[Imports] Update of german Aral petrol stations

Wed Feb 18 13:04:01 UTC 2015

Thanks for thinking of a solution with us.

When referring to Avia, I meant Avia International (fuel retail): http://www.avia-international.com/

I definitely agree that the most useful POI’s are those for tourists (hotels, restaurants, shops etc.)
Nonetheless, a lot of navigation systems rely upon OSM and thus OSM’s POI layer for navigation purposes. Therefore I think it would be a good idea to include as many POI’s as possible (not just for tourists, seeing as some people sometimes struggle to find things in their area, myself included haha).

I most definitely concur that a direct dump is not an option. When we give out our data, we will definitely format it in a way that is most accessible for everyone, no worries. I think we might even have some data that includes the OSM tags, but I will check this out soon, quite a busy day today.
I will get back to everyone concerning further data/ bulk uploads once we have cleared what we’re allowed to share.

Concerning the API: this is what we are planning on building. If you have any input, feel free to let me know!

The approach you use in Flanders seems quite interesting, could you maybe tell me more about that? What kind of file does the Belgian government supply you with? Is this a PAF file? And what locations does this include?

Also a small update concerning the official BP authorization document: I have contacted the person in charge of this to send me an email/document saying that they give us authorization, but I just heard that he will be back in the office on the 23rd of February, so a little more patience is required.

If anyone has any thoughts on this, please feel free to discuss this with me, I’m open to ideas!

From: Sander Deryckere [mailto:sanderd17 at gmail.com]
Sent: Wednesday, February 18, 2015 11:57 AM
To: Bas van Kempen
Cc: Bryce Nesbitt; Holger Jeromin; Imports OpenStreetMap.org; talk-de
Subject: Re: [Imports] Update of german Aral petrol stations

IMO (and that's a very personal opinion), the most useful are POI for tourists.

Due to the community-driven quality assurance, and the general difference in quality between regions, data analysts and other professionals often don't like to use OSM data (unless when investigating OSM itself, or when offering services around OSM). As such, regular people are most of the time the end users of OSM data, and this happens in the form of general applications (general web maps, smarphone apps, ...), and even travel guides. Regular people also use a map most often when they're abroad.
This target is rather well overlapping with your audience: companies that want to make local adds.

  *   BP/Aral : It's always handy to know where to find gas for your car as a tourist
  *   Best Western Hotels: Also rather touristic
  *   Avia: Not sure what brand you mean here, there seem to be multiple brands called Avia
  *   Costa Coffee: Also a nice one for tourists
  *   Lidl Supermarkets / Carrefour : general supermarkets are also a good thing to find when you're a tourist
  *   Hornbach: I don't know the brand, but from a quick search, I guess it's a DIY-shop. Less interesting for tourists, so it has a lower priority IMO, but any shop is more than welcome
  *   Glassdrive: a car-glass repair service? We can always hope tourists don't need this, but when they do, it's very handy to find on a map.
  *   OKQ8: banks and ATMs are also a rather interesting feature
  *   Movenpick restaurants/hotels/bars: Obviously interesting for tourists.

So there seems to be only one in your list that's moderately interesting, and one I don't know, the other ones are all very interesting.

Of course, a direct dump is no option, so the only way to go is to serve the data in a format that can be queried by any OSM editor. That can be raw data, and let the OSM community build all tools around it. It can be with a simple API on your side (get a number of POI around a certain coordinate), in which case the community can build a map view around it that compares your data with OSM data, this way you have more access control, but still a rather low maintenance. Or it can be a complete solution on your side, but that would probably require you to learn the OSM data tagging customs, and translate your DB into OSM tags.

For comparing data, I think it would be best to use the Overpass API service. With Overpass API, it's possible to load a selection of raw OSM data (f.e. all gas stations in a certain bbox) directly in the browser, and thus execute all comparison code client-side. This would save processing power, but also maintenance.

It's also the sort of approach we use in Flanders for the address import and maintenance. The government gives us a file with all address positions. I split that file geographically into parts that are small enough so it doesn't choke the browser. Then every mapper who wants to import addresses visits the import webpage, that webpage loads all address data from the government and from OSM for that region, compares it with each other, and reports the results on a slippy map. So if it's possible with addresses, it should certainly be possible with less dense data.

In any case, thanks for looking into it.
Regards,
Sander

2015-02-18 11:00 GMT+01:00 Bas van Kempen <bvankempen at navads.eu<mailto:bvankempen at navads.eu>>:
Hi Sander/everyone else,

I’m the person working for NavAds here in Amsterdam. What kind of data are you thinking about? Ofcourse we can’t share ALL of our data right now, but I’m curious as to which you would consider useful to potentially publish to OSM. I can name a few customers that we currently have in our database, namely:

•        BP/Aral

•        Best Western Hotels

•        Avia

•        Costa Coffee

•        Lidl Supermarkets

•        Carrefour

•        Burger King Nordic

•        Hornbach

•        Glassdrive

•        OKQ8

•        Movenpick restaurants/hotels/bars

Just to name a few.  We’re currently investigating if there is potential interest in these POI’s. If there is interest in uploading these to OSM, we could see which data we can potentially make available for download. I personally think it would benefit the entire OSM mapping community if we could add all these places, seeing as they are relevant to people using OSM for navigation purposes. If you let me know what kind of data would be relevant to upload to OSM (eg. Petrolstations, hotels, supermarkets, etc.) I can have a look at what we can publish through our website. Ideally, it would be all of our data, but I will leave that up to the good people of OSM. :)

From: Sander Deryckere [mailto:sanderd17 at gmail.com<mailto:sanderd17 at gmail.com>]
Sent: Wednesday, February 18, 2015 10:08 AM
To: Bryce Nesbitt
Cc: Holger Jeromin; Imports OpenStreetMap.org; talk-de
Subject: Re: [Imports] Update of german Aral petrol stations

I do wonder what datasets other than the BP stations are available on navads.nl<http://navads.nl>
If every region has some data, I think it would be great to think of a more global solution.
F.e. try to make an automated QA-style map, that shows where OSM POI are missing, where tags are missing, or where OSM might contain already deleted POI.
That way, by combining the data from different sources into one QA map, it would be possible for contributors to work in their local neighbourhoods (instead of a small team doing a big import of one feature). Together with user feedback, it might even improve the original database.
This would be achievable when navads shares all (or more) datasets openly, or when the QA-style map is hosted on the navads site itself, where the number of requests can be limited, a login can be required, ... I hope this fits somehow in the business plan of navads, and it's legally possible.
Regards,
Sander

2015-02-17 18:10 GMT+01:00 Bryce Nesbitt <bryce2 at obviously.com<mailto:bryce2 at obviously.com>>:
What about the case where the station changed owners, and is currently "brand=Something Else"?
How will those match up?

_______________________________________________
Imports mailing list
Imports at openstreetmap.org<mailto:Imports at openstreetmap.org>
https://lists.openstreetmap.org/listinfo/imports

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/imports/attachments/20150218/80491f3b/attachment-0001.html>