Trends in Web Mapping: Dispelling the Myths around Web Services

What are Web Services?
From the OGC Website: "Web services are self-contained, self-describing, modular applications that can be published, located, and invoked across the Web.Web services perform functions that can be anything from simple requests to complicated business processes.Once a Web service is deployed, other applications (and other Web services) can discover and invoke the deployed service."

Web mapping is a rapidly evolving field, and new applications and technological possibilities are emerging on an almost daily basis.At the same time, the market and user base are exploding, Google Earth being a case in point.However, Web mapping is more than just a pretty face.As usual, it would be nothing without underlying data.Here too, a (less visible) technological revolution is happening.

The maturing of Web services and interoperability standards allow users to have access to a framework that enables them to access spatial data and applications from anywhere across the network, as was recently demonstrated in the Australian Spatial Interoperability Demonstration Project (SIDP) project.This means that data can be obtained from point-of-truth custodians, avoiding duplication and enabling business continuity solutions, saving significant time and money and delivering better information products.

Despite these obvious benefits, this approach highlights a number of issues around security, connectivity, privacy, licensing, etc.These need to be, and can be, dealt with effectively.However, sometimes such issues begin to lead lives of their own.They turn into myths that risk resonating with the decision-makers, hampering much needed innovation.

This article presents and dispels three such myths, which I call the 1) bandwidth; 2) reliability; and 3) security myths.We will see that the concerns raised may be real, but that there are effective and efficient ways of dealing with them.In the end, the only thing we have to fear is fear itself.

1.The Bandwidth Myth: "These solutions rely on high-bandwidth connections, and will never work in the bush or over wireless communications."
Though often heard, this myth is only true for badly designed solutions.Yes, Web services do sometimes exchange large volumes of spatial data.And if your solution relies on users consuming or entering data over low bandwidth connections, it is important that your architecture is designed in such a way that the high-volume payloads are exchanged between servers and services, and communication with end-user clients is lightweight.This can be achieved by simple mechanisms such as transmitting only map images to the end-user (JPGs tend to be under 100Kb in size).Many such solutions have been implemented already and have been shown to function efficiently over dial-up and wireless connections.

Furthermore, bandwidth these days is the most rapidly growing IT infrastructure component, as shown in the graph that accompanies this article in Scientific American.With bandwidth capacity doubling every nine months, it is highly likely this concern will disappear completely in the next couple of years.For instance, this article is being written in my local cafe, where they've just installed 1MB/sec free wireless Internet access available for all customers.

2.The Reliability Myth: "Our solutions are mission critical; we cannot rely on the Internet, what if the network goes down?"
Actually, this myth is a combination of three concerns, all relating to business continuity issues.

• What if the entire network goes down?
• What if the data service I rely on becomes unavailable?
• What if my connection to the network is cut off?

The first one is the "doomsday" scenario.Some catastrophic event is going to disable the entire network, be it your organization's Intranet or the world-wide Internet.Apart from the frequency considerations of this occurring, it is also important to realize that the Internet is specifically designed for survival, maintaining connectivity during full-scale nuclear war.Redundancy, peer-to-peer connections and automatic rerouting are all inherent design elements that make complete failure highly unlikely, if not impossible.

Dependency on a single service for a particular data source is always a bad idea, especially for mission critical applications.Even in centralized storage situations, redundancy and off-site mirroring should be built in.You wouldn't want that one meteor to take out your data bunker, would you? This is no different in distributed, Web service based solutions.On the contrary: Web services, especially interoperable (i.e.standardized, vendor independent) ones, make this much easier to achieve.If, for instance, the Australian Bureau of Meteorology's weather service goes down today, you can already connect to several other (commercial) weather services within seconds to get the same information.Likewise, there are already many cadastral and topographic standardized Web services available around Australia that provide you with the required redundancy and fall-back.

Failure of your own connection to the network is the most realistic concern in this myth.It happens most frequently (I'm sure you can remember the last time it happened to you), and has the most potential impact.

Nevertheless, it need not be as serious as it's perceived.There are several ways to deal with it, depending on the mission criticality of your solution.First, there's the physical network connection.When your broadband connection goes down, there's always a phone line to fall back on.And you could also consider building further redundancy into your connectivity by signing up to multiple physical network providers or Internet service providers.Second, caching strategies will ensure ongoing availability during service outages.Finally, it is important to note that the reliability myth is conspicuously absent in other communities that rely on business continuity for their mission critical systems, such as banking and insurance.There it is widely recognized that networked, interoperable systems are a prerequisite for, rather than a threat to, business continuity.The spatial community certainly has something to learn here.

3.The Security Myth: "The only secure connection is no connection."
Sure, and if you lock yourself up in your home, you'll never get run over by a bus.Of course, there are very genuine security concerns with (spatial) data.Commercial, licensing, potential misuse and privacy issues all need to be taken into consideration.Each dataset has its own level of security risk attached to it, and in some cases (but I'd wager only a very small proportion), the risk level is such that the data can only be exchanged across a quarantined network.For the vast majority however, there are a great number of security options available to manage these risks.Here are a few:

• Restricting access to trusted users and/or IP addresses;
• Intranet-only access;
• Using Virtual Private Networks (VPNs);
• Restricting scale levels at which your data can be accessed;
• Restricting the number of features that can be retrieved at one time.

Sure, not a single one of these is as secure as "no connection." And each organization should make its own risk assessment. But you do need to ask yourself if you can afford not to be connected. Let's look at the banking world again.The security risks inherent to Internet banking are very real and cost banks and their customers millions of dollars.Despite this, Internet banking is cheap and convenient and continues to increase in popularity.The banks are certainly not considering shutting it down, or making it prohibitively secure.Nor are the customers demanding this.The benefits just outweigh the risks.

Speed, reliability and security are relevant aspects of Web based solutions and need to be addressed in earnest, depending on risk-benefit tradeoffs for each individual situation.This article shows that there are many mechanisms available to do this the myths surrounding Web mapping solutions can be easily dispelled.

Ed.Note: Maurits Van Der Vlugt was responsible for delivering the Australian Spatial Information Business Association (ASIBA) / Open Geospatial Consortium-Australasia (OGC) SIDP Project.This article is partly inspired by audience feedback received about SIDP.In addition, this article first appeared in Position Magazine.

Share