederik Ramm wrote:
>>
>>> Hi,
>>>
>>> SteveC wrote:
 "nodes on a way but not a relation" schema needs to have
 addr:street=foo on the addressing way ?
>>>
>>> The original idea was to (by default) assume that the street name is
>>> the "name" tag of the nearest highway (i.e. no addr:* required), and
>>> that you can optionally set addr:street on the node to make it 100%
>>> clear.
>>
>> I don't mean the node, I mean when you have numbers on a way, it  
>> would
>> seem to make sense to make the way have addr:street rather than all
>> the nodes.
>
> Are you talking about the addr:interpolation way?

yes...

> We specifically
> decided to not put any more information on the addr:interpolation-way,
> because that would mean that every software had to look for the
> addr:*-Tags in the interpolation way *and* on the nodes which makes
> things a bit more complex.

Um ok. So I don't see why the nodes have to have addr:street. I'm  
assuming that the way would have addr:street on it, and the nodes  
would have the numbers. I don't see that picking the closest ways to  
match it as a good solution.

Best

Steve


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Matt Amos]

On Tue, Dec 9, 2008 at 3:35 PM, Dave Stubbs <[EMAIL PROTECTED]> wrote:
> Where a block had a single entrance equipped with doorbell/entry
> phone/postboxes I added the wonderful tag addr:flats=1-10 to the
> building. ie: I treated it as a black box, and merely stated what
> numbers were inside. Often these blocks have names or a street number:
> I've been adding these as name= and addr:housenumber=

i did a similar thing where there was a building of flats but the
flats were accessible via different stairwells. e.g: each flat was
flat N, Henry house, coin street. the stairwells i entered as
interpolation lines linked to the building (hey, its a way) and the
building i then linked to the street.

my attempts can be seen here (much less beautiful than dave's)
http://tools.geofabrik.de/osmi/?view=addresses&lon=-0.10960&lat=51.50568&zoom=18

cheers,

matt

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Dave Stubbs]

2008/12/9 Douglas Furlong <[EMAIL PROTECTED]>:
> 2008/12/8 SteveC <[EMAIL PROTECTED]>
>>
>> Some notes on using Karlsruhe schema...
>>
>> "nodes on a way but not a relation" schema needs to have
>> addr:street=foo on the addressing way ?
>>
>> addr:housename needs to be added, or addr:housenumber needs to be
>> explicit that it can be a name, or made a more generic term to
>> encompass both
>>
>> wiki pages for address, addressing, addresses needs to point to the
>> schema
>>
>> needs a simple "how to do addressing in potlatch" video a-la the old
>> ones I did, as if you ignore relations it is essentially trivial
>>
>> I'm happy to do all that to the wiki page, but just wondered if I'm
>> totally off track on any of them?
>
> I'd like to throw a quick question in here, now that this is being
> discussed, and people with a much better idea of tagging and related issues
> are reviewing the thread :)
>
> I live in an area of London that has a massive number of residential tower
> blocks, some having 10 flats, others having over 100.
>
> What would be the best way of putting in these house numbers, so that they
> can be mapped?
>
> Linked are two examples where i've put address data in, so that if nothing
> else it is stored, but it's not being drawn which is a pity, and problem
> wouldn't be brought up on a search.
>
> http://openstreetmap.org/?lat=51.478197&lon=-0.110314&zoom=18&layers=B000FTF
> (you'll need to view the data, as they buildings are not rendered).
> http://openstreetmap.org/?lat=51.53264&lon=-0.09008&zoom=17&layers=B000FTFT
>
> I'd love to be able to put in the buildings and where possible easily map
> the house number ranges.
>
> Any idea's?


I was having similar problems in Putney. One of the first observation
of the Karslruhe Schema is that it doesn't really handle anything
finer than street level addresses.
First up, just add the building=yes tag to get them to render. It's
not cheating, they are buildings :-)

Where a block had a single entrance equipped with doorbell/entry
phone/postboxes I added the wonderful tag addr:flats=1-10 to the
building. ie: I treated it as a black box, and merely stated what
numbers were inside. Often these blocks have names or a street number:
I've been adding these as name= and addr:housenumber=

Where a building had more than one number, i.e. 12-20 Camomile Street,
but that's just because they knocked down 12-20 to build the new
building, then I added it as addr:housenumber=12-20. You usually find
something like 12-20 stuck on the front of the building, and it
doesn't really represent house numbers 12 to 20, but rather the number
"12-20". Maybe this should actually be under addr:housename instead.
meh.

Where a building contains flats with street addresses I added lots of
nodes (or interpolation nodes) in the right place (ish) with
addr:floor=0 / 1 / 2 etc to indicate what floor (although that's
rarely part of the address as such) on top of the building, and
unconnected to it in any way. Or I got bored and added them as more
black boxes like above.

Where you just have a row of houses all in the same "building", i.e.
terraces or their 1960s concrete equivalent then I just used
interpolation lines as on the wiki page.

To see my attempts in all their glory have a look at:
http://tools.geofabrik.de/osmi/?view=addresses&lon=-0.21376&lat=51.45866&zoom=17

Dave

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing - Tower blocks

[D Tucny]

2008/12/9 Douglas Furlong <[EMAIL PROTECTED]>

> 2008/12/8 SteveC <[EMAIL PROTECTED]>
>
>>
>> I'd like to throw a quick question in here, now that this is being
> discussed, and people with a much better idea of tagging and related issues
> are reviewing the thread :)
>
> I live in an area of London that has a massive number of residential tower
> blocks, some having 10 flats, others having over 100.
>
> What would be the best way of putting in these house numbers, so that they
> can be mapped?
>
> Linked are two examples where i've put address data in, so that if nothing
> else it is stored, but it's not being drawn which is a pity, and problem
> wouldn't be brought up on a search.
>
>
> http://openstreetmap.org/?lat=51.478197&lon=-0.110314&zoom=18&layers=B000FTF(you'll
>  need to view the data, as they buildings are not rendered).
> http://openstreetmap.org/?lat=51.53264&lon=-0.09008&zoom=17&layers=B000FTFT
>
> I'd love to be able to put in the buildings and where possible easily map
> the house number ranges.
>
> Any idea's?
>

You'll see that some buildings in that view are showing up... they key
difference between the ones that are showing up and the ones that aren't is
that the ones that are have the building=yes tag... the others don't...
Adding the building tag would allow them to be rendered and recognised as
buildings... In the osmarender layer the numbers would even show up...

The numbering however would be an issue with those there...

The interpolation scheme is specified as being a linear way containing at
least two nodes with numbers... So, for example, on the building that has 1
- 21 Mandela street, the building could be tagged as it is now, but without
the interpolation tag... This should then be rendered in osmarender as 1 -
21 over a solid circle...

Alternatively, if each of the numbers is reachable from the street from one
end of the building to the other on the side facing Mandela Street, you
could create an additional way over that edge of the building containing 2
nodes, 1 at each end, the one at the lowest numbered end with a housenumber
of 1, the one at the highest end with a housenumber of 21 then the way
itself having the interpolation tag set to all... What would be rendered by
osmarender then would be a 1 at one end and 21 at the other with a line,
potentially dotted or dashed, not sure, joining them... A tool to locate an
address should in theory then be able to guestimate how far along the way a
specific number is using this...

The final (best?) approach would be to tag each individual house number by
it's entrance, so if this building has 7 entrances with 3 flats per entrance
you could add a node for each entrance to the building, tagging the first
entrance as 1-3, the second as 4-6 etc...

This should all be fine for places that have street addresses, but, I'm
struggling with tower blocks or other buildings that are part of private
estates at the moment which I've posted on the wiki about here...
http://wiki.openstreetmap.org/wiki/Talk:Proposed_features/House_numbers/Karlsruhe_Schema#Apartment_Blocks.2C_Blocks_of_Flats.2C_Communities.2C_Business_Parks.2C_Campuses_etc...

Anyway... hope that helps...

d
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Douglas Furlong]

2008/12/8 SteveC <[EMAIL PROTECTED]>

> Some notes on using Karlsruhe schema...
>
> "nodes on a way but not a relation" schema needs to have
> addr:street=foo on the addressing way ?
>
> addr:housename needs to be added, or addr:housenumber needs to be
> explicit that it can be a name, or made a more generic term to
> encompass both
>
> wiki pages for address, addressing, addresses needs to point to the
> schema
>
> needs a simple "how to do addressing in potlatch" video a-la the old
> ones I did, as if you ignore relations it is essentially trivial
>
> I'm happy to do all that to the wiki page, but just wondered if I'm
> totally off track on any of them?
>

I'd like to throw a quick question in here, now that this is being
discussed, and people with a much better idea of tagging and related issues
are reviewing the thread :)

I live in an area of London that has a massive number of residential tower
blocks, some having 10 flats, others having over 100.

What would be the best way of putting in these house numbers, so that they
can be mapped?

Linked are two examples where i've put address data in, so that if nothing
else it is stored, but it's not being drawn which is a pity, and problem
wouldn't be brought up on a search.

http://openstreetmap.org/?lat=51.478197&lon=-0.110314&zoom=18&layers=B000FTF(you'll
need to view the data, as they buildings are not rendered).
http://openstreetmap.org/?lat=51.53264&lon=-0.09008&zoom=17&layers=B000FTFT

I'd love to be able to put in the buildings and where possible easily map
the house number ranges.

Any idea's?
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Ed Loach]

> i prefer using relations, but clearly both methods have merits.
> how
> hard is it to support both?

Having read this discussion I know of a location locally where both
methods are probably required. There is a road called "New Way"
where the houses on one side actually have postal addresses of
"Eastern Promenade", although there is no vehicular access to the
"front" of these houses. Tagging the nodes with addr:street=Eastern
Promenade, but then using a relation to link the nodes with "New
Way" seems the perfect solution. At present I've only added the way
with a note about the properties on one side having a different
street address, rather than actually marking any of them.

Ed



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Ben Laenen]

On Tuesday 09 December 2008, Christoph Eckert wrote:
> Hi,
>
> > But meanwhile - mostly because it is so easy to do in JOSM - many
> > people simply tag the whole shebang (addr:country, addr:town,
> > addr:post_code, addr:street, addr:house_number) onto every house
> > node.
>
> though I attended the workshop where we developed the scheme, I think
> we meanwhile could do better. Using the proposal of a street
> relation[1], we could easily make linking housenames to streets much
> more robust and reliable. Just add the nodes with the house names to
> those relations. Counting them in the planet file shows a significant
> amount: bzcat relations-latest.osm.bz2 | grep -i 'tag k="type"
> v="street"' | wc -l 1767
>
> I just posted an example:
> http://www.christeck.de/wp/?p=131
>
> I also put it as a proposal to the wiki. If it's not appreciated,
> feel free to edit it[2] :) .

+1 for the type=street relation. It's badly needed, but let's keep it 
separate from house number proposals for now. There are reasons enough 
why a street relation is useful apart from house numbers:

* name description: not really necessary data, but fun to map anyway: 
extra information written on a street sign telling where the street 
name comes from (who the person was, or what it means). You just can't 
add that to each highway that belongs to the street when it exists out 
of a dozen ways.

* other data related to the street, like pointers to decisions made by 
the municipal government with traffic rules for that road etc. (which 
could help in keeping things updated later on)

(so basically that means losing data duplication, but it of course means 
that renderers need an update to handle it if we also put the 
streetname in there as well and no longer in the highway=X ways).

* you can add all nodes with things that belong to that street to the 
relation (shops, bus stops, etc.) and there's no more ambiguity in some 
cases in which street something is.

And these house numbers are just another reason (at least Karlsruhe 
scheme). But I must say I really dislike the name "associatedStreet" as 
is currently written down in that proposal. It looks pretty bad to me 
as it's by name not universal enough since it can be used for much 
more. Just add it to the street relation and be done with it.

Ben

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Jochen Topf]

On Tue, Dec 09, 2008 at 03:00:04PM +0100, Frederik Ramm wrote:
> Matt Amos wrote:
> > exactly - so why say that the nearest road with a particular name is
> > the addressable road? better to indicate exactly which way is the
> > addressing element and leave routing for the "via" elements.
> 
> No, not "indicate which way is the addressing element" (because the 
> object may even have a street name without the street existing in OSM) - 
> simply put the street name with the object and that's it. In my eyes, 
> the address is a name, not a pointer to an object that has "addressing 
> quality".

The Czech Republic seems to be one of the busiest users of the Karlsruhe
model. They seem to have imported house numbers from somewhere as there
are many cases where there are house number nodes, but the corresponding
street is missing.

Jochen
-- 
Jochen Topf  [EMAIL PROTECTED]  http://www.remote.org/jochen/  +49-721-388298


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Matt Amos]

On Tue, Dec 9, 2008 at 1:41 PM, Frederik Ramm <[EMAIL PROTECTED]> wrote:
> Matt Amos wrote:
>> except when it doesn't - e.g: misspelled streets, deleted "in use"
>> streets, etc...
>
> A little redundancy doesn't hurt - on the contrary, it makes spotting
> mistakes easier.

and prevention is better than a cure. don't spot the mistakes - use
the API's features to help prevent their occurrence.

> And about deleted "in use" streets: If a house has a
> certain address then it has that address, even if the street which used to
> pass by the house is physically removed. There is no automatism in the real
> world that links house addresses to streets, so why should there be in OSM?
> If you remove the road next to a house node with an address, then the
> address will of course remain unchanged until this is done explicitly...
> just like in the real world.

there are two uses for addressing: navigation and geo-location. the
situation you describe would make navigation very difficult, so i
think it does not happen much in the real world.

"oh hais, my house address is so-and-so street, but that doesn't exist
any more. please go via such-and-such road."?

the two tagging methods are essentially the same (i.e: they both link
elements to a way - and not necessarily the closest way) and result,
after parsing, in essentially the same data model. the fundamental
differences are:

1) the relatedStreet cannot refer to streets that do not exist and
will prevent anyone attempting (possibly accidentally) to do this.
2) the addr:street represents address information textually, and
therefore requires searching (which may fail, or may result in
unintended items found) to lookup the street.

i prefer using relations, but clearly both methods have merits. how
hard is it to support both?

cheers,

matt

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Frederik Ramm]

Hi,

Matt Amos wrote:
> exactly - so why say that the nearest road with a particular name is
> the addressable road? better to indicate exactly which way is the
> addressing element and leave routing for the "via" elements.

No, not "indicate which way is the addressing element" (because the 
object may even have a street name without the street existing in OSM) - 
simply put the street name with the object and that's it. In my eyes, 
the address is a name, not a pointer to an object that has "addressing 
quality".

Bye
Frederik


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Andy Robinson (blackadder-lists)]

Frederik Ramm wrote:
>Sent: 09 December 2008 1:37 PM
>To: Matt Amos
>Cc: talk@openstreetmap.org
>Subject: Re: [OSM-talk] addressing
>
>Hi,
>
>Matt Amos wrote:
>> sure, editor support isn't 100% yet, but why re-create a poor-man's
>> relations with name-based references, when we already have "proper"
>> relations?
>
>In my eyes an address is not a relation. It comes close, but a house
>*can* have an address that has nothing to do with the road that passes
>it. The address "31 So-and-So Street" does not mean that this is the
>31st house on So-and-So Street, it doesn't even necessarily mean that
>the entrance is via So-and-So Street or that it is in the vicinity of
>30, 32, or 33... I view an address as an individual attribute of a
>certain property that is often similar to addresses of neighbouring
>properties, but need not be.

I've spotted that in the UK the street name really isn't anything to do with
the street per se. In many cases the street sign only appears at the
start/end of a run of properties. Historically it was a way of relating the
property to a location. SO taking that a step further its really the street
that should be related to the property and not the other way around.

Cheers

Andy


>
>Bye
>Frederik
>
>
>___
>talk mailing list
>talk@openstreetmap.org
>http://lists.openstreetmap.org/listinfo/talk
>
>No virus found in this incoming message.
>Checked by AVG - http://www.avg.com
>Version: 8.0.176 / Virus Database: 270.9.15/1838 - Release Date: 08/12/2008
>6:16 PM


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Matt Amos]

On Tue, Dec 9, 2008 at 1:37 PM, Frederik Ramm <[EMAIL PROTECTED]> wrote:
> Matt Amos wrote:
>> sure, editor support isn't 100% yet, but why re-create a poor-man's
>> relations with name-based references, when we already have "proper"
>> relations?
>
> In my eyes an address is not a relation. It comes close, but a house *can*
> have an address that has nothing to do with the road that passes it.

exactly - so why say that the nearest road with a particular name is
the addressable road? better to indicate exactly which way is the
addressing element and leave routing for the "via" elements.

> The
> address "31 So-and-So Street" does not mean that this is the 31st house on
> So-and-So Street, it doesn't even necessarily mean that the entrance is via
> So-and-So Street or that it is in the vicinity of 30, 32, or 33... I view an
> address as an individual attribute of a certain property that is often
> similar to addresses of neighbouring properties, but need not be.

this is not an argument against using relations. this is only an
argument against assuming that the nearest street is the addressing
street, which i am not advocating.

cheers,

matt

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Frederik Ramm]

Hi,

Matt Amos wrote:
> except when it doesn't - e.g: misspelled streets, deleted "in use"
> streets, etc...

A little redundancy doesn't hurt - on the contrary, it makes spotting 
mistakes easier. And about deleted "in use" streets: If a house has a 
certain address then it has that address, even if the street which used 
to pass by the house is physically removed. There is no automatism in 
the real world that links house addresses to streets, so why should 
there be in OSM? If you remove the road next to a house node with an 
address, then the address will of course remain unchanged until this is 
done explicitly... just like in the real world.

Bye
Frederik


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Frederik Ramm]

Hi,

Matt Amos wrote:
> sure, editor support isn't 100% yet, but why re-create a poor-man's
> relations with name-based references, when we already have "proper"
> relations?

In my eyes an address is not a relation. It comes close, but a house 
*can* have an address that has nothing to do with the road that passes 
it. The address "31 So-and-So Street" does not mean that this is the 
31st house on So-and-So Street, it doesn't even necessarily mean that 
the entrance is via So-and-So Street or that it is in the vicinity of 
30, 32, or 33... I view an address as an individual attribute of a 
certain property that is often similar to addresses of neighbouring 
properties, but need not be.

Bye
Frederik


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Dave Stubbs]

2008/12/9 Elena of Valhalla <[EMAIL PROTECTED]>:
> On Tue, Dec 9, 2008 at 12:45 PM, Dave Stubbs <[EMAIL PROTECTED]> wrote:
>>> sure, editor support isn't 100% yet, but why re-create a poor-man's
>>> relations with name-based references, when we already have "proper"
>>> relations?
>> Because editor support is almost non-existent, and exceptionally confusing 
>> :-)
>
> then we should fix the editor(s), not add poor data
>

It's not poor data.
It's a different data model. If you want to argue the data model is
inferior then you have a whole range of good arguments to chose from.

Dave

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Dave Stubbs]

>
> addr:street=Foo on an addressable element is basically the same thing
> as route=LCN:4 on a way. and i thought it was well understood why
> using this method for routes is A Bad Idea, even if it is easier.


Actually that has far more to do with having bus 37, bus 337, bus 270,
bus 44, bus N44, ncn 4, ncn 5, lcn 37, lcn 3, lcn 5, lcn 25 all
sharing the same way (just a random example, don't try to find it!).
Cramming all of that into one tag would become an exercise in text
parsing, and some of the solutions on offer involved building turing
complete languages into the format. I don't really consider it the
same as addr:street which mainly has issues of integrity and
duplication (with maybe the added bonus of having to find the street
object), neither of which were so much of a problem with routes.

Anyway, the primary reason people weren't using relations (when they
actually existed of course) was that editor support sucked. Hence why
the relation handling in Potlatch came in... it's no accident it's
geared primarily for adding routes.
Before that ncn, rcn, lcn tags were generally the way to go for
bicycle routes. And if you still find it easier then I say go for it.
It's a wiki, someone else'll fix it :-)

Dave

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Matt Amos]

On Tue, Dec 9, 2008 at 11:45 AM, Dave Stubbs <[EMAIL PROTECTED]> wrote:
>> sure, editor support isn't 100% yet, but why re-create a poor-man's
>> relations with name-based references, when we already have "proper"
>> relations?
>
> Because editor support is almost non-existent, and exceptionally confusing :-)

so lets fix the real problem! i'm sure everyone would like better
relation support in all editors :-)

> The advantage of using addr:street is that it does just work.

except when it doesn't - e.g: misspelled streets, deleted "in use"
streets, etc...

> The only
> problem being that JOSM isn't clever enough at the moment to
> autocomplete the field based on existing streets in the area. It
> actually should be a drop down combo box letting you select the
> relevant street. And it's entirely possible to cleverly rename local
> objects' addr:street when the street name changes. Of course once
> you've gone to all that trouble you might as well have made the
> backend use a relation instead.

exactly! so why construct such a relation internally, where only JOSM
can use it?

> I like the clean relation data model, but find the addr:street thing
> much easier at the moment.

it is much easier at the moment, and supported by OSM inspector,
etc... which makes relatedStreet a hard sell, but i think it really is
the best way to do it.

addr:street=Foo on an addressable element is basically the same thing
as route=LCN:4 on a way. and i thought it was well understood why
using this method for routes is A Bad Idea, even if it is easier.

cheers,

matt

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Elena of Valhalla]

On Tue, Dec 9, 2008 at 12:45 PM, Dave Stubbs <[EMAIL PROTECTED]> wrote:
>> sure, editor support isn't 100% yet, but why re-create a poor-man's
>> relations with name-based references, when we already have "proper"
>> relations?
> Because editor support is almost non-existent, and exceptionally confusing :-)

then we should fix the editor(s), not add poor data

> The advantage of using addr:street is that it does just work. The only
> problem being that JOSM isn't clever enough at the moment to
> [...] Of course once
> you've gone to all that trouble you might as well have made the
> backend use a relation instead.

that's exactly the point: the addr:street solution would need quite
some ad hoc work in the editors to become really feasible for ease of
adding, maintenance and robustness; the relation solution would need
some work in the editors that will benefit also when working with
other relations

-- 
Elena of Valhalla

homepage: http://www.trueelena.org
email: [EMAIL PROTECTED]

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Richard Fairhurst]

Dave Stubbs wrote:
> I like the clean relation data model, but find the addr:street 
> thing much easier at the moment.
> Maybe I'll add an addressing button (sorry, I meant obscure 
> keypress :-P) to Potlatch sometime then there'll be no excuses left.

When this accursed API 0.6 is out of the way, I'll be (finally) starting on
the new tagging panels and stuff for Potlatch so you won't need any obscure
keypresses. Well, ok, not many.

cheers
Richard
-- 
View this message in context: 
http://www.nabble.com/addressing-tp20903650p20913252.html
Sent from the OpenStreetMap - General mailing list archive at Nabble.com.


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Elena of Valhalla]

On Tue, Dec 9, 2008 at 12:48 PM, Frederik Ramm <[EMAIL PROTECTED]> wrote:
> [...]
> Congratulations on constructing something that is likely to break as soon as
> the first inexperienced mapper touches it ;-) it would have been perfectly
> ok to just add nodes for the individual houses and tag them with the full
> address data - no relations, no headache, and easy to understand for
> everyone.

a name based approach would be as likely to break, and it wouldn't
even give any warning like the relation approach does

>> An even better alternative would probably be to add the collected
>> street relation to the associatedStreet one, but I'm not sure there is
>> support for relations in relations in the api / editors
> JOSM does support relations within relations but there is still a bug that
> causes problems if both the containing and the contained relation are
> created in the same session.

that will have to wait, then, I suppose

> Still my advice is not to use relations wherever there is an easier way.

Actually, I don't understand what's hard with the relations: sure it
is painful to add them with no autocompletion in josm, but that's just
a minor editor issue that will be hopefully fixed.

Maybe it is because I started mapping with the 5 api already in use,
so relations were just a fact and not something new that I had to
learn and changed the way I was used to map.

>  One
> set of address tags per address, nothing could be easier, no relations
> required.

easier? maybe, but a maintenance nightmare

the street I did yesterday was just a small test, but the next one I'm
going to try is a 200+ houses one: i'm not going to do it at once, so
I can't just insert all of the data and be done with it, I would have
to painfully copy it at least once per session, hoping I'm not adding
some spelling error

-- 
Elena of Valhalla

homepage: http://www.trueelena.org
email: [EMAIL PROTECTED]

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Dave Stubbs]

> sure, editor support isn't 100% yet, but why re-create a poor-man's
> relations with name-based references, when we already have "proper"
> relations?


Because editor support is almost non-existent, and exceptionally confusing :-)

The advantage of using addr:street is that it does just work. The only
problem being that JOSM isn't clever enough at the moment to
autocomplete the field based on existing streets in the area. It
actually should be a drop down combo box letting you select the
relevant street. And it's entirely possible to cleverly rename local
objects' addr:street when the street name changes. Of course once
you've gone to all that trouble you might as well have made the
backend use a relation instead.

I like the clean relation data model, but find the addr:street thing
much easier at the moment.
Maybe I'll add an addressing button (sorry, I meant obscure keypress
:-P) to Potlatch sometime then there'll be no excuses left.

Dave

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Matt Amos]

On Tue, Dec 9, 2008 at 11:48 AM, Frederik Ramm <[EMAIL PROTECTED]> wrote:
>> An even better alternative would probably be to add the collected
>> street relation to the associatedStreet one, but I'm not sure there is
>> support for relations in relations in the api / editors
>
> Still my advice is not to use relations wherever there is an easier way.
> One set of address tags per address, nothing could be easier, no
> relations required.

of the three people i know who've tried out the karlsruhe schema, all
very experienced mappers, one added an incorrect addr:street tag which
did not match the name of the street it was supposed to. it turned out
that the street signs were punctuated differently at each end of the
street. this is common enough in the UK that i've seen it many times
and, if it sometimes bites experienced mappers, i expect it will bite
inexperienced mappers harder and more often.

lets say there are 30 houses along a street (all nodes) and the way
name needs to be changed (either because it really changed, or because
it was misspelled in the first place). if there is a relation for the
addressing, the user needs to alter the name tag on the way. if
addr:street is used, the user needs to further select all the houses,
check that all 30 nodes are present and all have the same addr:street
tag and update them.

this means at least 31 objects are updated, whereas only one needs be
updated for the relatedStreet relation.

furthermore: with the name-based relation, if the user deletes the
existing way and re-draws it with a different name, everything
silently breaks. with the relation method the user is informed that
that street is "in use", which (hopefully) will prevent accidental
breakage.

sure, editor support isn't 100% yet, but why re-create a poor-man's
relations with name-based references, when we already have "proper"
relations?

cheers,

matt

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Richard Fairhurst]

SteveC wrote:
> needs a simple "how to do addressing in potlatch" video a-la 
> the old ones I did, as if you ignore relations it is essentially trivial

I'll be committing some new presets next week with all the fields ready and
waiting.

cheers
Richard
-- 
View this message in context: 
http://www.nabble.com/addressing-tp20903650p20912791.html
Sent from the OpenStreetMap - General mailing list archive at Nabble.com.


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Frederik Ramm]

Hi,

> that sounds elegant and solves the problem I had yesterday when trying
> to add some street numbers to a street made of many small ways that
> branch from the main one, i.e. something like this:
> 
>   | |
>   |  |  |
> ---+--++-+--
>   |
> 
> there is now at least one such relation on the map :)

Congratulations on constructing something that is likely to break as 
soon as the first inexperienced mapper touches it ;-) it would have been 
perfectly ok to just add nodes for the individual houses and tag them 
with the full address data - no relations, no headache, and easy to 
understand for everyone.

> An even better alternative would probably be to add the collected
> street relation to the associatedStreet one, but I'm not sure there is
> support for relations in relations in the api / editors

JOSM does support relations within relations but there is still a bug 
that causes problems if both the containing and the contained relation 
are created in the same session.

Still my advice is not to use relations wherever there is an easier way. 
One set of address tags per address, nothing could be easier, no 
relations required.

Bye
Frederik


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Jochen Topf]

On Tue, Dec 09, 2008 at 01:07:16AM +0100, Christoph Eckert wrote:
> Counting them in the planet file shows a significant amount:
> bzcat relations-latest.osm.bz2 | grep -i 'tag k="type" v="street"' | wc -l 
> 1767

Actually thats totally insignificant compared to over 400.000 house
numbers in the whole of Europe tagged without relations.

Jochen
-- 
Jochen Topf  [EMAIL PROTECTED]  http://www.remote.org/jochen/  +49-721-388298


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Jochen Topf]

On Mon, Dec 08, 2008 at 03:37:00PM -0800, SteveC wrote:
> On 8 Dec 2008, at 13:41, Frederik Ramm wrote:
> 
> > Hi,
> >
> > SteveC wrote:
> >> "nodes on a way but not a relation" schema needs to have   
> >> addr:street=foo on the addressing way ?
> >
> > The original idea was to (by default) assume that the street name is  
> > the "name" tag of the nearest highway (i.e. no addr:* required), and  
> > that you can optionally set addr:street on the node to make it 100%  
> > clear.
> 
> I don't mean the node, I mean when you have numbers on a way, it would  
> seem to make sense to make the way have addr:street rather than all  
> the nodes.

Are you talking about the addr:interpolation way? We specifically
decided to not put any more information on the addr:interpolation-way,
because that would mean that every software had to look for the
addr:*-Tags in the interpolation way *and* on the nodes which makes
things a bit more complex. Also there are more cases where this can go
wrong if there is conflicting information. And everybody who will not
do the interpolation anway can just ignore those interpolation ways
alltogether.

Jochen
-- 
Jochen Topf  [EMAIL PROTECTED]  http://www.remote.org/jochen/  +49-721-388298


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Elena of Valhalla]

On Tue, Dec 9, 2008 at 1:07 AM, Christoph Eckert <[EMAIL PROTECTED]> wrote:
> though I attended the workshop where we developed the scheme, I think we
> meanwhile could do better. Using the proposal of a street relation[1], we
> could easily make linking housenames to streets much more robust and
> reliable. Just add the nodes with the house names to those relations.

that sounds elegant and solves the problem I had yesterday when trying
to add some street numbers to a street made of many small ways that
branch from the main one, i.e. something like this:

  | |
  |  |  |
---+--++-+--
  |

there is now at least one such relation on the map :)

http://openstreetmap.org/?lat=45.79049&lon=8.85152&zoom=16&layers=0B00FTF

An even better alternative would probably be to add the collected
street relation to the associatedStreet one, but I'm not sure there is
support for relations in relations in the api / editors

-- 
Elena of Valhalla

homepage: http://www.trueelena.org
email: [EMAIL PROTECTED]

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[SteveC]

On 8 Dec 2008, at 13:41, Frederik Ramm wrote:

> Hi,
>
> SteveC wrote:
>> "nodes on a way but not a relation" schema needs to have   
>> addr:street=foo on the addressing way ?
>
> The original idea was to (by default) assume that the street name is  
> the "name" tag of the nearest highway (i.e. no addr:* required), and  
> that you can optionally set addr:street on the node to make it 100%  
> clear.

I don't mean the node, I mean when you have numbers on a way, it would  
seem to make sense to make the way have addr:street rather than all  
the nodes.

Either way, I don't think magically figuring out what street some  
nodes or ways housenumbers belong to should be left to closeness, it  
doesn't off the top of my head strike me as something that's  
computationally  trivial - lots of edge cases. I figure a relation is  
nice, and requiring addr:street is a good lesser option. not purring  
addr:street should be discouraged.

> But meanwhile - mostly because it is so easy to do in JOSM - many  
> people simply tag the whole shebang (addr:country, addr:town,  
> addr:post_code, addr:street, addr:house_number) onto every house  
> node. It adds redundancy but that is not necessarily bad - gives the  
> OSM inspector a chance to find fishy corners, and makes everything  
> really simple for anybody dealing with the data.
>
>> addr:housename needs to be added, or addr:housenumber needs to be   
>> explicit that it can be a name, or made a more generic term to   
>> encompass both
>
> The JOSM presets already supply addr:housename even though (to my  
> surprise) the wiki page doesn't list it.

ah cool

Best

Steve


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Frederik Ramm]

Hi,

Christoph Eckert wrote:
> Counting them in the planet file shows a significant amount:
> bzcat relations-latest.osm.bz2 | grep -i 'tag k="type" v="street"' | wc -l 
> 1767

Many of these do not contain a single node; they are just "superway" 
relations.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail [EMAIL PROTECTED]  ##  N49°00'09" E008°23'33"

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Christoph Eckert]

Hi,

> But meanwhile - mostly because it is so easy to do in JOSM - many people
> simply tag the whole shebang (addr:country, addr:town, addr:post_code,
> addr:street, addr:house_number) onto every house node.

though I attended the workshop where we developed the scheme, I think we 
meanwhile could do better. Using the proposal of a street relation[1], we 
could easily make linking housenames to streets much more robust and 
reliable. Just add the nodes with the house names to those relations. 
Counting them in the planet file shows a significant amount:
bzcat relations-latest.osm.bz2 | grep -i 'tag k="type" v="street"' | wc -l 
1767

I just posted an example:
http://www.christeck.de/wp/?p=131

I also put it as a proposal to the wiki. If it's not appreciated, feel free to 
edit it[2] :) .


HTH & best regards,

ce


[1] http://wiki.openstreetmap.org/wiki/Relations/Proposed/Collected_Ways
[2] http://wiki.openstreetmap.org/wiki/Proposed_features/
  House_numbers/Karlsruhe_Schema

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] addressing

[Frederik Ramm]

Hi,

SteveC wrote:
> "nodes on a way but not a relation" schema needs to have  
> addr:street=foo on the addressing way ?

The original idea was to (by default) assume that the street name is the 
"name" tag of the nearest highway (i.e. no addr:* required), and that 
you can optionally set addr:street on the node to make it 100% clear.

But meanwhile - mostly because it is so easy to do in JOSM - many people 
simply tag the whole shebang (addr:country, addr:town, addr:post_code, 
addr:street, addr:house_number) onto every house node. It adds 
redundancy but that is not necessarily bad - gives the OSM inspector a 
chance to find fishy corners, and makes everything really simple for 
anybody dealing with the data.

> addr:housename needs to be added, or addr:housenumber needs to be  
> explicit that it can be a name, or made a more generic term to  
> encompass both

The JOSM presets already supply addr:housename even though (to my 
surprise) the wiki page doesn't list it.

Agreed that the Wiki page needs a makeover, ideally by someone who 
writes better English than the assorted band of Germans who did the 
initial version ;-)

Bye
Frederik

-- 
Frederik Ramm  ##  eMail [EMAIL PROTECTED]  ##  N49°00'09" E008°23'33"

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk