Hello,

I don't have nearly enough information to do an honest assessment of
your project and such a design program is a huge amount of work so you
will have to take that on yourself.

The good news is that you are on the right path---there is a mix of
technology which you could assemble to solve your needs. The bad news is
that you are a year or two too early so this work is currently harder
than it will eventually be.

The design concerns will then revolve around the timeframe of your work.
If your project is a decade long effort, then I'd encourage you to get
into the heart of GeoAPI/Geotools/uDig. If you need a quick and dirty
for the next six months, these projects will require a bunch of work and
the *JUMP* projects (OpenJUMP...) may be easier to get up to speed
with. 

For user interfaces, you have quite a few choices uDig, *JUMP*, or
thuban (in python) as well as many others. For the back end, the PostGIS
database is the place to work if you have vector data of any significant
size. Interface to C/C++ is relatively easy but I have *no* idea how to
link to VB. For .Net, the spring will be critical since Java will become
free software and that will affect the IKVM project. I suspect one could
program in .Net and have the whole Geotools stack running on an IKVM
instance---a massive memory hog but one could get that to work. There
was an effort to port Geotools to C# but I can't imagine that effort
could last given the scale of the work involved.

The canonical source for projects you can evaluate for your use is the
freegis.org web site. There are projects in many languages and in
various states of development. For example, I just saw that Thuban had a
new release

An aside on licensing, I suspect the 'free software' issue will prove
more important than the 'open source' for the success of your project.
While, for you, reading the source will be important to evaluate
algorithms and understand code, for your users the freedom to re-mix
will be critical. In that context, having code under LGPL may be
important; the GPL is lovely but does prevent companies from mixing the
old business style 'we sell added-value software' with the new style 'we
support your particular uses of this free software.' You get the idea.

Your needs will be met, eventually, by the GeoAPI interfaces, the
Geotools library implementing those interfaces, and the uDig user
application. GeoAPI (the packages in the opengis.org namespace) will
eventually become the only thing that users need to learn: the
interfaces define geospatial features, how to access their content, and
how to manipulate them. Geotools is still a mix of quick-and-dirty
functionality and polished design. uDig is a great foundation for user
interfaces which allows easy user selection of objects and then for
developers to hookup, relatively simply, a small operation to act on
that selection. uDig aims to provide all of the 'common GIS
functionality' you describe buffers, merges, joins... So GeoAPI/Geotools
will be like arc objects eventually, but for now has a bigger scope and
is focused on implementing the ISO geospatial model so is taking its
time to develop a comprehensive solution.

The next year of work on GeoAPI/Geotools/uDig may get us to the point
where we have all the pieces you need available through elegant API. The
GeoAPI interfaces to geometric elements are currently being implemented
in Geotools. uDig is currently seriously evaluating user workflows for
spatial operations. (The operations themselves are well understood and
basically implemented, we are currently struggling to see what the best
way will be for users to ask for the operations.) 

best of luck,
--adrian


On Mon, 2007-04-02 at 18:15 -0700, [EMAIL PROTECTED] wrote:
> Hi, Adrian. I saw a posting by you concerning which open source
> product/technology was useful for what purposes. I would like to tell
> you about what I am attempting to do in the hopes you can help me
> decide which road to take.
> 
> I am trying to develop a standalone geospatial application for Windows
> that has the capacity to allow users to select point, lines, and
> polygon features on a map. This selection and its numeric attributes
> will be sent as an array(s) to already-existing VB/C/C++ functions
> that will compute various calculations and return them to the user
> either as a thematic map or tabular file.
> 
> One of the things I want to be able to do is spatial analysis
> on-the-fly, namely, common GIS operations such as intersect,
> attribute/spatial joins, buffering, and perhaps others such as
> raster-to-feature conversions, etc. I am becoming familiar with the
> multiplicity of open source GIS technologies, but I am not sure if it
> is developed enough to provide an API or SDK to programatically carry
> out such operations. Of course, ESRI already has all this
> functionality through ArcObjects and its .NET framework, but I want to
> take the open source route. :-)
> 
> Your thoughts and comments will be immensely helpful.
> 
> Regards,
> Juan Declet
> 


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Geotools-gt2-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geotools-gt2-users

Reply via email to