[OSM-dev] Introduction

[Graham Jones]

Waldemar,

It is good to see that you are interested in participating in GSoC
with OpenStreetMap!

As you say in your note, this scope project has the potential to be very
wide, but it will be important to set some achievable goals to turn it into
a suitable project for GSoC.

The basis will be to establish a 'render on demand' web interface, and as
you say a Danjo based interface like that used for Maposmatic looks
promising, and gives very professional looking results.  It has also been
done already, so should not be too difficult to set up, leaving you to
concentrate on providing options for the user to customise their maps.

The key to making the project a success will be to define what rendering
options you provide to the user.   The sort of things that you could
consider include (in no particular order):

   - Map projection (you could use a custom map projection to give the most
   accurate map for the area being drawn) (I am not sure if anyone does this at
   the moment?)
   - Road colours (either completely selectable, or you could define
   different styles for different countries based on local conventions)
   (Similar to the http://www.cloudmade.com style editor, but for paper
   output) .
   - Contours / hill shading (like http://www.opencyclemap.org or
   http://www.free-map.org.uk).
   - Options to emphasise cycling/walking routes (again like
   http://www.opencyclemap.org).
   - Points of Interest (like http://www.townguide.webhop.net) - select
   which to display, and add index.  Need to deal with which icon to use for
   the POIs, and deal with overlapping icons.
   - Output format (e.g. PDFs of various page sizes and resolutions).

Arranging for high resolution raster output will mean either collaborating
with the mapnik GSoC project that will add output resolution as an option to
mapnik (assuming someone takes it on!), or your code will have to write the
style file to take account of the required output resolution.

Well, that is the sort of thing that I would think about (but the above list
may be too long for a GSoC project, so I would recommend choosing a few as
your main project with some others as "if time allows" options, but
structure your code so that it can be extended in the future to include
other options without a major re-write.

I hope this helps - others may have other suggestions or comments?

Regards


Graham.

On 22 March 2010 20:02, waldemar quevedo <waldemar.quevedo at gmail.com> wrote:

> Hello OSM Community! I thought about introducing myself properly.
> My name is Waldemar Quevedo, and I'm a student at the Tec de Monterrey
> in Guadalajara,Mexico. On my last year of university I have been focusing
> on learning
> artificial intelligence and computer graphics and also did an internship
> at a startup here in Guadalajara,Mexico were working on FatStudent,
> a web application to help students from London to find part-time jobs.
>
> One of the core parts of FatStudent is a Google Maps Flex/Rails application
> for
> browsing the jobs [0]. While working on this, me and the FatStudent team
> discussed
> what were some of the limitations of using Google Maps, mainly because
> they're not free as in speech, and this got me interested on the
> OpenStreetMap project.
> Then I realized that it could get even more interesting if I go ahead and
> participate
>  on this year's GSoC by helping OSM!
>
> Reviewing this year's GSoC ideas, there was an interesting discussion
> regarding
> the 'Easy Printable Output'. There has also been some discussion regarding
> this project
> on the Mapnik side of things (Mr.Springmeyer has been very kind on
> providing info
> on what needs to be done to improve Mapnik regarding the subject on
> Mapnik's Trac)[1].
> I'm interested on creating a web application that takes advantage of the
> new developments
> from Mapnik and integrate the ideas of other projects that have achieved
> something similar
> (townguide, Maposmatic with ocitysmap), to make it easier to print
> customizable maps.
>
> Since Mapnik has already some Python bindings, and learning from Maposmatic
> approach,
> Django would be a good option to develop the application.
>
> Anyway, there is a lot to be thought with respect to this idea and its
> scope
> could get very broad. What I'm thinking is for now to work on setting up a
> service based upon
> what Graham Jones has already done with townguide[2], and creating a
> service a la Maposmatic
> but for POIs while discussing with the community which features could add a
> greater value
> to the project. (Themes for the maps like the ones provided by Cloudmade
> could be great I think)
>
> Please feel free to give your suggestions or tell me if I am doing it
> wrong,
>
> - Waldemar Quevedo
>
> [0]: http://www.fatstudent.co.uk/jobs/map
> [1]: http://trac.mapnik.org/wiki/GSOC2010/Ideas#BetterPrintSupport
> [2]: http://www.townguide.webhop.net/
>
> _______________________________________________
> dev mailing list
> dev at openstreetmap.org
> http://lists.openstreetmap.org/listinfo/dev
>
>


-- 
Dr. Graham Jones
Hartlepool, UK
email: grahamjones139 at gmail.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/dev/attachments/20100322/b6714737/attachment.html>