ApacheCon talk proposal: API in cloud environments

[Martin Desruisseaux <ma...@geomatys.com>]

Hello all

Below is my proposal for a talk at the next ApacheCon in Montréal. The
talk will be partially about Apache SIS, but I think that the main part
could be of wider scope. Is there any comment, proposal for changes or
collaboration with other talk proposal?

    Martin

------------------------------------------------------------------------

The Open Geospatial Consortium (OGC) defines many international
standards that make interoperability possible between different
geospatial applications. Most standards are articulated around data
formats (Well Known Text, Geographic Markup Language, etc.) and web
services (Web Map Service, Web Feature Service, etc.) Those standards
enable data transfers between server machines, where data are stored,
and client machines, where data are typically processed (except with Web
Processing Service). But in a world of petabytes of Earth Observation
data, bringing data to the algorithm is not always practical; there is
sometime a need to bring algorithm to data instead. Google Earth Engine,
Open Data Cube, OpenEO and Amazon lambdas are examples of environments
where data are hosted and computed remotely. In those environments, the
OGC standards for data transfers do not apply as much as in the
"classical" situation. Consequently each cloud environment defines its
own, non-standard API for handling geospatial data.

This presentation will show how an old OGC effort — GeoAPI — could apply
to the cloud environment for some kinds of problems. An example of
remote execution using the same standard API in both Java and Python
languages will be shown. We will present advantage and inconvenient of
using a standard API. In particular the perceived complexity of
international standards should be weighted against the problems of
popular simple alternatives. Apache Spatial Information System (SIS)
will be presented as a GeoAPI implementation with a focus on new
features, some of them resulting from evolution in standards.

This talk is aimed to peoples having an interest in international
standards applied to geospatial data, their implementation in Apache
SIS, and how cloud environments may impact those standards. This talk
will introduce some advanced features like dynamic datums in spatial
referencing, but mainly as illustrations of the expertise contained in
international standards.

Re: ApacheCon talk proposal: API in cloud environments

[George Percivall <gp...@opengeospatial.org>]

Martin,

This looks great.  

Hope we get many excellent abstracts like this.

George


> On Mar 29, 2018, at 10:54 AM, Martin Desruisseaux <ma...@geomatys.com> wrote:
> 
> Hello all
> 
> Below is my proposal for a talk at the next ApacheCon in Montréal. The
> talk will be partially about Apache SIS, but I think that the main part
> could be of wider scope. Is there any comment, proposal for changes or
> collaboration with other talk proposal?
> 
>     Martin
> 
> ------------------------------------------------------------------------
> 
> The Open Geospatial Consortium (OGC) defines many international
> standards that make interoperability possible between different
> geospatial applications. Most standards are articulated around data
> formats (Well Known Text, Geographic Markup Language, etc.) and web
> services (Web Map Service, Web Feature Service, etc.) Those standards
> enable data transfers between server machines, where data are stored,
> and client machines, where data are typically processed (except with Web
> Processing Service). But in a world of petabytes of Earth Observation
> data, bringing data to the algorithm is not always practical; there is
> sometime a need to bring algorithm to data instead. Google Earth Engine,
> Open Data Cube, OpenEO and Amazon lambdas are examples of environments
> where data are hosted and computed remotely. In those environments, the
> OGC standards for data transfers do not apply as much as in the
> "classical" situation. Consequently each cloud environment defines its
> own, non-standard API for handling geospatial data.
> 
> This presentation will show how an old OGC effort — GeoAPI — could apply
> to the cloud environment for some kinds of problems. An example of
> remote execution using the same standard API in both Java and Python
> languages will be shown. We will present advantage and inconvenient of
> using a standard API. In particular the perceived complexity of
> international standards should be weighted against the problems of
> popular simple alternatives. Apache Spatial Information System (SIS)
> will be presented as a GeoAPI implementation with a focus on new
> features, some of them resulting from evolution in standards.
> 
> This talk is aimed to peoples having an interest in international
> standards applied to geospatial data, their implementation in Apache
> SIS, and how cloud environments may impact those standards. This talk
> will introduce some advanced features like dynamic datums in spatial
> referencing, but mainly as illustrations of the expertise contained in
> international standards.
> 
>