[Geowanking] ArcView 3.1 replacement
ebwolf at gmail.com
Fri Apr 18 16:39:05 PDT 2008
I've always thought that GPL'ing ArcView would be both an outstanding and
astounding move by ESRI. ArcView represents the state of GIS about 10 years
ago. That was a good point in time because the fundamentals of GIS were well
established by that point. A GPL'd ArcView could become a powerful
educational tool that, ultimately, would drive greater demand for ArcGIS.
As far as abandoning analysts with Avenue... ArcGIS through .NET and COM
exposes considerably more functionality than you ever got with Avenue. The
reason Avenue was so simple, semantically, was that the exposed
functionality was very limited. Consider that ArcGIS Desktop (ArcView,
ArcEditor, ArcInfo 9.2) are essentially just implementations of the
functionality exposed by the ArcGIS COM objects. The objects are also
available without the ArcGIS Desktop interface. The transition from ArcView
3.3 to ArcGIS was really hampered by it's complexity. This caused alot of
delays that resulted in people getting really closely tied to Avenue and not
feeling like they can adopt the ArcGIS COM structure.
Overall, I'd say ESRI mostly has an identity problem. They still haven't
fully realized that they are a software company. And they are a software
company on the scale of Microsoft. They need people who are very good at
structuring APIs. And they also need people who better understand software
As for an opensource interpreter for Avenue... Most of the osgeo
environments already have a solid scripting component - usually in Java.
Instead of trying to reinvent the wheel, I think it's better for open source
to move forward with things like JUMP. Of course, an Avenue interpreter in
JUMP might be an interesting project!
On Fri, Apr 18, 2008 at 5:12 PM, Howard Butler <hobu.inc at gmail.com> wrote:
> On Apr 18, 2008, at 5:15 PM, percy wrote:
> I think the key here is to discuss the Extensions to Arcview and Arcmap.
> > Much of the base functionality can be covered by these FOSS desktop clients
> > like gvSIG and QGIS and UDIG, but extensions like Spatial analyst and 3D
> > analyst, for some of us, is where it becomes difficult to our work done
> > using only FOSS.
> I think a key component would be an open source interpreter for Avenue
> implemented for either QGIS, uDig, or gvSIG. It wouldn't need to be
> complete parity with what ArcView 3.x had, but if it were at least an 80/20
> solution, folks with scripts and utilities that they're using for basic data
> manipulation and simple analysis would have an upgrade path.
> ESRI abandoned lots of smalltime developers when they moved from ArcView
> 3.x to ArcGIS and didn't bring the Avenue folks along for the ride. Rather
> than face the migration cost of rewriting all of their Avenue code, they're
> just sitting there hoping things will continue to work unsupported for a
> while. ESRI's not going to bring Avenue forward -- they've moved on to .NET
> and COM Python wrappers -- and I think there's a significant audience of
> analysis-oriented folks that were left behind. If the open source world
> gave these folks a potential migration path, it might generate enough
> interest and funding to bootstrap it.
> Alternatively, if ESRI wanted to totally fluster the open source desktop
> market, they could GPL ArcView (the pieces of it they own and could actually
> license) and let all of us freetards froth over it for two or three years
> trying to beat it into some kind of shape. By the time we actually get it
> to work, they'll probably be end-of-life'ing ArcGIS and on to the next
> Geowanking mailing list
> Geowanking at lists.burri.to
Eric B. Wolf 720-209-6818
PhD Student CU-Boulder - Geography
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Geowanking