[OSM-dev] OSM-3D tagging info

Matthias Meißer digi_c at arcor.de
Wed Dec 21 18:29:18 GMT 2011


Am 16.12.2011 14:17, schrieb Jaime Crespo:
> El 16/12/2011 11:15, "Frederik Ramm" <frederik at remote.org
> <mailto:frederik at remote.org>> escribió:
>  >
>  > Ander,
>  >
>  >
>  > On 12/16/11 10:09, Ander Pijoan wrote:
>  > Such information is almost certainly not suitable for OSM.
>  > Under no circumstances should you import data for all of Spain while
> sitting at your desk - if at all, then you can help local users to
> import cadastre data where there's interest and where there *are* local
> users to begin with.
>
> Even if I agree that I like user contributed content above cold imports,
> when I learnt about the availability of this dataset, I knew it could be
> very useful. We are talking here about official administrative staff
> (similar to boundaries) which cannot be surveyed: landuses and building
> shapes. Buildings could be traced from ortophotography, but with the
> relation stuf it is both boring and very imprecise (compared to
> 1-metre-acurate always-updatable import). Sincerely, I prefer people
> working in surveying amenities, mountain tracks, names, routing, etc.
> This import have nothing of this.
I understand, that this import is a good chance for you, to get 
informations, that are hard to trace exactly.
Here in Rostock (city in Germany), we got a similar import, maybe you 
can benefit on our experiences:
http://wiki.openstreetmap.org/wiki/Import/Catalogue/Houses_Rostock
In short it was a great way to get a lot of details on the map (which 
makes working with data more difficult by the way ;)). The positions had 
to be altered for legal reasons, so the buildings weren't exatcly 
rectified. Later by adding housenumbers, it was a great help, that the 
import included all sperated parts of a house, so you could double check 
if your number records were exact.
We did a manual import, so everybody picked up a piece of cake and 
imported the data to the DB, while he made sure, that there were no 
heavy overlappings. So basicaly we added the geometry, the hight and the 
building use, to get a good help for mappers.

Now more than two years later (and with Bing available) I'm not that 
sure if it was wise:
-no real exact positions (but good enough)
-no good splitted parts/heigts for 3D rendering
-some outdated informations, hard to find
-some to generelized informationos/categories that can be modelled 
better with OSM
-mostly to detailed buildings (balconies,...)
-hard to motivate people to add further attributes to the buildings 
(roof, colours ...)

+good for est. housenumbers
+great help for remembering amenities and stuff by non mappers (osmbugs.org)
+super detailed informations (power transformators,...)

So my suggestion is, to do a import only together with a crew on the 
ground, to make sure, that they accept this 'mechanical' adding of 
informations and to make sure that they don't override the hard manual 
of all of you. And of course to work on this data if the import was done...
>
> I, as one of the "local users", am the one that is fearing any problem
> and checking progresses. But I want also thank Ander for his work,
> because he is doing all the hard work that nobody had time to do for
> this import (and even answering all suggestions).
Definitly! We all aren't talking to him, to stop any action, but to help 
you to make a wise decission so OSM comes better and better in your area :)

bye
Matthias
(user:!i!)

>
> Anyway, I am the first person that could see problems for future
> contributors. But my beliefs are that they should be addressed within
> the editors' scope: things like better handling of layers and relations
> for buildings and landuses. 90% of the corrections I point to new users
> have to do with multipoligons.
>
> --
> Jaime Crespo
>
>
>
> _______________________________________________
> dev mailing list
> dev at openstreetmap.org
> http://lists.openstreetmap.org/listinfo/dev




More information about the dev mailing list