On 2014-10-06 06:48, Marc Gemis wrote:
After 3.5 years and 34.000 addresses, I have to admit that an import
is the only way to get addresses in OSM fast.
But is the conclusion that we have to make that a crowed-sourced model
for map making failed ? That we have to move to an import of
third-party databases model ? One were the source is maintained
elsewhere and OSM is nothing more than a combination of all those
databases ?
I don't think anyone is claiming that. You seem to be fond of
exaggerations.
Imports are not bad in principle and OSM has not failed in places where
there are no imports.
Are we going to a Google-like model were "ordinary" mappers cannot
change stuff and see the result immediately ? Will everything have to
be "approved" by the original data source ? Are those the kind of
questions that the OSM-future group is thinking about ?
I sure hope not.
Another, related question is how do you want to deal with updates on
BAG data now. Can a (tourist) mapper change the building outline or
remove it ? Or will it be reverted with the next BAG update when the
change is not reported back to BAG ? That's what I've read on the
forum when I reported the mistake.
OSM can be changed by everyone who has registered. So yes, a tourist
mapper can change BAG data. AFAIK there is no process in place for
updates of BAG data.
Regards,
Maarten
On Sun, Oct 5, 2014 at 9:56 PM, Johan C <osm...@gmail.com> wrote:
Marc, I agree with Maarten. Let's hope that our address data helps
users appreciate OSM apps more. And that these apps are smart enough
to draw more users into mapping.
You have a good point on maintenance versus the number of mappers.
On the BAG data it's luckily quite simple: thousands of paid people
are constantly working on updating that data. Challenge for the OSM
community is to get these updates efficiently into OSM. That's being
worked on, but it needs some programming on a JOSM plugin.
2014-10-05 21:39 GMT+02:00 Marc Gemis <marc.ge...@gmail.com>:
For maintenance, the way we do it now doesn't work IMHO. At least
not with only a handful of (40-50) mappers in both Belgium and The
Netherlands. We can't keep up with the changes.
IMHO The amount of work to add new data is approx. the same as
verifying and keeping it up to date.
I wonder how many mappers we really need to keep a country such a
The Netherlands up-to-date. Not only on addresses, but also on POIs
(companies, shops, benches, traffic signs, schools, sidewalks,
paths, access permissions, etc. etc.)
I also see a "cultural" difference between The Netherlands (having
done 2 major imports already) and Belgium (more averse towards big
imports). So it's also interesting to see how this has an impact on
those things.
Happy mapping
m
p.s. Hopefully you can drop by some time and remove some of those
phantom buildings the AND import introduced in the north of Belgium
:-)
On Sun, Oct 5, 2014 at 9:09 PM, Maarten Deen <md...@xs4all.nl>
wrote:
On 2014-10-05 20:58, Marc Gemis wrote:
You'll beat us on numbers, that's true.
Maybe the problem that I see is not so much the imports, but the
maintenance of all that data (imported or manually added).
Who is going to see all those mistakes, changes, etc. when all the
data is there ? The one that I saw was a building in a forest.
That's
easy to spot when it is demolished. But how will you notice that
the
shape of a building is incorrect when there are thousands of
buildings?
How will you do this if there has been no import? I see no
difference in maintaining OSM data after import or after manual
mapping. In both cases the environment can change and it will only
be changed in OSM if someone notices.
The only difference is that after an import there is more data. If
you think maintaining more data is a bad thing, then we all have to
stop entering in OSM.
At this moment, most communities are still adding data, the Dutch
import speeds up this process, so the Dutch community will be in
the
next phase before the others. You can rely on other parties (BAG)
for
the updates, but will they catch mistakes ? The building I'm
talking
about is demolished 2 or 3 years ago. Why was the BAG not updated
yet
? Will people be looking for mistakes when all data seems to be
there
?
Again, I see no difference with manually entered data. There too
you can have features that are years out of date. If no one notices,
it is no problem. If a mapper notices it, it will be fixed.
So the challenge is to find people that want to find those
mistakes.
Less interesting than adding new data.
I'm looking forward to see how the Dutch community is going to
tackle
this.
As said: the same as now. If you are not interested in maintaining
the data you entered then that data will be outdated too. I'm just
so glad we have a nice map in the Netherlands where I only need to
fix minor things so I can spend time mapping somewhere else.
Regards,
Maarten
On Sun, Oct 5, 2014 at 8:10 PM, Johan C <osm...@gmail.com> wrote:
<Anyway this is not doing anything positive for my feelings on
imports.>
In terms of having data for a routing engine (like OsmAnd) a
definition can be that any missing address in a country is an
error.
The number of missing addresses in the Netherlands is calculated
recently: on a total of approx. 8,5 million addresses 60.000 are
missing at the moment (0,7% error). How does that compare to
Belgium
or Germany?
2014-10-05 19:49 GMT+02:00 Marc Gemis <marc.ge...@gmail.com>:
On Sun, Oct 5, 2014 at 4:49 PM, Johan C <osm...@gmail.com> wrote:
The BAG should contain the correct building outline, since this is
Cadastral information, nowadays updated very often. But as any
database, the BAG might incidentally have errors. Satellite imagery
though is at risk of being well outdated. So in these cases it's
possibly best to have groun truth info to determine the correct
building outline.
Funny that Florian found an error, and that I also found 2. And
that
for holiday mappers. :-)
Anyway this is not doing anything positive for my feelings on
imports.
regards
m
_______________________________________________
Talk-nl mailing list
Talk-nl@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-nl [1] [1]
_______________________________________________
Talk-nl mailing list
Talk-nl@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-nl [1] [1]
Links:
------
[1] https://lists.openstreetmap.org/listinfo/talk-nl [1]
_______________________________________________
Talk-nl mailing list
Talk-nl@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-nl [1]
_______________________________________________
Talk-nl mailing list
Talk-nl@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-nl [1]
_______________________________________________
Talk-nl mailing list
Talk-nl@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-nl [1]
_______________________________________________
Talk-nl mailing list
Talk-nl@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-nl [1]
Links:
------
[1] https://lists.openstreetmap.org/listinfo/talk-nl
_______________________________________________
Talk-nl mailing list
Talk-nl@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-nl
_______________________________________________
Talk-nl mailing list
Talk-nl@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-nl
