I've updated the address data on http://qa.poole.ch/addresses/ch/ to the current GWR address data.
There is one, larger, change from the previous update (mid-2019). As it seems to be a trend to add ancillary buildings now with house numbers in the format/number/*.*/number /I've removed them from both the statistics and the generated files, this removes a good 200'000 addresses from the current dataset (note if somebody wants this data I can generate files with them included).
This, naturally, doesn't address the problem with non-postal house numbers and the different ways of handling them in general, put should limit the effect it has to the places and cantons that are not using a numbering scheme that doesn't differentiate these in format.
Simon
Hallo Simon,
hallo alle,
in Ramsen/SH ist mir folgendes Hausnummernproblem überdeutlich geworden: Die Gemeinde vergibt schon immer die Versicherungsnummer als Hausnummer, was für alle Ortsfremden auf der Suche nach einem bestimmen Haus frustrierend ist.
Ob die übliche, aufsteigende Nummerierung entlang der jeweiligen Strasse - als Zahl mit Punkt vor die Versicherungsnummer gesetzt sowie an den Häusern sichtbar angebracht - das Problem lösen würde, weiss ich nicht.....
Bezieht sich Dein Hinweis unten, Simon, aber vielleicht auf etwas ganz anderes?
mhG Ludwig
On 06.12.19 10:22, Simon Poole wrote:
I've updated the address data on http://qa.poole.ch/addresses/ch/ to the current GWR address data.
There is one, larger, change from the previous update (mid-2019). As it seems to be a trend to add ancillary buildings now with house numbers in the format/number/*.*/number /I've removed them from both the statistics and the generated files, this removes a good 200'000 addresses from the current dataset (note if somebody wants this data I can generate files with them included).
This, naturally, doesn't address the problem with non-postal house numbers and the different ways of handling them in general, put should limit the effect it has to the places and cantons that are not using a numbering scheme that doesn't differentiate these in format.
Simon
talk-ch mailing list talk-ch@openstreetmap.ch http://lists.openstreetmap.ch/mailman/listinfo/talk-ch
Hallo Ludwig
Nein, dass hat damit zu tun, dass je nach Gemeinde auch Nebengebäude im Adressauszug des GWR auftauchen. Wenn wir dies generell nicht vermeiden können (da wir keinen Zugriff auf die entsprechenden Metainformationen haben), so kann man wenigstens die herausfiltern, die eben dieses spezielles Format haben. Da seit Ende 2017 -alle- Gebäude im GWR erfasst werden sollen, gehe ich davon aus, dass das Problem eher zunehmen wird, dies also eine sinnvolle Massnahme ist (ich hab zwar schon ein Problem damit entdeckt bei meinem aktuellen Mappingprojekt, aber dazu später was) .
Simon
Am 06.12.2019 um 12:16 schrieb Ludwig Baumgart:
Hallo Simon,
hallo alle,
in Ramsen/SH ist mir folgendes Hausnummernproblem überdeutlich geworden: Die Gemeinde vergibt schon immer die Versicherungsnummer als Hausnummer, was für alle Ortsfremden auf der Suche nach einem bestimmen Haus frustrierend ist.
Ob die übliche, aufsteigende Nummerierung entlang der jeweiligen Strasse - als Zahl mit Punkt vor die Versicherungsnummer gesetzt sowie an den Häusern sichtbar angebracht - das Problem lösen würde, weiss ich nicht.....
Bezieht sich Dein Hinweis unten, Simon, aber vielleicht auf etwas ganz anderes?
mhG Ludwig
On 06.12.19 10:22, Simon Poole wrote:
I've updated the address data on http://qa.poole.ch/addresses/ch/ to the current GWR address data.
There is one, larger, change from the previous update (mid-2019). As it seems to be a trend to add ancillary buildings now with house numbers in the format/number/*.*/number /I've removed them from both the statistics and the generated files, this removes a good 200'000 addresses from the current dataset (note if somebody wants this data I can generate files with them included).
This, naturally, doesn't address the problem with non-postal house numbers and the different ways of handling them in general, put should limit the effect it has to the places and cantons that are not using a numbering scheme that doesn't differentiate these in format.
Simon
talk-ch mailing list talk-ch@openstreetmap.ch http://lists.openstreetmap.ch/mailman/listinfo/talk-ch
talk-ch mailing list talk-ch@openstreetmap.ch http://lists.openstreetmap.ch/mailman/listinfo/talk-ch
Hello Simon,
first of all thank you for taking into account the previous feedback and filtering the streets supposedly inhabited but without address. it helps a lot to focus on the most useful streets to add first (in my opinion).
However, I saw the opposite problem: a new street with 2 inhabited houses, is missing from the GWR<>osm summary but available both on the cantonal and federal site. is it possible to add the date of the GWR data on the page http://qa.poole.ch/addresses/ch/ or to indicate the frequency of download of the federal file? it would help to know if it is because the tool has not download the file recently or if it is "out of osm" that the update has not yet been done.
Regards, Marc
On 06.12.19 10:22, Simon Poole wrote:
I've updated the address data on http://qa.poole.ch/addresses/ch/ to the current GWR address data.
There is one, larger, change from the previous update (mid-2019). As it seems to be a trend to add ancillary buildings now with house numbers in the format/number/*.*/number /I've removed them from both the statistics and the generated files, this removes a good 200'000 addresses from the current dataset (note if somebody wants this data I can generate files with them included).
This, naturally, doesn't address the problem with non-postal house numbers and the different ways of handling them in general, put should limit the effect it has to the places and cantons that are not using a numbering scheme that doesn't differentiate these in format.
Simon
The date I updated the street list on is visible here https://wiki.openstreetmap.org/wiki/OSM_-_GWR_Street_and_Place_Names_Compari...
I don't update it automatically* and adding that as this stage wouldn't make sense anyway as on the one hand the BfS isn't officially distributing it anyway any more (as in linking to it from a website) and as we can expect the list from swisstopo to become available sometime this year (currently they are planning to release it on CC BY 4.0 terms which is a bit of a bummer). In any case while the list is generated from the GWR it was never closely synchronized to the GWR address data releases so differences are expected.
Simon
* this is slightly more involved than it sounds, as particularly at the beginning of the year there is no guarantee that the municipality data in the list matches up with what OSM has, actually we currently have that issue with the address data from the GWR.
Am 20.01.2020 um 20:40 schrieb marc marc:
Hello Simon,
first of all thank you for taking into account the previous feedback and filtering the streets supposedly inhabited but without address. it helps a lot to focus on the most useful streets to add first (in my opinion).
However, I saw the opposite problem: a new street with 2 inhabited houses, is missing from the GWR<>osm summary but available both on the cantonal and federal site. is it possible to add the date of the GWR data on the page http://qa.poole.ch/addresses/ch/ or to indicate the frequency of download of the federal file? it would help to know if it is because the tool has not download the file recently or if it is "out of osm" that the update has not yet been done.
Regards, Marc
On 06.12.19 10:22, Simon Poole wrote:
I've updated the address data on http://qa.poole.ch/addresses/ch/ to the current GWR address data.
There is one, larger, change from the previous update (mid-2019). As it seems to be a trend to add ancillary buildings now with house numbers in the format/number/*.*/number /I've removed them from both the statistics and the generated files, this removes a good 200'000 addresses from the current dataset (note if somebody wants this data I can generate files with them included).
This, naturally, doesn't address the problem with non-postal house numbers and the different ways of handling them in general, put should limit the effect it has to the places and cantons that are not using a numbering scheme that doesn't differentiate these in format.
Simon
talk-ch mailing list talk-ch@openstreetmap.ch http://lists.openstreetmap.ch/mailman/listinfo/talk-ch
-
Ludwig Baumgart -
marc marc -
Simon Poole