[talk-au] New CC-BY datasets due Monday 28 September on Government 2.0 Taskforce website

[Brendan Morley]

On Sat, 3 Oct 2009 23:25:44 +1000, John Smith wrote:

>2009/10/3 Brendan Morley <morb.gis at beagle.com.au>:
>> How do you mean "consumed"?

>Some land owners have taken over the land when it's probably still
>crown land. For example:

>http://maps.google.com.au/?ie=UTF8&ll=-26.158115,152.64636&spn=0.007184,0.013937&z=17

>Horswoord Road mostly doesn't exist, yet the land isn't listed as a
>property boundary at all, so the land has probably been squatted by
>nearby land owners.

Yes, that often happens when a row of lots has multiple road frontages.  The "back" frontage is often left au naturel, presumably less maintenance for the 
council.  But if you want to subdivide later, you don't have to resume somebody else's backyard as well.  Just build the "back" road when you actually need 
it.

If you look closer (allowing for the photographic offset in that area) there's a bushier strip of vegetation about where the road casement passes through.

That's one of the reasons why I'm tagging DCDB-derived roads with "note=DCDB indicates a right of way in this location. Needs a field survey to confirm 
highway type and actual alignment."  Sometimes the road formation does not exist!

In that case I'm actually intending to remove the highway tag completely, but leaving the way included in the database.  This would indicate that 
"somebody" evaluated that the legal right of way existed and that "somebody" visited the location and found no transportation infrastructure.  Maybe there is 
a highway tag after all that fits?  In some other datasets it's seems to be either called "unformed" or a "construction line" - not sure which definition fits here 
though.

Another alternative is to add an area rather than a way and tag as natural=grass or =wood (and maybe access=yes?)

>Yup, the errors are based on paper boundaries verse real world boundaries.

But the real world can be legally forced to match the paper boundary in this case.  Admittedly the absolute positioning can be out at times in the DCDB (I've 
seen a 15m error in a previous edition, since corrected) but the relative arrangements are meant to be a direct reflection of the thousands of individually-
lodged survey plans since the "dawn of time".

So there's "should" be nothing stopping you from traversing the length of Horswood Road, formed or not.  If you get challenged, just say you're intending to 
follow Horswood Road.  Having an easily-attainable copy of the DCDB map in your hand may help, which is another reason why having a CC-BY DCDB is 
such an epic win.


Brendan


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-au/attachments/20091004/cc231b95/attachment.html>