I hope by now we have established that getting started with GIS is ever so slightly more complex than simply choosing and pricing software packages. That is not to say budgeting and procurement are not important components of any GIS plan. But buying GIS software without a plan is like putting Descartes before the horse, so to speak.
In the two previous articles of this series, August 2007 and October 2007, we began planning by looking at data and proposed data products. Then we examined our staffing issues. From these two component exercises we developed a needs assessment. Now we are ready to begin the process of matching our hardware and software requirements with our fiscal realities.
Once you truly grasp GIS you will realize that, just as in surveying, the tools of the craft become dated. Remember, it’s all about the data. As an example, a surveyor in the 21st century using 21st-century tools can often accurately reconstruct a survey first executed in the 19th century using the tools of that period because he has the data in the form of a deed or a plat. Technology changes. Data endures.
With that in mind, let us proceed to construct our GIS with the best available technology tools of the day. But we must also proceed with the understanding that these technologies as well as our business needs will change from time to time, and we will need to adapt and plan for those changes as well.
One of the harsh rules of adaptation is fiscal constraint. Before investing in a product, you should consider the return on that investment as well as hidden costs. For example, if you want to find out how much a copy of ArcView “costs” you could simply navigate to the ESRI Web site (www.esri.com), click on the Products tab, and see that the going price for a single use (unkeyed) copy of ArcGIS, ArcView is $1,500. A single network seat is $2,500.
Now we can just multiply that times the number of staff we have identified as users in our personnel interviews and voila! We have completed the budgeting process, right? Not so fast. Remember when we interviewed those potential GIS users? Did we evaluate their existing hardware in that process? Let’s take another look at those standard questions.
- Which of my functions needs GIS?
- What data is needed?
- What hardware do I need?
- What software do I need?
- Will existing procedures be changed by GIS?
That fourth bullet point mentioning software certainly looms large in any GIS plan. But No. 3 about hardware is the one that can put a hole in your plan if you don’t see it coming. Indeed, what hardware do you need? When you buy any professional business software, you are not just buying a piece of software. You are investing in a system. And this software is just one component of that system. Every piece of software sold comes packaged with a set of system requirements.
To get the maximum benefit from your software investment, your hardware must be up to the task of operating it. And not just today, but that hardware must also accommodate the never-ending reality of patches, upgrades and future releases. Computer hardware has a natural tipping point at which it makes business sense to replace a piece of hardware rather than upgrade or retrofit existing equipment. The value estimated for that tipping point is known in the trade as MTBF (Mean Time Before Failure).
Since we are talking about a GIS plan here, we must always look ahead. File exchanging and sharing is an integral part of any Geographic Information System. It is one of the core activities. A wide variety of file formats are typically provided and received on a daily basis.
Yes, file sharing. We didn’t just open a can of worms here; we popped open a whole six-pack. How do I share thee? Let me count the ways. The express purpose of most GIS files is to make them available to a wide audience of users, and the vast majority of these will be end users. (We discussed types of users in Part 2.) To make our data widely available and accessible we need to consider two additional hardware issues: networks and file servers.
NetworksThe old “Sneakernet” will always maintain a presence, but it is obviously no longer an option as a system. If your development environment has a network in place, the task is somewhat easier. We need first to acquire dedicated, assigned storage space. Then we need to get our user accesses and permissions identified, defined and set.
If we don’t have an existing network to work with, we will need to construct one. The operative word here is scalable. The size, type and bandwidth of your network need to be matched to your user needs. Scalable means having the capacity to be expanded by modification rather than replacement.
Here are five examples:
The Fileserver Client. In this model, data is accessed via a desktop GIS client from a file server data source. (“Client” is techno jargon for a standard desktop PC workstation.) In this environment the user typically sees the data source as a network drive letter and the files themselves are viewed in their native format under their original file names. File permissions are established to determine who can see and edit the data, and who can “view only.” This is the most fundamental approach. A rudimentary form of this type of network can be constructed simply using a good workstation for the file server and a generic router kit.
The Database Client. In this environment the desktop client accesses data from a database management system (DBMS). The user normally needs a password to access the database. This configuration typically uses published data (e.g., the content is controlled by the administrators, and users can access but not edit the data). This configuration most often employs a dedicated server.
The Browser Client. The Browser Client works similarly to the Database Client but it employs a browser on a standard Web image map service. And no GIS software is required. The Web site can be open access or enabled with password-protected security. ArcIMS, Map Objects and Autodesk Map Guide are some examples of how this technology is deployed. The Browser Client can be deployed in both intranet and Internet versions.
The Web GIS Client. In the Web GIS Client environment, the Desktop Client can access standard Web image map services and add them to his project. This is newer technology and gaining in popularity. ArcServer by ESRI is an example of using this configuration.
Wireless Networks. More recently wireless networks have become fairly common. Wireless networks utilize modem and cell phone technology. Wireless networks using Bluetooth technology are already making mobile GIS a working reality for some organizations.
Networks come in a wide variety of configurations. There are LANs (Local Area Networks), WANs (Wide Area Networks), HANs (Home Area Networks) and others. Networks also have some descriptive characteristics we need to be familiar with. Topology is the geometric arrangement of a computer system. Network topology, though not identical to the mathematical term topology I have discussed in other previous columns, uses the terms topology and node essentially in the same fashion. The protocol defines a common set of rules and signals that computers on the network use to communicate. Network architecture can be broadly classified as using either a peer–to-peer or client/server configuration. Computers on a network are referred to as clients or nodes. Computers and other devices that allocate resources for a network are called servers.
Integrated SystemsThe integration of GIS functionality with other enterprise systems to better manage resources and assets is another very important element to consider when developing a GIS. Many organizations have struggled with the integration and interoperability of CAD and GIS. Some have been successful.
But there is more, much more. Government mandated programs for inventories of managed assets have expanded the use of GIS in small- and medium-sized municipalities at a furious pace. Legislation has required reporting of and on assets managed by municipalities to the Governmental Accounting Standards Board (GASB). The GASB monitors the financial reporting methods for state and local governments. Most of the GASB compliance is accomplished through asset management software systems such as Cityworks by Azteca Systems. These systems also link up with GIS products such as ArcView to provide the mapping component for these programs.
As we can now see, GIS is not just software. It is a combination of the data, the people and the equipment. So then, how much does a copy of ArcView cost? It really does depend on what you plan to do with it. In my next column in March we’ll review what we have covered so far and try to put it all together.