[gdal-dev] Vector Tiles in OGR

[Stefan Keller]

Hi all

No there's no standard about non-Webmercator usage of vector tiles.
TileJSON spec. leaves open how CRS is handled which was a wise decision.
So it's up to us to come up with a proposal.

But what's bad is, that TileJSON is not even up-to-date to describe
vector tiles. At least MBTiles spec. recently got updated to 1.3.
I'm a bit frustrated about Mapbox having "no time" not even to update
the MVT specs. It's just delaying innovation.

:Stefan



2018-02-13 10:39 GMT+01:00 Even Rouault <even.rouault at spatialys.com>:
> On mardi 13 février 2018 02:06:47 CET jratike80 wrote:
>
>> Even Rouault-2 wrote
>
>>
>
>> > On lundi 22 janvier 2018 23:40:55 CET Even Rouault wrote:
>
>> >> Hi,
>
>> >>
>
>> >> > P.S. @Even: Any news on a GDAL/OGR MVT reader?
>
>> >>
>
>> >> Yes, a new OGR MVT reader has just been committed in GDAL trunk:
>
>> >> http://gdal.org/drv_mvt.html
>
>> >
>
>> > And the writer part (for MVT and MBTILES) driver is now in trunk
>
>> >
>
>> > http://gdal.org/drv_mvt.html
>
>> > http://gdal.org/frmt_mbtiles.html
>
>> >
>
>> > Even
>
>>
>
>> Hi,
>
>>
>
>> At the moment we create MVT tiles from PostGIS with ST_AsMVT to suit with
>
>> the national gridset that is based on ETRS-TM35FIN (EPSG:3067) coordinate
>
>> system. How difficult would you consider to enhance the driver to split
>> and
>
>> write vector tiles according to other gridsets than the Web Mercator one?
>
>
>
> Not that much. But is there a standard way of deriving the tiling scheme for
> that CRS ? I guess not. You'd probably need to manually define the
> coordinate of the left-corner of the tile (0,0) in the CRS and the
> resolution of the tile at zoom level 0.
>
>
>
> Even
>
>
>
> --
>
> Spatialys - Geospatial professional services
>
> http://www.spatialys.com
>
>
> _______________________________________________
> gdal-dev mailing list
> gdal-dev at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/gdal-dev