[mapserver-dev] "ambiguity" of gml_[item name]_type = Date

[Steve Lime]

Makes complete sense to me at first glance.

On Tue, Oct 22, 2019 at 3:48 PM Even Rouault <even.rouault at spatialys.com>
wrote:

> Hi,
>
> when serving vector through WFS currently, we can use the gml_[item
> name]_type
> to define the type of a field. Currently we have a value "Date" whose
> semantics is rather loose. All input backends (OGR, Oracle, MSSQL,
> PostGIS)
> use it whereas they have distinct types for Date (without time), DateTime,
> and
> Time (without date). MapServer having internally a single type makes that
> when
> using OGR output (for example GeoJSON), we'll have lost some information,
> and
> thus the OGR output currently prefers to use a OGR OFTDateTime to avoid
> potentially losses. But "2019-10-22T00:00:00Z" is not semantically
> equivalent
> to "2019-10-22".
> So my proposal would be to change (define) Date to be a Date without time
> information, and add DateTime and Time types.
>
> Thoughts ?
>
> (note: some backends like Oracle and PostGIS have distinctions between
> DateTime with or without timezone, but we probably don't need to go to
> that
> level of subtelty)
>
> Even
>
> --
> Spatialys - Geospatial professional services
> http://www.spatialys.com
> _______________________________________________
> mapserver-dev mailing list
> mapserver-dev at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/mapserver-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/mapserver-dev/attachments/20191022/70d9efed/attachment.html>