Hello all
An OGC meeting happened in Toulouse (France) November 18th to 21th. Key
slides from the meeting are available at [1]. Those slides are often
extracted from longer presentations; they give an idea of what has been
discussed, but if a slide seems particularly interesting it may be
necessary to get the full presentation from the relevant OGC working
group for more context. Below are my personal notes covering only a tiny
part of the meeting.
GeoTIFF standard
A test suite is under development and is currently about 80% done. The
tests will be available as OGC CITE tests ("org.opengis.cite" Java
package). This is relevant to Apache SIS among others, which is
developing a GeoTIFF reader.
For future GeoTIFF evolution, a project has been created on GitHub [2].
GeoTIFF 1.1 was aimed to fix issues regarding EPSG codes and clarify the
definition of vertical coordinates, while doing as little changes as
possible compared to 1.0. GeoTIFF 1.2 or maybe 2.0 may apply some of the
proposals listed on GitHub.
Coordinate Reference System (CRS)
Now that ISO 19111 and ISO 19162 (Well Known Text format) are completed,
the next task is a standard file format for the exchange of gridded
geodetic data. ESRI presented the current state of their work, based on
netCDF 4 format (almost same as HDF5 format). They started this work a
few years ago. The OSGeo PROJ project also started a work with similar
goal after raising about 10,000$ in crowdfunding. The timing between OGC
and PROJ effort is a coincidence (OGC started this effort because it was
a next task planned after ISO 19162 completion, not because of PROJ
activity). PROJ did a comparisons of some file formats and selected
GeoTIFF [3], but my personal analysis is that more discussion on [3]
would be needed (many reported GeoTIFF advantages apply also to HDF5)
and that their selection was based largely on PROJ/GDAL convenience,
which is of course a valid reason for them. For an international
standard, if ESRI continues to propose a format based on netCDF/HDF5, I
would prefer to explore such approach.
OSGeo activity: lexicon glossary
OSGeo reported their activity in the creation of a glossaries of terms
across projects. The problem they want to solve is that different
projects use inconsistent terms. They want to maintain a master OSGeo
glossary for harmonizing a working set of terms between OSGeo projects
and OGC/TC211. ISO already has a glossary with about 1000 entries
translated in 14 languages [4]. OSGeo uses a sub-domain of that glossary
[5]. Bugs can be filled on GitHub for ISO entries [6] or OSGeo entries [7].
Architecture group / Big data / Datacube
Abstract specification is drafted for tiling in two dimensional plane in
Euclidean space [8]. This draft specification tries to put in common the
practice of various standard (GeoPackages, etc.). It handles commonly
used tessellations: triangles, rectangles or hexagons.
I mentioned in a previous "Report on OGC meeting" email that a paper on
big data community practice is being drafted at OGC. Since that time,
the group agreed to said that current paper represents the view of
subset of the community and that other approaches will be explored in
future work. One question to discuss further is how to put analytic
(e.g. interpolation) in big data. One approach is to use Web Processing
Service (WPS). Others take inspiration for example from geoxarray [9].
Another approach is ISO/IEC 9075-15:2019 — Multi-dimensional arrays
(SQL/MDA) [10], a new standard providing some functionalities similar to
Web Coverage Processing Service (WCPS). That standard allows to perform
some analytics on a datacube stored in a SQL database.
An Indexed 3D Scene Layers (I3S) specification is maintained by ESRI
[11] and published as an OGC "community practice" document (not a
standard). Next work in progress is Point Cloud Scene Layers.
Moving features: JSON encoding
There is two closely related OGC documents for moving features in JSON:
* MF-JSON trajectory encoding best practice paper: this document keeps
the MF-JSON file as a valid GeoJSON file.
* MF-JSON prism encoding standard: this document adds more
functionalities that can not be represented in a JSON file.
The MF-JSON prism encoding has been approved as an OGC standard and
should be published soon on [12].
GeoAPI
The web site has been revamped [13]. We abandoned the use of Maven for
generating the web site, which make easier to focus on more important
information. Developments of Python interfaces is continuing, together
with a Java-Python bridge for connecting any GeoAPI implementations in
those languages. A Java Native Interface (JNI) binding for PROJ 6 is
available at [14]. Both Apache SIS and PROJ-JNI can be used together in
the same application if desired. It is even possible to mix them in a
coordinate operations.
Varia
Work on portrayal started during OGC TestBed 15. The intent is to
improve the entire framework of portrayal specifications. It includes a
web API for discovery of style resources. A style could be published and
offered for use by other systems.
A discussion during OGC closing plenary was that web services lost a lot
of metadata like (keywords, etc.).
Martin
[ 1] https://portal.opengeospatial.org/files/?artifact_id=91407
[ 2] https://github.com/opengeospatial/geotiff/projects
[ 3]
https://github.com/rouault/PROJ/blob/rfc4_remote_and_geotiff_grid/docs/source/community/rfc/rfc-4.rst#discussion-on-choice-of-format
[ 4] https://www.geolexica.org/
[ 5] https://osgeo.geolexica.org/ and https://osgeodev.geolexica.org/
[ 6] https://github.com/ISO-TC211/TMG
[ 7] https://github.com/geolexica/osgeo.geolexica.org
[ 8] http://docs.opengeospatial.org/DRAFTS/19-014.html
[ 9] https://github.com/geoxarray/geoxarray/
[10] https://www.iso.org/standard/67382.html
[11] https://github.com/Esri/i3s-spec
[12] http://docs.opengeospatial.org/
[13] http://www.geoapi.org/
[14] https://github.com/Kortforsyningen/PROJ-JNI
