addr:interpolation ways not along road
Hi all, I wonder if mkgmap should treat this a valid addr:interpolation way: http://www.openstreetmap.org/way/32922365 The wiki says: "How to interpolate the house numbers belonging to the way along the respective street." but this way connects house numbers which are up to 170 m away from the road named Bergweg, so I think it is not "along" the respective street. With r3502 I've added checks to ignore numbers which belong to invalid addr:interpolation ways, and this detection says it is invalid. I've kept this with r3503. This also happens for some other addr:interpolation ways for the way 25748155. Please let me know when you think that this is too aggressive. Gerd -- View this message in context: http://gis.19327.n5.nabble.com/addr-interpolation-ways-not-along-road-tp5838... Sent from the Mkgmap Development mailing list archive at Nabble.com.
Hi, On Mon, Mar 23, GerdP wrote:
Hi all,
I wonder if mkgmap should treat this a valid addr:interpolation way: http://www.openstreetmap.org/way/32922365
The wiki says: "How to interpolate the house numbers belonging to the way along the respective street."
but this way connects house numbers which are up to 170 m away from the road named Bergweg, so I think it is not "along" the respective street.
The addr:interpolation way is valid in the regard that it matches reality. If this is good tagging is another question. And if the street really stops that early: I don't know, the pictures of bing and google are not good enough for this. But I know that this will not really help you. Thorsten -- Thorsten Kukuk, Senior Architect SLES & Common Code Base SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nuernberg, Germany GF: Felix Imendörffer, Jane Smithard, Jennifer Guild, Dilip Upmanyu, Graham Norton, HRB 21284 (AG Nürnberg)
Hi Thorsten, I agree that the data doesn't look that bad. My problem is this: 1) We have to interpret the info in addr:interpolation ways because it helps to find the right intervals in many cases, for example it normally gives a clear hint to which part of a road a pair of houses belongs when muultiple roads form a # shape. See http://www.openstreetmap.org/node/427691367 which is closer to a different road. 2) On the other hand, a lot of the addr:interpolation ways in Canada are wrong, e.g. http://www.openstreetmap.org/way/75809612 crosses the road. If mkgmap only ignores this wrong addr:interpolation way it will still use the obviously wrong odd house number 3367 for the road side that has even numbers. So, mkgmap should also ignore that number. As it is sometimes difficult to decide which of the two numbers is wrong, I decided that it is okay to ignore both. I guess I have to find more detailed rules :-( Gerd
Date: Tue, 24 Mar 2015 08:35:22 +0100 From: kukuk@suse.de To: mkgmap-dev@lists.mkgmap.org.uk Subject: Re: [mkgmap-dev] addr:interpolation ways not along road
Hi,
On Mon, Mar 23, GerdP wrote:
Hi all,
I wonder if mkgmap should treat this a valid addr:interpolation way: http://www.openstreetmap.org/way/32922365
The wiki says: "How to interpolate the house numbers belonging to the way along the respective street."
but this way connects house numbers which are up to 170 m away from the road named Bergweg, so I think it is not "along" the respective street.
The addr:interpolation way is valid in the regard that it matches reality. If this is good tagging is another question.
And if the street really stops that early: I don't know, the pictures of bing and google are not good enough for this.
But I know that this will not really help you.
Thorsten
-- Thorsten Kukuk, Senior Architect SLES & Common Code Base SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nuernberg, Germany GF: Felix Imendörffer, Jane Smithard, Jennifer Guild, Dilip Upmanyu, Graham Norton, HRB 21284 (AG Nürnberg) _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Hi, On Tue, Mar 24, Gerd Petermann wrote:
2) On the other hand, a lot of the addr:interpolation ways in Canada are wrong, e.g. http://www.openstreetmap.org/way/75809612 crosses the road.
Yes, North-America is suffering a lot from this automatic imports. They hoped to have quick a base and then that mapper will fix it, but in the end nobody was interested in fixing data, what was there was good enough for them ...
If mkgmap only ignores this wrong addr:interpolation way it will still use the obviously wrong odd house number 3367 for the road side that has even numbers. So, mkgmap should also ignore that number. As it is sometimes difficult to decide which of the two numbers is wrong, I decided that it is okay to ignore both.
If you find a wrong addr:interpolation, you should ignore not only the way, but all the points, too. If the way is wrong, then the points are most likely wrong, too. Thorsten
I guess I have to find more detailed rules :-(
Gerd
Date: Tue, 24 Mar 2015 08:35:22 +0100 From: kukuk@suse.de To: mkgmap-dev@lists.mkgmap.org.uk Subject: Re: [mkgmap-dev] addr:interpolation ways not along road
Hi,
On Mon, Mar 23, GerdP wrote:
Hi all,
I wonder if mkgmap should treat this a valid addr:interpolation way: http://www.openstreetmap.org/way/32922365
The wiki says: "How to interpolate the house numbers belonging to the way along the respective street."
but this way connects house numbers which are up to 170 m away from the road named Bergweg, so I think it is not "along" the respective street.
The addr:interpolation way is valid in the regard that it matches reality. If this is good tagging is another question.
And if the street really stops that early: I don't know, the pictures of bing and google are not good enough for this.
But I know that this will not really help you.
Thorsten
-- Thorsten Kukuk, Senior Architect SLES & Common Code Base SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nuernberg, Germany GF: Felix Imendörffer, Jane Smithard, Jennifer Guild, Dilip Upmanyu, Graham Norton, HRB 21284 (AG Nürnberg) _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
-- Thorsten Kukuk, Senior Architect SLES & Common Code Base SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nuernberg, Germany GF: Felix Imendörffer, Jane Smithard, Jennifer Guild, Dilip Upmanyu, Graham Norton, HRB 21284 (AG Nürnberg)
Hi Thorsten,
Yes, North-America is suffering a lot from this automatic imports. They hoped to have quick a base and then that mapper will fix it, but in the end nobody was interested in fixing data, what was there was good enough for them ...
I think the problem is that the data is so detailed that one cannot believe that it is totally wrong. I'll look at this later.
If mkgmap only ignores this wrong addr:interpolation way it will still use the obviously wrong odd house number 3367 for the road side that has even numbers. So, mkgmap should also ignore that number. As it is sometimes difficult to decide which of the two numbers is wrong, I decided that it is okay to ignore both.
If you find a wrong addr:interpolation, you should ignore not only the way, but all the points, too. If the way is wrong, then the points are most likely wrong, too.
Yes, I agree. I just have to find better rules to decide what addr:interpolation way is wrong. Gerd
participants (3)
-
Gerd Petermann -
GerdP
-
Thorsten Kukuk