[GRASSLIST:1665] Learning GRASS

[Wes Jennings]

I recently signed on to the GRASSLIST. I was lurking in the background
getting impressed with the knowledge level of the emails submitted. But
I had some questions to ask, but I didn't want to bore the whole list
with elemental questions. I decided to pick on some poor soul and write
them directly. I picked out Rich Shepard to direct my private
questions. His response so impressed me that I ask him if we could
submit it to the GRASSLIST. He graciously agreed. My system does not
have my original email. However, Rich has quoted my basic questions for
his responses. I hope others get as much encouragement from this as I
did. Again, Rich, thanks for your well measured responses.

Wes Jennings  Email: wes at taclug.org

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

On Tue, 20 Mar 2001 wes at adrift.harbornet.com wrote:
(Note: My email address is now  wes at taclug.org)

> I am concerned about two particular points. (1) Is it really feasible to
> learn something as complicated as a GIS from just the spearfish database
> and the tutorial?

Wes,

  Your question has two parts, and the answers are a) yes and b) no. The
tutorial will start to give you a feel for the mechanics of running the
software. Starting to understand the different modules and how to use the
atrocious interface will give you the basis for experimenting and learning
on your own data. So, from this perspective, the answer is "yes, you can
learn the mechanics of the software from the tutorial".

  However, there's MUCH more to using a GIS than the mechanics of
manipulating the program. This has been one of my big issues for more than a
decade (I started learning/using GIS back in 1987). Look at it this way:
teaching someone how to use a word processor does not make him a writer.
Similarly, teaching someone how to use GRASS, MapInfo, ARC/Info, ARCview, ER
Mapper, etc. does not make her a spatial analyst. So, the answer to that
part of your question is "no". Books and coursework are the answer here.

> (2) Just how bright is the future for the GRASS GIS system? Will it be
> around in 5 years healthy and kicking?

 A year or so ago, the estimate I received from Baylor was 40,000-45,000
users world wide. GRASS has been around since the late 1980s, and I expect
it will be around longer than I. Why not? You get the source code and can
do with it as you wish.

  The current version in development (5) retains a lot of the baggage from
the program's earliest days. That'll all go away in version 6 (next year
sometime). GRASS handles every type of spatial data: vector and raster. It
is an image processing package with many of the features of ER-Mapper, ERDAS
and others; more capability can be added when there's a need. It handles
vector data as does ARC/info, ARCview, MapInfo and Maptitude, but right now
the vector analytical capabilities are limited in comparison to the raster
analytic capabilities. Raster's where it happens for most of the
environmental/natural resource analyses for which GRASS was invented. (GRASS
started life as the Ft. Hood Information System and grew into the
environmental management tool the Army used (uses?) for all military
installations.)

  Then there are the hydrologic and other modules that are (or could be)
tightly integrated with GRASS. I suppose you could get comparable analytic
capabilities from ARC/Info and some of the expensive add-ons to it, but
you'd spend >$30,000 for a single-user license on NT.

  Bugs in GRASS get fixed quickly. There's tremendous support via the 'Net
for it. So, I do believe that GRASS will be around 5 years from now. As a
matter of fact, down the hall from a friend of mine in Boulder, CO, is a
company that adopted GRASS about 5 years ago (a much earlier 4.x version).
They've modified it themselves and have become experts with it. They sell
spatial analyses and visualization based on GRASS to developers and they are
making a ton of money. Because they can do what others cannot do. This
company doesn't care where GRASS is going because they have what they need
and they have the ability to make it into whatever they need in the future,
too.

> Any info about your own personal use and Applied Ecosystems use of GRASS
> would be appreciated.

  I've not used GRASS too much for projects because almost all my project
data are in MapInfo format and the MI->GRASS translator is still not a
finished product (we can do points and lines just fine, but the polygons are
very complex to translate -- especially having to develop topological
relationships for GRASS when MapInfo doesn't use topology).

  However, I am using GRASS to develop and environmental decision support
system for a client and will be using GRASS to prepare Environmental Impact
Statements (and EAs) for 2-3 clients in the very near future. I am also
putting together a hydrologic/hydraulic model of the Willamette River basin
to look at the influence of land uses on erosion, sediment dynamics, fish
habitats and other issues of concern. All new projects will use GRASS and
the older ones will be transferred as needed.

> I have written to you personally, instead of submitting to the list, since
> I'm not aware of how well accepted such elemental questions would be
> received on the list.

  Everyone on the list is friendly and helpful. I would recommend that you
go to the local technical bookstore and carefully look at their selection of
books on GIS. Joe Berry's books (compilations of his columns) are a good
place to start. What you want are books that teach you how to apply GIS to
solve problems, not how to do specific tasks using one vendor's brand of
GIS.

  There are also introductory books that describe the components of a GIS.
Those may also be helpful so you can see how GRASS is put together and how
the pieces relate.

Hope this helps,

Rich

Dr. Richard B. Shepard, President

                       Applied Ecosystem Services, Inc. (TM)
            2404 SW 22nd Street | Troutdale, OR 97060-1247 | U.S.A.
 + 1 503-667-4517 (voice) | + 1 503-667-8863 (fax) | rshepard at appl-ecosys.com