GIS Monitor Apr 5, 2001


-ESRI Retires ArcCAD and Atlas GIS
-Manifold GIS Chooses ECW Over MrSid
-Making Money in Location-based Services
-Plagiarism on the Rise

Departments: Points of Interest, Week in Review, Back Issues, Advertise, Contact, Subscribe/Unsubscribe

This issue sponsored by: ERDAS (


ESRI will end sales and support for two of its products: ArcCAD and Atlas GIS. Atlas GIS sales ended March 31 and ArcCAD sales will end with April 30. Technical support will be available until the end of the year, according to letters sent to users.

As I have a long history with these products, I can empathize with the users of these products. I was a member of the ArcCAD team for many years. I received my first live demo of Atlas GIS in 1990, then owned by Strategic Mapping.

However, it is hard to argue that these products should be kept alive. ArcCAD was originally built to work with AutoCAD 11, back in 1992. The unprecedented collaboration of ESRI and Autodesk back then surprised many. I spent many days as an ESRI employee in the Autodesk booth showing ArcCAD and gathering feedback on where the product should go in the future. But ESRI had a tough time keeping up with Autodesk’s fast pace -- as did many other developers. ESRI survived AutoCAD’s move to Windows, the unlucky R13 and even AutoCAD Map, Autodesk’s competing entry into the CAD/GIS space. However, after the launch of AutoCAD Map, Autodesk decided that ESRI was no longer welcome in the Autodesk developer program. Even without developer support or early access to new releases, I was pleased that ESRI continued to support the existing user base for years.

But, let’s face it, CAD-based GIS --anyone’s CAD based GIS -- just did/does not deliver on its promise. All of the attempts, from Dennis Klein’s FMS/AC (gone), to Bill McKenzie’s Geo/SQL (still around, but tiny), to Intergraph’s MGE (no longer in development), to ESRI’s ArcCAD, to Autodesk’s AutoCAD Map/Land Development Desktop to Bentley’s MicroStation Geographics (originally Mizar’s MicroGIS) were quite similar. The integration between CAD and GIS was never complete –- you could never just pick up the CAD tools and work on the data. Extra procedures created entirely new workflows and made learning difficult. Try as they might, splicing GIS genes into CAD package did not make graceful creatures.

To make matters worse, these products failed to deliver in the marketplace.

Consider the best selling of these products, AutoCAD Map. According to figures from their Autodesk GIS 2000 Conference, held last fall, users of Autodesk mapping products are distributed as follow:

- over 600,000 AutoCAD-only GIS users
- over 150,000 AutoCAD Map and AutoCAD Land Development Desktop (LDD)users
- more than 17,000 Enterprise GIS users (Vision)
- 2000 Autodesk MapGuide installations with users measured in the millions

With the price difference of only a few hundred dollars between Map and AutoCAD, Autodesk probably expected most of its GIS Users to switch to it. Map probably did win some AutoCAD users over because it supports MicroStation DGN files – something AutoCAD still does not do. More may have had Map fall into their laps after buying Land Development Desktop for engineering purposes. To me, this proves there just isn’t that much demand in this niche. CAD based GIS did not, and does not, rule the world.

Look for other solutions to solve the CAD/GIS integration challenge including Windows-based desktop GIS (GeoMedia, for example) and client server solutions, like Autodesk’s own GIS Design Server.

Among the Windows-based desktop products was Atlas GIS. One of the early products introduced by Strategic Mapping (SMI) in the late 1980s, this was a fine piece of software. Simple, intuitive, inexpensive, Atlas was very much the “Excel” of mapping. When SMI realized that selling inexpensive software and expensive data was not working out well, they sold off their data to Claritas and the software, Atlas GIS, to ESRI. ESRI more than likely wanted the Atlas market share, which was mostly in business geographics, a “hot” market in 1996. Indeed, ESRI did get a few hot Atlas resellers with big accounts in the deal.

But the business geographics market on the desktop did not live up to expectations and MapInfo ruled whatever there was of it into the mid ‘90s. ESRI ended up with TWO Windows desktop GISs and had to simultaneously market and support Atlas and ArcView, the rising star of the company. How long could Atlas last? I feel strongly that it actually survived longer at ESRI than it might have elsewhere.

Perhaps the most important thing to remember is that this is a business decision, not an evaluation of the importance or quality of the products. The fact that so many users did so much with the software products is testament to their worthiness of the products and their role in the industry. The work put into Atlas and ArcCAD continues to drive GIS development and the data and maps developed in them will no doubt have a life elsewhere.

====== a message from our sponsor ===============
Do You Use ERDAS Software?

Then join us for our annual ERDAS User Day, Monday, April 23, in St. Louis, Missouri! Not an ERDAS client? Stop by our booth (#101) at ASPRS 2001 to get your solution.

For more information, see


Earth Resource Mapping (ERM) recently issued a press release stating that Manifold chose their ECW image compression format over LizardTech’s MrSid. As there was no explanation of why, I got a hold of Dmitri Rotow, Manifold product manager, who gave me the details. He explained that work occurred during the time that LizardTech was involved with a lawsuit against ERM for patent infringement, illegal use of MrSID trademarks, illegal copying of a LizardTech SDK, making false statements when comparing ECW to MrSid.

“We actually did do a comparison between ERMapper and MrSid to consider which SDK (if any) we would incorporate into 5.00 before we were aware of the lawsuit. That analysis considered a variety of factors, including the technology, the licensing terms and conditions of the SDK, the strategic factors, etc. From our perspective as a software developer it was hands down a choice for ERMapper.

“In a nutshell the key issues for us were:

1) Core technology. ECW really is supported by an open community that exchanges many sophisticated ideas and actual code. Anybody (including ERMapper and us) can't help but benefit from that. We could educate ourselves in a deep and serious way on ECW using the resources of the open community. We couldn't do that with MrSid. MrSID, in our view, was likely to fall further behind with their approach.

2) Licensing. The ERMapper SDK makes it easy for a third party to read/write ERMapper ECW formats. Their legal approach is surprisingly deft and low key.

3) Strategic factors. We think compression is a good idea. ECW makes it easy to add compression, as we've done in our own ECW implementation (we're pretty good at math so this is no big deal given the open community discussions about ECW). However, our business focus is not ECW nor are we licensing our stuff to the entire world for free. ERMapper has a focus on this and is doing a good job making it easy for third parties to use their approach. It's a safe bet that ERMapper's SDK will become a widely-adopted industry standard for interchange so we wanted to have it in 5.00 whether or not we did our own. In contrast, it is difficult to see how MrSid will become an interchange standard on the scale ERMapper's SDK makes possible.”


To end-users location-based services sound like a good idea. People need to find out where the closest drycleaner is. Many take this one step further and believe people are willing to pay to get that information. Questions: when we do pay and where does the money go? Or in other words, who needs to be paid for the subscriber to get that information?

It’s actually fairly complicated. First, the carrier has to know where the device is. Wireless carriers are currently scrambling to be sure they meet the FCC deadline of October 1 to meet regulations in the US. Some are “doing it themselves” and others are contracting with companies to provide just the “locate me” piece. CellLoc, based in Canada, is a company addressing just that part of the equation, though a sister company will deliver services. CellLoc is building a network designed not to carry phone traffic, but just to pinpoint locations of devices.

Once the device’s location is known, step two is mapping it and the querying for the information of interest. There’s a bit of geocoding (locating the latitude/longitude or the cell the device inhabits) and then the GIS part: doing the spatial query and providing candidate “hits.”

Step three is delivering that response – in a map, or text or perhaps voice. Now, it is entirely possible for three or more companies and technologies to come together to answer the drycleaner query. So, what happens to my $10 a month (I made that number up)? How much goes to each piece? I suppose we’ll learn more about that as LBS comes online, but for now, the GIS part is but a sliver of the whole.


A few incidents of content “borrowing” have left me less than comfortable. Stung was our own site, TenLinks, one of our partner sites, and another GIS specialty site. I cannot believe that the individuals were not aware that they were in fact taking intellectual property of others and passing it on as their own. Perhaps they thought they would not be found out.

A year ago, I uncovered a case of plagiarism. I had agreed to help a GIS Masters degree candidate with her thesis. She interviewed me and I sent her to several sites with white papers written by technologists at many of the top GIS companies. When I saw her draft, I immediately recognized huge passages taken verbatim without attribution. I told her that I’d have nothing further to do with her education. Where I went to school, such behavior resulting in expulsion, no questions asked.

But the writers involved in the recent website incidents were not students but professionals. The Internet makes stealing even easier than in the past, but it makes it no less unethical.

We at TenLinks do receive requests to republish our content and in most cases, we comply, asking only that credit be given.


-A reader requested some details on the CeBIT Bluetooth failure. It turns out the application was designed to give visitors to CeBIT indoor map-based directions between booths. And, this detailed article suggests, the failures were due not only to incompatibilities, but also overworked servers and interference. On another note: Microsoft has decided against in the box support for Bluetooth in its Windows XP due later this year.

- Palm and Sanyo Fashion House have introduced a new raincoat with a special pocket for a Palm Pilot. Prices range from $185 to $695 at specialty stores nationwide.


You can reach more than 6,000 GIS professionals every issue by sponsoring GIS Monitor. For more information, email

Please send comments and suggestions to:
Adena Schutzberg
GIS Monitor Editor
Ultimate Map/GIS Directory - Your search is over!

If you wish to subscribe, unsubscribe or change your preferences visit our
subscription page.