Re: [OSM-talk] Large cadastral polygons

[Martin Koppenhoefer]

sent from a phone

> On 12. Jun 2020, at 17:39, Paul Johnson  wrote:
> 
> Not just that, but cadastral type tags probably shouldn't be spanning large 
> areas to start with.  If your landuse or landcover polygon is crossing an 
> unclassified or higher highway, you're probably making a big mistake.  
> landuse=residential is NOT a substitute for place=neighborhood (something I 
> see a lot).


+1, I would go even further, and at least in countries like Germany (roads not 
on parcels), no urban landuse should ideally contain public roads. This makes 
editing much easier, multipolygons are not or very rarely needed for landuse 
mapped in this way. Easier to understand and more detailed. 
Then, use place for settlement parts.



> 
> If you have reasonably high resolution (like, at least Esri Clarity in the 
> midwest) imagery available, it's often possible to see lot stakes and 
> fenceposts on property boundaries...that seems like a reasonable edge for a 
> landuse polygon.


+1

Cheers Martin 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Large cadastral polygons

[Paul Johnson]

On Fri, Jun 12, 2020 at 6:32 PM Warin <61sundow...@gmail.com> wrote:

> On 13/6/20 1:37 am, Paul Johnson wrote:
>
> On Fri, Jun 12, 2020 at 10:28 AM Florian Lohoff  wrote:
>
>> On Fri, Jun 12, 2020 at 02:14:15PM +0200, Mateusz Konieczny via talk
>> wrote:
>> >
>> >
>> >
>> > Jun 12, 2020, 13:59 by f...@zz.de:
>> >
>> > > Changeset envelopes which span more than 100s of km² are broken.
>> > >
>> > Except cases where you edit/delete already created huge objects or you
>> create
>> > huge object that actually should be created.
>> >
>>
>> These types of objects should be pretty exceptional. I try to split
>> landuses to sub 1km² because i also feel the pain for
>> rendering tiles. As soon as someone touches those areas you invalidate
>> tons of tiles. So breaking this down also benefits us long term
>> concerning workload on the tile servers.
>>
>
> Not just that, but cadastral type tags probably shouldn't be spanning
> large areas to start with.  If your landuse or landcover polygon is
> crossing an unclassified or higher highway, you're probably making a big
> mistake.  landuse=residential is NOT a substitute for place=neighborhood
> (something I see a lot).
>
>
> Hummm.. the following military areas are used as a rocket test firing
> area. Spiting them up into 1 km square areas would be a lot of work! Total
> area is approximately 122,188 km².
>

Odds are that's a single parcel.  Nobody's expecting that to get broken
up.  But when you have landuse=residential spanning entire city districts,
or entire towns and crossing other landuses like retail, commericial,
industrial, highway and railway, that's clearly a misuse.
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Large cadastral polygons

[Warin]

On 13/6/20 1:37 am, Paul Johnson wrote:
On Fri, Jun 12, 2020 at 10:28 AM Florian Lohoff > wrote:


On Fri, Jun 12, 2020 at 02:14:15PM +0200, Mateusz Konieczny via
talk wrote:
>
>
>
> Jun 12, 2020, 13:59 by f...@zz.de :
>
> > Changeset envelopes which span more than 100s of km² are broken.
> >
> Except cases where you edit/delete already created huge objects
or you create
> huge object that actually should be created.
>

These types of objects should be pretty exceptional. I try to split
landuses to sub 1km² because i also feel the pain for
rendering tiles. As soon as someone touches those areas you invalidate
tons of tiles. So breaking this down also benefits us long term
concerning workload on the tile servers.


Not just that, but cadastral type tags probably shouldn't be spanning 
large areas to start with.  If your landuse or landcover polygon is 
crossing an unclassified or higher highway, you're probably making a 
big mistake. landuse=residential is NOT a substitute for 
place=neighborhood (something I see a lot).



Hummm.. the following military areas are used as a rocket test firing 
area. Spiting them up into 1 km square areas would be a lot of work! 
Total area is approximately 122,188 km².


Ways 544274312, 436690530 and 436098551.


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

[OSM-talk] Large cadastral polygons

[Paul Johnson]

On Fri, Jun 12, 2020 at 10:28 AM Florian Lohoff  wrote:

> On Fri, Jun 12, 2020 at 02:14:15PM +0200, Mateusz Konieczny via talk wrote:
> >
> >
> >
> > Jun 12, 2020, 13:59 by f...@zz.de:
> >
> > > Changeset envelopes which span more than 100s of km² are broken.
> > >
> > Except cases where you edit/delete already created huge objects or you
> create
> > huge object that actually should be created.
> >
>
> These types of objects should be pretty exceptional. I try to split
> landuses to sub 1km² because i also feel the pain for
> rendering tiles. As soon as someone touches those areas you invalidate
> tons of tiles. So breaking this down also benefits us long term
> concerning workload on the tile servers.
>

Not just that, but cadastral type tags probably shouldn't be spanning large
areas to start with.  If your landuse or landcover polygon is crossing an
unclassified or higher highway, you're probably making a big mistake.
landuse=residential is NOT a substitute for place=neighborhood (something I
see a lot).

If you have reasonably high resolution (like, at least Esri Clarity in the
midwest) imagery available, it's often possible to see lot stakes and
fenceposts on property boundaries...that seems like a reasonable edge for a
landuse polygon.  Also makes it easier to deal with landuse changes in the
future, as lots rarely change shape, more commonly get subdivided or
consolidated; that's a trivial change.
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk