On 12/10/2012 09:57, Pieren wrote:
On Fri, Oct 12, 2012 at 10:15 AM, Simone Saviolo
<simone.savi...@gmail.com> wrote:
I think that contact info of an amenity (allow me to group shops,
restaurants, bars, and companies under that umbrella just for a minute)
should be considered all of equal importance.
They are amenities that stay for long time like post offices. And
amenities that are changing very fast like shops and restaurants. I
think OSM should concentrate on the first. OSM is not a google-like
bot scanning the web to build automatically a yellow-page db.
Surveying things like commercial streets are inevitably outdated
quickly. I would say that "amenities" that cannot be found in
wikipedia should be considered as commercial advertising.
Pieren
_______________________________________________
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging
Hi Pieren
While the purist inside me likes the idea of keeping OSM as up to date as possible I must disagree with the idea off de-prioritising amenities that could be considered as commercial advertising from
OSM. In fact I believe that almost the opposite approach is constructive to OSM in the longer run for the following reasons.
For OSM data to reach a tipping point and become a tool and resource that is widely known and used by the public in the way that Google maps is then it needs to offer information on a similar scale.
The general public don't just want a map to find where they are geographically, they expect this as standard these days, they want to know in detail about the things around them; what they can do,
where and how they can do it. Data on shops, hotels and general businesses is exactly the sort of thing people are looking for in this respect. As such OSM must offer these locations for other people
in order to build tools and apps off them. Providing information links on top of these locations as suggested in Alex's original post is only going to extend the usefulness and I believe is the right
way to do it.
Linking the extra none mapping info rather than try and provide it all as key value pairs in OSM is a great approach and fits neatly in with Tim Bernard Lees talk on the expansion of the web as a form
of linked data ( http://www.ted.com/talks/tim_berners_lee_on_the_next_web.html ). For example if we start adding tags with hard links to review sites for a particular place it might even be possible
to provide a positive feedback loop where a site that combined the OSM geo data and the review site data could then scan the link to see if there where any reviews marking the place as closed and in
turn feed this back into OSM and remove or update the appropriate node if the business closed down (a little far fetched but you have to dream). Drawing up a simple schema that avoids namespace
collisions and the like early on seems a good way to go (I would encourage Alex to add the info in his email to a page on the wiki).
There may be a number of dedicated mappers out there that keep their local areas up to date as far as shops and day to day amenities but we need to reach a tipping point where it becomes a competitive
disadvantage for a business not to have their information in OSM due to the large number of services using OSM data for the public good. As such I think we should be tagging businesses as fast as we
can. If people start turning up at a business find the wrong one they are more likely to update it themselves or ask the shop keeper who has good incentive to change it than if they never found it in
the first place and never let the shop keeper know about it.
As I said before we should be driving towards building a service with a strong commercial intensive to free data under our licence if we are ever going to compete with the Google and other
propitiatory mapping services. Without dedicated mappers starting the ball rolling we will be forever stuck in catch 22.
Cheers - Richard
_______________________________________________
Tagging mailing list
Tagging@openstreetmap.org
http://lists.openstreetmap.org/listinfo/tagging
