Re: [OSM-talk-be] Importing polygons of administrative boundaries for Belgium into OSM

[Tim Couwelier]

Dear all,

recent discussion on the forum has given some insight in WHY certain people
feel they need to adjust admin boundaries, at it seemingly makes no sense a
house or company site can be in two different municipalities at once.

Personally, I feel we should NOT edit based on such logic, for such cases
us belgians seem to have a way to just 'deal with it' or 'not make a big
fuss', as long as mail arrives where it should we're happy.
I've also strongly defended my point there.

I do have to agree you cannot possibly survey the borders in the field:
- where they run along / cross roads
- where they decided to put the sign where it seemed 'best fit' to put it,
with imprecision of multiple meters.
- ...

I'm not in favor of a 'blind replacement', but in cases where there's
differences, reïnstating the admin boundaries would get my support.


Note:
I once reported to Roeselare their sign of 'bebouwde kom' (built-up area)
was put up BEYOND the sign for the border with Hooglede. Seems to make no
sense, given the boundaries should be in the same position.
So rather then 'fixing' it by having the two signs on the same pole...  I
noticed later they just switched them around! So I pointed out that again...
Current situation: Driving from Roeselare to Hooglede, the signs are about
10m apart but in correct direction. From Hooglede to Roeselare the 'border'
sign is missing, there's only the sign F1 for the built-up area.

The above just to show how unreliable field survey is for those.

Op di 7 mei 2019 om 11:12 schreef Lionel Giard :

> I know that it's not always possible but in some cases, you can find the
> boundaries via the boundary stones (when they exists like in
> buildings/chapels). But they are generally well hidden - it often need
> local knowledge like from local notary of the village or local
> historican... And if you have contact with the municipality, you can get
> the information directly when they made a small change in the cadastre to
> reflect reality (like when they correct huge mistakes in the cadastre from
> more than a century ago).
>
> But i'm pretty sure that cadastre is not yet up to good accuracy (compare
> to reality) at the moment. They are improving their data but, as far as i
> know, it is not yet corrected everywhere.
> For example, in a lot of rural areas, the cadastre was made from the
> center of the agglomeration (i.e. the best accuracy is found near the
> church or the center of the village) and if you go more and more away of
> this center, you'll find big errors or incertitude in the data (like the
> cadastre parcels can be few dozen meters off reality).  Relative to other
> cadastre parcels, the cadastre data is correct but not if you put our OSM
> data on top, it will not match for the comparable data like the location of
> streets, ...
> That's why i say that i will not *YET *import blindly the cadastre
> administrative data, as it *can *be "imprecise" relative to the reality
> at the moment.
>
> Le mar. 7 mai 2019 à 10:42, K W via Talk-be  a
> écrit :
>
>> Good morning,
>> thanks a lot for your reply. It is difficult to see for me how you could
>> see "by field survey" where administrative boundaries run. This is
>> impossible, as they are not a physical feature. Therefore, the only valid
>> reference are the official data. I agree that OSM is quite good, but there
>> are definitely instances where the borders deviate by "a few metres" or
>> more – which, for certain purposes, makes a big difference. Indeed, so far
>> I have corrected mistakes by hand, based on the cadastre. As the only valid
>> data one could imagine when it comes to administrative boundaries are the
>> official ones (be it from the cadastre or from ING) it would still, in my
>> opinion, be better to import those into OSM rather than correcting manually
>> here and there.
>>
>> On Tuesday, May 7, 2019, 9:58:05 AM GMT+2, Lionel Giard <
>> lionel.gi...@gmail.com> wrote:
>>
>>
>> Hello,
>>
>> To me this dataset is not necessarily better than what we have in OSM (as
>> we corrected some border by field survey  where this dataset is wrong).
>> I would never "replace" blindly all existing geometry by this one. I think
>> we can say that we are more or less correct (i never saw errors of more
>> than a few meters) - and i find it better to improve the current boundaries
>> by hand : checking the official sources, the reality on the ground, ... For
>> example, i had a part of a street that i surveyed recently where the
>> addresses changed of municipality (to have the whole street in the same
>> municipality and not cut it before the last three houses). We are correct
>> in OSM, but not yet in the official data (as they don't update often
>> enough).
>>
>> Technically, it is not the official source for administrative boundaries
>> in Belgium, as the official source is the Cadastre - which is also open -
>> but it is not yet accurate enough (even if they are improving it a lot
>> recently).
>>
>> Le mar. 7 ma

Re: [OSM-talk-be] Importing polygons of administrative boundaries for Belgium into OSM

[Lionel Giard]

I know that it's not always possible but in some cases, you can find the
boundaries via the boundary stones (when they exists like in
buildings/chapels). But they are generally well hidden - it often need
local knowledge like from local notary of the village or local
historican... And if you have contact with the municipality, you can get
the information directly when they made a small change in the cadastre to
reflect reality (like when they correct huge mistakes in the cadastre from
more than a century ago).

But i'm pretty sure that cadastre is not yet up to good accuracy (compare
to reality) at the moment. They are improving their data but, as far as i
know, it is not yet corrected everywhere.
For example, in a lot of rural areas, the cadastre was made from the center
of the agglomeration (i.e. the best accuracy is found near the church or
the center of the village) and if you go more and more away of this center,
you'll find big errors or incertitude in the data (like the cadastre
parcels can be few dozen meters off reality).  Relative to other cadastre
parcels, the cadastre data is correct but not if you put our OSM data on
top, it will not match for the comparable data like the location of
streets, ...
That's why i say that i will not *YET *import blindly the cadastre
administrative data, as it *can *be "imprecise" relative to the reality at
the moment.

Le mar. 7 mai 2019 à 10:42, K W via Talk-be  a
écrit :

> Good morning,
> thanks a lot for your reply. It is difficult to see for me how you could
> see "by field survey" where administrative boundaries run. This is
> impossible, as they are not a physical feature. Therefore, the only valid
> reference are the official data. I agree that OSM is quite good, but there
> are definitely instances where the borders deviate by "a few metres" or
> more – which, for certain purposes, makes a big difference. Indeed, so far
> I have corrected mistakes by hand, based on the cadastre. As the only valid
> data one could imagine when it comes to administrative boundaries are the
> official ones (be it from the cadastre or from ING) it would still, in my
> opinion, be better to import those into OSM rather than correcting manually
> here and there.
>
> On Tuesday, May 7, 2019, 9:58:05 AM GMT+2, Lionel Giard <
> lionel.gi...@gmail.com> wrote:
>
>
> Hello,
>
> To me this dataset is not necessarily better than what we have in OSM (as
> we corrected some border by field survey  where this dataset is wrong).
> I would never "replace" blindly all existing geometry by this one. I think
> we can say that we are more or less correct (i never saw errors of more
> than a few meters) - and i find it better to improve the current boundaries
> by hand : checking the official sources, the reality on the ground, ... For
> example, i had a part of a street that i surveyed recently where the
> addresses changed of municipality (to have the whole street in the same
> municipality and not cut it before the last three houses). We are correct
> in OSM, but not yet in the official data (as they don't update often
> enough).
>
> Technically, it is not the official source for administrative boundaries
> in Belgium, as the official source is the Cadastre - which is also open -
> but it is not yet accurate enough (even if they are improving it a lot
> recently).
>
> Le mar. 7 mai 2019 à 09:21, K W via Talk-be  a
> écrit :
>
> EN: Good morning, from time to time I make adjustments in OSM where the
> boundaries between municipalities in OSM are not exactly following the
> official data. However, this is a tedious process, and it would be nice, of
> course, to have the exact data for all Belgium municipalities (and former
> municipalities – "deelgemeenten"/"sections") in one go.
> Now, I have seen that it is possible to download the official border
> polygons from the site of the National Geographic Institute for free: IGN
> - produits - Données numériques  . The
> only problem is that I have no clue whatsoever how this could be done
> technically, i.e. to replace the existing municipal boundaries in OSM by
> the correct polygons from IGN. Could anybody help with this?
>
> IGN - produits - Données numériques
>
> 
>
>
> NL: Goedemorgen, af en toe maak ik correcties in OSM waar de grenzen
> tussen gemeenten in OSM niet precies de officiële gegevens volgen. Maar dat
> is een moeilijke procedure en het zou natuurlijk leuk zijn om de precieze
> gegevens voor alle Belgische gemeenten (en vroegere gemeenten –
> "deelgemeenten") in een keer te hebben.
> Nu heb ik gezien dat het mogelijk is om de officiële grenspolygonen van de
> webpagina van het Nationaal Geografisch Instituut gratis te downloaden:
> http://www.ngi.be/FR/FR1-5-2.shtm. Het enige probleem is dat ik geen idee
> heb hoe je dat technisch moet doen, dus hoe de huidige gemeentegrenzen op
> OSM te vervangen door de correcte polygonen van het NGI. Zou iemand kunnen
> he

Re: [OSM-talk-be] Importing polygons of administrative boundaries for Belgium into OSM

[K W via Talk-be]

 Good morning,thanks a lot for your reply. It is difficult to see for me how 
you could see "by field survey" where administrative boundaries run. This is 
impossible, as they are not a physical feature. Therefore, the only valid 
reference are the official data. I agree that OSM is quite good, but there are 
definitely instances where the borders deviate by "a few metres" or more – 
which, for certain purposes, makes a big difference. Indeed, so far I have 
corrected mistakes by hand, based on the cadastre. As the only valid data one 
could imagine when it comes to administrative boundaries are the official ones 
(be it from the cadastre or from ING) it would still, in my opinion, be better 
to import those into OSM rather than correcting manually here and there.
On Tuesday, May 7, 2019, 9:58:05 AM GMT+2, Lionel Giard 
 wrote:  
 
 Hello,
To me this dataset is not necessarily better than what we have in OSM (as we 
corrected some border by field survey  where this dataset is wrong). I 
would never "replace" blindly all existing geometry by this one. I think we can 
say that we are more or less correct (i never saw errors of more than a few 
meters) - and i find it better to improve the current boundaries by hand : 
checking the official sources, the reality on the ground, ... For example, i 
had a part of a street that i surveyed recently where the addresses changed of 
municipality (to have the whole street in the same municipality and not cut it 
before the last three houses). We are correct in OSM, but not yet in the 
official data (as they don't update often enough). Technically, it is not the 
official source for administrative boundaries in Belgium, as the official 
source is the Cadastre - which is also open - but it is not yet accurate enough 
(even if they are improving it a lot recently). 
Le mar. 7 mai 2019 à 09:21, K W via Talk-be  a écrit 
:

EN: Good morning, from time to time I make adjustments in OSM where the 
boundaries between municipalities in OSM are not exactly following the official 
data. However, this is a tedious process, and it would be nice, of course, to 
have the exact data for all Belgium municipalities (and former municipalities – 
"deelgemeenten"/"sections") in one go.Now, I have seen that it is possible to 
download the official border polygons from the site of the National Geographic 
Institute for free: IGN - produits - Données numériques . The only problem is 
that I have no clue whatsoever how this could be done technically, i.e. to 
replace the existing municipal boundaries in OSM by the correct polygons from 
IGN. Could anybody help with this?

| 
| 
| 
|  |  |

 |

 |
| 
|  | 
IGN - produits - Données numériques


 |

 |

 |



NL: Goedemorgen, af en toe maak ik correcties in OSM waar de grenzen tussen 
gemeenten in OSM niet precies de officiële gegevens volgen. Maar dat is een 
moeilijke procedure en het zou natuurlijk leuk zijn om de precieze gegevens 
voor alle Belgische gemeenten (en vroegere gemeenten – "deelgemeenten") in een 
keer te hebben.Nu heb ik gezien dat het mogelijk is om de officiële 
grenspolygonen van de webpagina van het Nationaal Geografisch Instituut gratis 
te downloaden: http://www.ngi.be/FR/FR1-5-2.shtm. Het enige probleem is dat ik 
geen idee heb hoe je dat technisch moet doen, dus hoe de huidige 
gemeentegrenzen op OSM te vervangen door de correcte polygonen van het NGI. Zou 
iemand kunnen helpen?
FR: Bonjour, du temps en temps je fais des corrections dans OSM où les 
frontières entre les communes dans OSM ne suivent pas les données officielles. 
Mais cela est un processus pénible et il serait agréable, bien sûr, avoir les 
données exactes pour toutes les communes belges (et communes anciennes – 
"sections") d’un seul coup.Or, j’ai vu qu’il est possible télécharger 
gratuitement les polygons officiels des frontières du site de l’Institut 
National Géographique: http://www.ngi.be/FR/FR1-5-2.shtm. Le seul problème est 
que je n’ai aucune idée comment faire cela techniquement, c’est à dire comment 
remplacer les frontières communales existantes dans OSM par les polygons 
corrects de l’IGN. Est-ce que quelqu’un pourrait aider ?
DE: Guten Morgen, ab und zu bringe ich Korrekturen in OSM an, wo die Grenzen 
zwischen den Gemeinden in OSM nicht genau den offiziellen Daten folgen. Aber 
das ist natürlich ein mühsamer Prozess, und es wäre natürlich schön, die 
genauen Daten für alle belgischen Gemeinden (und früheren Gemeinden – 
"Teilgemeinden") auf einmal zu haben.Jetzt habe ich gesehen, dass es möglich 
ist die offiziellen Grenzpolygone auf der Webseite des Nationalen 
Geographischen Instituts gratis herunterzuladen: 
http://www.ngi.be/FR/FR1-5-2.shtm. Das einzige Problem ist, dass ich keine 
Ahnung habe, wie man das technisch machen müßte, mit anderen Worten wie man die 
Gemeindegrenzen, die in OSM derzeit existieren, durch die korrekten Polygone 
des ING ersetzen könnte. Kann mir da jemand helfen?
Many thanks/hartelijk dank/merci beaucou

Re: [OSM-talk-be] Importing polygons of administrative boundaries for Belgium into OSM

[Lionel Giard]

Hello,

To me this dataset is not necessarily better than what we have in OSM (as
we corrected some border by field survey  where this dataset is wrong).
I would never "replace" blindly all existing geometry by this one. I think
we can say that we are more or less correct (i never saw errors of more
than a few meters) - and i find it better to improve the current boundaries
by hand : checking the official sources, the reality on the ground, ... For
example, i had a part of a street that i surveyed recently where the
addresses changed of municipality (to have the whole street in the same
municipality and not cut it before the last three houses). We are correct
in OSM, but not yet in the official data (as they don't update often
enough).

Technically, it is not the official source for administrative boundaries in
Belgium, as the official source is the Cadastre - which is also open - but
it is not yet accurate enough (even if they are improving it a lot
recently).

Le mar. 7 mai 2019 à 09:21, K W via Talk-be  a
écrit :

> EN: Good morning, from time to time I make adjustments in OSM where the
> boundaries between municipalities in OSM are not exactly following the
> official data. However, this is a tedious process, and it would be nice, of
> course, to have the exact data for all Belgium municipalities (and former
> municipalities – "deelgemeenten"/"sections") in one go.
> Now, I have seen that it is possible to download the official border
> polygons from the site of the National Geographic Institute for free: IGN
> - produits - Données numériques  . The
> only problem is that I have no clue whatsoever how this could be done
> technically, i.e. to replace the existing municipal boundaries in OSM by
> the correct polygons from IGN. Could anybody help with this?
>
> IGN - produits - Données numériques
>
> 
>
>
> NL: Goedemorgen, af en toe maak ik correcties in OSM waar de grenzen
> tussen gemeenten in OSM niet precies de officiële gegevens volgen. Maar dat
> is een moeilijke procedure en het zou natuurlijk leuk zijn om de precieze
> gegevens voor alle Belgische gemeenten (en vroegere gemeenten –
> "deelgemeenten") in een keer te hebben.
> Nu heb ik gezien dat het mogelijk is om de officiële grenspolygonen van de
> webpagina van het Nationaal Geografisch Instituut gratis te downloaden:
> http://www.ngi.be/FR/FR1-5-2.shtm. Het enige probleem is dat ik geen idee
> heb hoe je dat technisch moet doen, dus hoe de huidige gemeentegrenzen op
> OSM te vervangen door de correcte polygonen van het NGI. Zou iemand kunnen
> helpen?
>
> FR: Bonjour, du temps en temps je fais des corrections dans OSM où les
> frontières entre les communes dans OSM ne suivent pas les données
> officielles. Mais cela est un processus pénible et il serait agréable, bien
> sûr, avoir les données exactes pour toutes les communes belges (et communes
> anciennes – "sections") d’un seul coup.
> Or, j’ai vu qu’il est possible télécharger gratuitement les polygons
> officiels des frontières du site de l’Institut National Géographique:
> http://www.ngi.be/FR/FR1-5-2.shtm. Le seul problème est que je n’ai
> aucune idée comment faire cela techniquement, c’est à dire comment
> remplacer les frontières communales existantes dans OSM par les polygons
> corrects de l’IGN. Est-ce que quelqu’un pourrait aider ?
>
> DE: Guten Morgen, ab und zu bringe ich Korrekturen in OSM an, wo die
> Grenzen zwischen den Gemeinden in OSM nicht genau den offiziellen Daten
> folgen. Aber das ist natürlich ein mühsamer Prozess, und es wäre natürlich
> schön, die genauen Daten für alle belgischen Gemeinden (und früheren
> Gemeinden – "Teilgemeinden") auf einmal zu haben.
> Jetzt habe ich gesehen, dass es möglich ist die offiziellen Grenzpolygone
> auf der Webseite des Nationalen Geographischen Instituts gratis
> herunterzuladen: http://www.ngi.be/FR/FR1-5-2.shtm. Das einzige Problem
> ist, dass ich keine Ahnung habe, wie man das technisch machen müßte, mit
> anderen Worten wie man die Gemeindegrenzen, die in OSM derzeit existieren,
> durch die korrekten Polygone des ING ersetzen könnte. Kann mir da jemand
> helfen?
>
> Many thanks/hartelijk dank/merci beaucoup/vielen Dank!
> ___
> Talk-be mailing list
> Talk-be@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-be
>
___
Talk-be mailing list
Talk-be@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-be

[OSM-talk-be] Importing polygons of administrative boundaries for Belgium into OSM

[K W via Talk-be]

EN: Good morning, from time to time I make adjustments in OSM where the 
boundaries between municipalities in OSM are not exactly following the official 
data. However, this is a tedious process, and it would be nice, of course, to 
have the exact data for all Belgium municipalities (and former municipalities – 
"deelgemeenten"/"sections") in one go.Now, I have seen that it is possible to 
download the official border polygons from the site of the National Geographic 
Institute for free: IGN - produits - Données numériques . The only problem is 
that I have no clue whatsoever how this could be done technically, i.e. to 
replace the existing municipal boundaries in OSM by the correct polygons from 
IGN. Could anybody help with this?

| 
| 
| 
|  |  |

 |

 |
| 
|  | 
IGN - produits - Données numériques


 |

 |

 |



NL: Goedemorgen, af en toe maak ik correcties in OSM waar de grenzen tussen 
gemeenten in OSM niet precies de officiële gegevens volgen. Maar dat is een 
moeilijke procedure en het zou natuurlijk leuk zijn om de precieze gegevens 
voor alle Belgische gemeenten (en vroegere gemeenten – "deelgemeenten") in een 
keer te hebben.Nu heb ik gezien dat het mogelijk is om de officiële 
grenspolygonen van de webpagina van het Nationaal Geografisch Instituut gratis 
te downloaden: http://www.ngi.be/FR/FR1-5-2.shtm. Het enige probleem is dat ik 
geen idee heb hoe je dat technisch moet doen, dus hoe de huidige 
gemeentegrenzen op OSM te vervangen door de correcte polygonen van het NGI. Zou 
iemand kunnen helpen?
FR: Bonjour, du temps en temps je fais des corrections dans OSM où les 
frontières entre les communes dans OSM ne suivent pas les données officielles. 
Mais cela est un processus pénible et il serait agréable, bien sûr, avoir les 
données exactes pour toutes les communes belges (et communes anciennes – 
"sections") d’un seul coup.Or, j’ai vu qu’il est possible télécharger 
gratuitement les polygons officiels des frontières du site de l’Institut 
National Géographique: http://www.ngi.be/FR/FR1-5-2.shtm. Le seul problème est 
que je n’ai aucune idée comment faire cela techniquement, c’est à dire comment 
remplacer les frontières communales existantes dans OSM par les polygons 
corrects de l’IGN. Est-ce que quelqu’un pourrait aider ?
DE: Guten Morgen, ab und zu bringe ich Korrekturen in OSM an, wo die Grenzen 
zwischen den Gemeinden in OSM nicht genau den offiziellen Daten folgen. Aber 
das ist natürlich ein mühsamer Prozess, und es wäre natürlich schön, die 
genauen Daten für alle belgischen Gemeinden (und früheren Gemeinden – 
"Teilgemeinden") auf einmal zu haben.Jetzt habe ich gesehen, dass es möglich 
ist die offiziellen Grenzpolygone auf der Webseite des Nationalen 
Geographischen Instituts gratis herunterzuladen: 
http://www.ngi.be/FR/FR1-5-2.shtm. Das einzige Problem ist, dass ich keine 
Ahnung habe, wie man das technisch machen müßte, mit anderen Worten wie man die 
Gemeindegrenzen, die in OSM derzeit existieren, durch die korrekten Polygone 
des ING ersetzen könnte. Kann mir da jemand helfen?
Many thanks/hartelijk dank/merci beaucoup/vielen Dank!___
Talk-be mailing list
Talk-be@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-be