Why ARC/INFO over GRASS?
We are long time GRASS users who are switching to ARC (using the relatively
cheap GIS II contract prices available to DOI agencies) because:
1) Standard database interface: we are really tired of one
attribute per map layer. (INFO is built in, not the best database,
but a database, none the less.) No additional multi-grand$$
database (ie, ORACLE) required (with its own learning curve). We
tried RIM -- 
-- no thanks.
2) Seamless integration with neighboring agencies. (Boulder County,
Larimer County, Region 2 of the USFS, including Arapaho/Roosevelt
& Routt NF's, all currently use ARC/INFO. The town of Estes Park
and Grand County will soon be on-line with ARC/INFO GIS's). The
public is getting pretty fed up with public agencies having
different, non-communicating inventories, and not being able to
come up with ecosystem-type approaches to solutions. This goes
beyond just software, but incompatable softwares and data
conversions have added to the problem.
3) Support. With the loss of the central NPS GIS division (GISD)
to the NBS, support for GRASS (with it's "broken code or program?,
user - heal thyself" philosophy) is harder to come by; many parks
are not willing to allow their GIS people the time to compile and
fix "open-GIS" software. The GRASS-user's list is GREAT, but not
all can access it, not all can understand or have the time to
implement the solutions offered by it. The cost of ARC (via
GIS II) versus the amount of time lost performing GRASS system
administration about even each other out.
4) Equipment. No extra cost involved. ARC/INFO will run on the
same workstation (SUN SPARCclassic/Solaris 2.x/Altek
digitizer/WYSE dumb terminal/etc.) that we use for GRASS.
5) Intergrated graphic output. GRASS output capabilities leave
MUCH to be desired. Right now, we have to keep a copy of GRASS4.0
with MAPGEN around, because we can't get the version of MAPGEN that
runs with GRASS4.1 compiled and running. Even if we could, we
still use MAPGEN just to dump vectors into CorelDraw, and use
CorelDraw to create/annotate maps (printed on a HP 650C plotter).
No driver for a plotter of this quality available for GRASS and its
"pmap" programs, and after using this method, "pmap" output looks
fairly sophmoric. One just needs to check out the "ARC/INFO MAPS
19xx" books that ESRI puts out each year to see the quality maps
that ARC is capable of.
I personally am glad we are converting because of the opportunities it will
open for me career-wise down the line (2, 3 years); I don't see many
job announcements that call for "X years of GRASS or open-GIS
experience required". True, GRASS has given me more systems admin.
experience than I might have had the chance to learn with a "canned" GIS,
but I've also been lucky enough to have a boss who has given me the time to
figure things out.
I also do not subscribe to the view that all we get rid of in government is
"the good stuff", and all we adopt is wasteful and inefficient. Good
people do good work. Of course, it's always easy to blame the supervisors
or some other "them" out there...
We will keep a current copy of GRASS up and running while I'm on the
ARC/INFO learning curve, and I still see GRASS as a viable GIS for those
with low/no budgets for commercial GIS's. The GRASS isn't dead, but it's not
growing as fast as before, either. It simply doesn't meet the needs of this
national park anymore.
Ronald Thomas ront@meeker.cfnr.colostate.edu
Natural Resource Spec. (GIS) ^^^ Phone: 303-586-1285
Resources Management Division ^^ ^^^^^ FAX: 303-586-1310
Rocky Mountain National Park ^^^ ^^^^^^^ Estes Park, CO 80517