[GRASS-dev] GSoC 2013, GRASS symbology idea

Hello,

I’m Tereza Fiedlerová, student of CTU in Prague, Czech Republic. I’d like to apply for Google Summer of Code 2013. I was looking on Grass SoC Ideas 2013 and I am interested in Symbology/Cartography [1] ideas there. I’ve read a little bit about some d. Grass modules, especially about d.legend, d.vect.thematic, d.area.thematic and something about cairo driver.

In my project, I would like to write new module for creating legend for vector map (something similar as d.legend for raster map). The other part of project could be new module for thematic

maps (something like d.vect.thematic and d.area.thematic, but more complex). For a nice and correct graphical output I could do some improvements in cairo driver. The target is not to need ps.map module, when user want to do some graphical output.

I’d like to hear some comments on this topic and maybe also some other ideas.

Thank You,

Tereza Fiedlerová

Faculty of Civil Engineering

CTU in Prague

[1] http://grasswiki.osgeo.org/wiki/GRASS_SoC_Ideas_2013#Symbology_.2F_Cartography

Hi Tereza,

In these days I'm thinking to substitute the GRASS d.* modules with mapnik [0].
Mapnik is the state of the art for high quality rendering of maps,
support several types of export formats and It is widely used.

The advantages to adopt Mapnik are:
- High quality output, symbols [1], text [2], legend[3] etc are
already implemented;
- we will have less code to maintain because part of the job will be
done by another library;
- export into different formats, export as tiles for example should
make easier to integrate an interactive viewer to a ipython notebook
[4,5,6], or with other local and remote services;
- should be quite easy integrate the mapnik layers and rules with a
GUI, we only need to generate an XML and call mapnik to generate the
figure.

The disadvantages are:
- at the moment there is not a direct interface between mapnik and
grass, it is possible to access to the grass maps through the gdal/ogr
interface but it is not working for GRASS7 sofar but may be would be
possible to add a GRASS plugin to mapnik [7].
- we add a new dependence to GRASS and may be will be more difficult
to install GRASS on Windows/Mac platforms and it will increase the
size of the final installation (~43 Mb).

What do you think?

Best regards.

Pietro

[0] http://mapnik.org/
[1] https://github.com/mapnik/mapnik/wiki/PointSymbolizer
[2] https://github.com/mapnik/mapnik/wiki/TextSymbolizer
[3] https://github.com/mapnik/mapnik/wiki/Legending
[4] http://ipython.org/ipython-doc/dev/interactive/htmlnotebook.html
[5] http://lists.osgeo.org/pipermail/grass-dev/2012-November/060768.html
[6] https://trac.osgeo.org/grass/ticket/1921
[7] https://github.com/mapnik/mapnik/wiki/PluginArchitecture

On Tue, Apr 23, 2013 at 4:48 PM, Tereza Fiedlerová
<tfiedlerova@gmail.com> wrote:

Hello,

I'm Tereza Fiedlerová, student of CTU in Prague, Czech Republic. I'd like to
apply for Google Summer of Code 2013. I was looking on Grass SoC Ideas 2013
and I am interested in Symbology/Cartography [1] ideas there. I've read a
little bit about some d. Grass modules, especially about d.legend,
d.vect.thematic, d.area.thematic and something about cairo driver.

In my project, I would like to write new module for creating legend for
vector map (something similar as d.legend for raster map). The other part of
project could be new module for thematic
maps (something like d.vect.thematic and d.area.thematic, but more complex).
For a nice and correct graphical output I could do some improvements in
cairo driver. The target is not to need ps.map module, when user want to do
some graphical output.

I'd like to hear some comments on this topic and maybe also some other
ideas.

Thank You,

Tereza Fiedlerová
Faculty of Civil Engineering
CTU in Prague

[1]
http://grasswiki.osgeo.org/wiki/GRASS_SoC_Ideas_2013#Symbology_.2F_Cartography

_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-dev

Hi,

2013/4/24 Pietro <peter.zamb@gmail.com>:

In these days I'm thinking to substitute the GRASS d.* modules with mapnik [0].
Mapnik is the state of the art for high quality rendering of maps,
support several types of export formats and It is widely used.

well, to "substitute" is a strong word. In any case it would be
interesting alternative to `ps.map` and d.* + cairo driver approach.

Generally speaking very interesting idea and worth to work on that.
But still I would say that closer mapnik integration will not replace
native GRASS display driver or `ps.map` module.

Martin

--
Martin Landa <landa.martin gmail.com> * http://geo.fsv.cvut.cz/~landa

Hi Tereza,

2013/4/23 Tereza Fiedlerová <tfiedlerova@gmail.com>:

and I am interested in Symbology/Cartography [1] ideas there. I've read a
little bit about some d. Grass modules, especially about d.legend,
d.vect.thematic, d.area.thematic and something about cairo driver.

In my project, I would like to write new module for creating legend for
vector map (something similar as d.legend for raster map). The other part of
project could be new module for thematic
maps (something like d.vect.thematic and d.area.thematic, but more complex).
For a nice and correct graphical output I could do some improvements in

beside improving existing d.* modules or creating new (d.vlegend),
improving cairo driver (support advanced symbology for lines, areas,
etc) I would like to highlight other issues:

1) better integration of d.* modules in wxGUI. Current support for
legends and other decorations in wxGUI map display need to be
fundamentally rewritten and improved.

2) Extend `v.class` and related lib (not documented at all, fns need
some prefix, ...) [1]. Also integrate this module in wxGUI.

Martin

[1] http://trac.osgeo.org/grass/browser/grass/trunk/lib/arraystats

Martin wrote:

Generally speaking very interesting idea and worth to work
on that. But still I would say that closer mapnik integration
will not replace native GRASS display driver

Mapnik works very closely with PostGIS, so your GRASS+PostGIS
efforts may make this a very clean & easy (trivial?) path for
people who want to do Mapnik rendering of GRASS data.

happy to see lots of rendering options to fill the many niches,
and glad to see all this get more attention,

Hamish

ps- if you haven't already tried out TileMill, give it a look..