[GRASSLIST:1249] Re: creating a desktop GIS application using GRASS

[Bill Dickinson Jr]

Hey there, Radim:

Looks like I need to look deeper into the GPL issue. If nothing else, 
if I get to the point of being able to release a commercial product 
but the GPL blocks me doing it with GRASS then there becomes a solid 
argument for changing the licensing. If the GRASS licensing does not 
make it easy to create a commercial product out of it then I would be 
forced to develop solely for ESRI, ENVI, and IDRISI. If commercial 
developers continue to run into the same problem then GRASS will not 
expand as us users would like and will continue to be a nifty little 
piece of freeware for the radical fringe to play with (my opinion, of 
course).

But, as you say, I won't give up. ;-)  We are still a year or two 
away from any commercial product using GRASS as the basic research 
still needs to be done, and a lot can happen in that time. If I get 
the initial research money, then I will be able to start asking for a 
larger pot from NASA that could include some GRASS development of our 
own. I imagine that would be helpful to the overall GRASS development 
effort no matter what else comes of it. ;-)

I am also getting ready to submit a research proposal to NASA for the 
visualization of environmental GIS data using GRASS and NVIZ. This 
will be mostly a research project to determine whether an open-source 
software solution is useful to NASA, though we are expecting to show 
a useful application of visualized GIS data in the daily operations 
of a federal installation.

I think these are the types of "success stories" with GRASS that need 
to happen before GRASS can become a more mainstream desktop GIS 
application. I imagine that there are already a lot of such stories 
and examples from people that have been using GRASS a lot longer then 
I have, they just need to get out to the masses somehow.

Bill

>It makes perfectly sense for me, but GPL does not allow that.
>Because GRASS libraries are covered by GPL, you cannot distribute
>proprietary SW linked to GRASS libraries.
>
>But don't give up immediately! I think that it is possible either:
>1) Fulfil old plan to release GRASS IO libraries under LGPL.
>    The problem is that we probably have to ask all developers
>    who contributed to such libraries after 'public domain / CERL' era.
>    Another question is how much will be allowed by GRASS developers
>    to put into 'IO' and if it is enough for any applications exept 
>import/export.
>2) Take old version of GRASS released as 'Public domain', release it
>    under some reasonable license (MIT/X) and do again all the bugfixing
>    porting etc. wich was done in current GRASS. Certainly it is not
>    possible to take improvements directly from current version, but
>    I think, that it is legal to compare old and current version to identify
>    what must be changed and then to code it without looking to current
>    GPL version, maybe not, maybe danger for future?
>3) Write completely new library under better license.
>
>Maybe start with 1) and later switch to 3).
>
>
>Radim
>
>
>On Wednesday 10 September 2003 15:06, Bill Dickinson Jr wrote:
>>  Not to muddy the waters any more on this topic, but I had a couple of
>>  pennies worth of thoughts to throw in since we are talking about
>>  creating a commercial product out of GRASS. ;-)
>>
>>  I have just submitted a proposal to NASA for a small research grant
>>  to apply remote sensing data to archaeology using GRASS as the GIS
>>  development environment to tie everything together (that's the
>>  Reader's Digest version of the proposal). The general process of
>>  creating archaeological predictive models within a GIS is pretty well
>>  documented, but our "proprietary" angle is in the advanced image
>>  processing and analysis technologies of our experts. Assuming that we
>>  can prove the basic concept of the research, we will then also submit
>>  for a larger pot of money from NASA to actually develop  a commercial
>>  product out of it. My company is very conservative with their
>  > business development dollars, but even they think my chance of
>>  funding is rather good and so they were willing to actually spend
>>  some internal money on the proposal effort. ;-)
>>
>>  Our basic theory for any commercial product is to develop, in ESRI
>>  terms, an extension to existing GIS software with our proprietary
>>  approach and techniques as the extra value added. The big target, of
>>  course, would be ESRI, but we will also look into ENVI and IDRISI, as
>>  well as our own GIS software, called DirectMet GIS
>>  (http://www.gst.com/prod_directmet.html). (Yes that is a shameless
>>  company plug - heh, forgive me.)
>  >
>  > Since we are doing the basic research in GRASS, and me being a
>  > current ESRI user trying desperately to get that monkey off of my
>  > back, we would also like to take a shot at a commercial GRASS
>>  product. In my mind, the way to sell a commercial GRASS product is
>>  along the lines of the OpenOSX cd-rom, just with a twist towards
>>  whatever your proprietary application may be - in our case,
>>  proprietary image processing and analysis technologies for
>>  archaeological purposes.
>>
>>  I would like to end up with a cd-rom that would contain (1) all of
>>  the packages necessary to install GRASS on the various operating
>>  systems (a separate cd-rom for each OS, not all on the same cd-rom),
>>  and (2) our proprietary "extension" to GRASS along with the
>>  documentation on how to use it. What we would be selling is the use
>>  of our proprietary stuff along with the cd-rom installation of GRASS
>>  (the same as OpenOSX), including the tech support for the entire
>>  bundle.
>>
>>  The next big issue would be the cost of such a product. My personal
>>  opinion is that it is ludicrous for any software to cost $6500+ (I
>>  believe we know of which software I speak) for a single license. I
>>  would  like to see our (hoped for) GRASS bundle go for something like
>>  $75-100, the value added to the open source GRASS packages being our
>>  proprietary application, the tech support, and the bundled cd-rom
>>  itself. I would imagine that any extensions for the other GIS
>>  software (ESRI, ENVI, IDRISI) would have a cost more in line with
>>  their price structures.
>>
>>  I think this type of approach also benefits the general GRASS
>>  development effort as well. One way that GRASS development will
>>  continue to advance is through solid commercial applications using
>>  the software. Put together a bundle like what I outline above, a
>>  bundle that allows you to pop in a cd-rom, install all the needed
>>  software and get working with minimal fuss, and you will see a lot
>>  more people using GRASS. Unfortunately, as already pointed out, this
>>  requires money to implement. Hopefully, NASA will see the value in my
>>  research proposal and give us some funding to at least take some
>>  steps in this direction. ;-)
>>
>>  Does anyone see a problem with this approach in light of this recent
>>  discussion on licensing?
>>
>>  Bill


-- 

Bill Dickinson
GIS Specialist
NASA Goddard Space Flight Center
Environmental & Safety Branch, Code 250
wdickins at pop700.gsfc.nasa.gov