[OSGeo-Standards] [TC-Discuss] [OSGeo-Discuss] Live DVD and OGC standards - GeoPackage
sfkeller at gmail.com
Tue Jul 9 15:53:59 PDT 2013
Vector tiling is about performance and styling.To me it's in an
experimental technological state. Citing Mapbox spec.  "Vector
tiles store a serialized version of the internal data that Mapnik uses
when rendering maps.". AFAIK it's still controversial if tiling before
hand is really needed as persisted data.
There are still many open issues which make it premature to be
* Handling of areas is rather unclear: What if an area is clipped or
larger than a tile? See also notes on Area Layers at .
* Labels at edges need to take extra care to keep things seamless
("Very large markers, very large text sizes, and very long labels
should be avoided" ).
* Vector tiles omit parks, water bodies, POIs, etc that are mapped
only as nodes (?).
2013/7/9 Pat Cappelaere <pat at cappelaere.com>:
> Hey Stefan,
> Would you care backing up your statement with some facts and explanation for the rest of us?
> I have been researching the topic and I have very impressed with the speed achieved (especially for mobile devices). You get immediate benefit with OpenStreetMap and GeoJSON.
> I am not sure what a desktop exchange format means and why vector-tiles would not be suited for such a thing.
> On Jul 9, 2013, at 3:22 AM, Stefan Keller <sfkeller at gmail.com> wrote:
>> Hi Pat
>> No, vector-tiles is'nt the kind of vector format one would wish as a
>> replacement of Shapefiles (aka "desktop exchange format") .
>> Though, MapBox made a valuable input to GeoPackage by proposing
>> MBTiles as a "desktop exchange format" for raster data.
>> Yours, Stefan
>> 2013/7/9 Pat Cappelaere <pat at cappelaere.com>:
>>> I was wondering what was going on with GeoPackage as well.
>>> There seems to be some better alternative with Mapbox TM2 anyway
>>> On Jul 8, 2013, at 6:04 PM, Cameron Shorter <cameron.shorter at gmail.com> wrote:
>>>> On 9/07/2013 6:59 AM, Stefan Keller wrote:
>>>>> Referring to the initial question, GeoPackage with SQLite/Spatialite
>>>>> "format" would have a chance to become "the Shapefile of the future".
>>>>> The OGC just recently had the chance to adopt this existing encoding -
>>>>> but unfortunately voted against in favor of an own spec.
>>>> Stefan, I missed the discussion related to this decision. (I actually didn't realise that the decision had been made).
>>>> Could you please recap why "GeoPackage was not considered as the shapefile of the future".
>>>> Was it discounted outright? Or identified as too immature to adopt yet?
>>>> Cameron Shorter
>>>> Software and Data Solutions Manager
>>>> Tel: +61 (0)2 8570 5050
>>>> Mob: +61 (0)419 142 254
>>>> Think Globally, Fix Locally
>>>> Geospatial & Data Solutions enhanced with Open Standards and Open Source
>>>> TC-Discuss mailing list
>>>> TC-Discuss at lists.opengeospatial.org
>>>> This is an auto-subscribe list. All OGC members are encouraged to maintain a subscription to this list. You may change your delivery options in the "Email Subscriptions" portion of "My Account" in the OGC Web Portal.
More information about the Standards