Incorrect dots and inaccuracies among city profiles

magpie · Post by **magpie** » Fri Oct 13, 2006 11:24 am

Gauss wrote:After above criticism relating to Asemissen and other dependent villages, I wonder whether inhabitants of, say, Sechtem share this point of view.

Then please merge e.g. Bornheim Roisdorf, Bornheim(roisdorf) and Roisdorf.

I've been ill the last days so I didn't say anything about this. But I cannot understand the critics: Since the locations would not be deleted if they were merged under Bornheim, it would still be possible to see, that

1. a note was entered in a particular part of Bornheim
2. distance calculation is accurate
etc.

Only the city statistics would be missing. And if you take a look at the statistics it is often only one or two users entering notes in that particular location (in Bornheim-merten it is one).

Düsseldorf e.g. has 49 parts. Some of them are quite rural or "on the wrong side" of the rhine. I could make this a criterium for entering my notes. Then I would not find them in Düssdeldorf but in Hamm, Hubbelrath, Eller, Vennhausen etc. I think this is too much.

We should only use official cities and villages as cities and subsume parts as locations. What else is a city-location conecpt for?

By the way, it would be a good idea, to show the locations of a city in it's profile.

Post by **Nerzhul** » Fri Oct 13, 2006 12:49 pm

magpie wrote:By the way, it would be a good idea, to show the locations of a city in it's profile.

Um, maybe for us for debugging, but not for the public. There's just so much (user generated) crap in the database. And you don't want to view the city profile of Amsterdam with its 2334 subsumed locations...

As for androls request, this was indeed a small glitch which is fixed. However, if there are multiple matching parent cities for a new location a new parent will always be created.

magpie · Post by **magpie** » Sun Oct 15, 2006 8:58 pm

Biberach

is wrong.

There is one Biberach in Baden (Postcode 77781) and Biberach an der Riß (Postcode 88400).

I think there are more than 20 cities of Biberach an der Riß in different writings. Please merge them. An leave one Biberach for Biberach in Baden.

Elmo · Post by **Elmo** » Sun Oct 15, 2006 9:36 pm

Hinterbichl is wrong. The dot is located at position 2 at Viamichelin, while
it should be position 1.

androl · Post by **androl** » Sun Oct 15, 2006 11:17 pm

Gauss wrote:
2. androl wrote:why do these three notes
http://www.eurobilltracker.eu/index.php ... d=22365077
http://www.eurobilltracker.eu/index.php ... d=22365084
http://www.eurobilltracker.eu/index.php ... d=22365099
create new city profiles [...]
Good question. In principle the heuristic is supposed to take care of capitalisation, as far as I understand it.

Edit: Androl, you will not expect a computer to recognise Ulm-<minorsubdivision> to be the same as Um-<minorsubdivision>.

What do you mean? where did I write Ulm-Mühlhofen? I don't think I ever did. I write UM-Mühlhofen, which means "Uhldingen-Mühlhofen-Mühlhofen", which contains a double "Mühlhofen", which is the reason why I abbreviate it.

magpie Fri Oct 13, 2006 8:55 am wrote:Following locations should be merged under the city of Bornheim (ID 12275)

Bornheim merten (ID 12280)
Bornheim merten (53332), Germany
Bornheim-Merten (53332), Germany

Bornheim-merten (ID 117864)
Bornheim-merten (53332), Germany

Bornheim/merten (ID 12285)
Bornheim/merten (53332), Germany

...

magpie Fri Oct 13, 2006 11:24 am wrote:By the way, it would be a good idea, to show the locations of a city in it's profile.

may I ask how you found out the locations-cities relationships for your list? (I don't think you entered notes in all of these locations) I could also make better comments on the locations if I had such an ability.
Also, my only possibility to find city profiles where I haven't entered a note is binary search on the id, which is also time consuming

yli · Post by **yli** » Mon Oct 16, 2006 12:24 am

Putte (2950) and

Putte (2580) aren't the same location.

2580 is Putte (mun. Putte)
2950 is Putte (mun. Kapellen)

both are in the province of Antwerpen

magpie · Post by **magpie** » Mon Oct 16, 2006 2:45 pm

androl wrote:may I ask how you found out the locations-cities relationships for your list? (I don't think you entered notes in all of these locations) I could also make better comments on the locations if I had such an ability.

I am as so called plotter. I am allowed to enter or change location coordinates.

Phaseolus · Post by **Phaseolus** » Tue Oct 17, 2006 9:13 pm

I am not sure to post on the correct place ; but, this evening when checking the user map (=the map showing all registered users) ; I spotted one green dot (i.e. quite a lot of users) in the Middle of the North Sea : the dot is between the Netherladns and UK, but closer to the Netherlands.

It looks strange ; I am not aware of a inhabited island on this place !

Dakkus · Post by **Dakkus** » Wed Oct 18, 2006 12:20 pm

There's a page for checking which zip codes belong to a certain dot, but I can't remember its address :(

EDIT: Found it.
There: http://www.eurobilltracker.eu/dots.php

Apparently there's nothing, so it's an average of some Dutch and some British towns with the same name, I suppose. Or maybe one of an Irish and a German one.

de*lerger · Post by **de*lerger** » Wed Oct 18, 2006 12:57 pm

St-Valéry-en-Caux lost its dot after the break.
http://www.eurobilltracker.eu/index.php ... d=21098310

Now there's a gap in the dotline on my map.

debemaks · Post by **debemaks** » Wed Oct 18, 2006 2:11 pm

Hi, I noticed that these dots are misplaced, they should be place as the same of Lerici - 19030 instead of somewhere in the nearby of Rimini, Ferrara.

http://www.eurobilltracker.eu/index.php ... d=17847965
http://www.eurobilltracker.eu/index.php ... d=22304353

androl · Post by **androl** » Wed Oct 18, 2006 10:26 pm

I don't understand why thisnote has a dot but its city profile does not have a map

Asiamaniac · Post by **Asiamaniac** » Wed Oct 18, 2006 10:44 pm

I also have got some corrections to suggest:

Wiesbaden-Biebrich; Wiesbaden-Igstadt; Mainz-Kastel and Mainz-Kostheim all must be part of Wiesbaden.

For the latter two this may sound confusing as these parts of town lead to expect that they are part of the city of Mainz.

Haah, this is indeed a needle that stings in the flesh of the Mainzers since the end of World War ll but all four mentioned town names officially and correctly belong to Wiesbaden.

Gauss · Post by **Gauss** » Wed Oct 18, 2006 11:01 pm

Phaseolus wrote:I am not sure to post on the correct place ; but, this evening when checking the user map (=the map showing all registered users) ; I spotted one green dot (i.e. quite a lot of users) in the Middle of the North Sea : the dot is between the Netherladns and UK, but closer to the Netherlands.

Fixed.

de*lerger wrote:St-Valéry-en-Caux lost its dot after the break.
http://www.eurobilltracker.eu/index.php ... d=21098310

Fixed. Ultimately caused by the (correct) accentuation.

debemaks wrote:Hi, I noticed that these dots are misplaced, they should be place as the same of Lerici - 19030 instead of somewhere in the nearby of Rimini, Ferrara.

Common phenomenon with places like "19035 Pozzuolo" which are so minor that neither Poste Italiane, nor www.nonsolocap.it, nor www.multimap.com, nor Google.it know.

androl wrote:I don't understand why thisnote has a dot but its city profile does not have a map?

0-0-locations are ignored when the map is generated; that's an indirect consequence of your own observation that 0-0-locations screw up many maps in the city profiles. Coordinates for 88069 Obereisenbach had already been entered ahead, when someone generated the location 88069 Tettnang-Obereisenbach, the two corresponding city profiles were simply merged. Such Kuhkäffer take up enough time, and merging is much faster than copying coordinates.

androl · Post by **androl** » Wed Oct 18, 2006 11:37 pm

Gauss wrote:0-0-locations are ignored when the map is generated; that's an indirect consequence of your own observation that 0-0-locations screw up many maps in the city profiles.

that's a good idea

Gauss wrote:Coordinates for 88069 Obereisenbach had already been entered ahead, when someone generated the location 88069 Tettnang-Obereisenbach, the two corresponding city profiles were simply merged. Such Kuhkäffer take up enough time, and merging is much faster than copying coordinates.

I don't understand. But this tells me, "88069 Obereisenbach" has coordinates and "88069 Tettnang-Obereisenbach" does not. And the profiles (1,2) are not merged. (should be merged to 104275)
That means, the note report map heuristics knows that "88069 Obereisenbach" matches best for "88069 Tettnang-Obereisenbach" and takes that location's coordinates, and since the coordinates are exact, it means that it only takes "88069 Obereisenbach" and not "88069 Tettnang"