Home » Publications » Commercial Open Source: Increase Web Mapping Capabilities While Controlling Costs

Commercial Open Source: Increase Web Mapping Capabilities While Controlling Costs
The Value of the OpenGeo Suite

1. Overview

Today, many enterprises and government agencies are finding themselves in a hole — they need to deploy more and better geospatial web services, but are faced with shrinking budgets. What’s more, they are often hampered with closed source software, raising the spectre of rising license costs as data and service volumes grow. The first law of holes states that if you find yourself in a hole, stop digging!

The upcoming release of OpenGeo Suite 3.0 provides enterprises with an immediate opportunity to stop digging and climb out this hole. This release adds significant spatial processing functionality to OpenGeo’s considerable advantages in reliability, scalability, costs and control. Until recently, geospatial web services have been dominated by a closed source provider, but increasingly open source alternatives have been closing the gap. The OpenGeo Suite 3.0, a commercial open source stack, can provide enterprises with clearly superior value in delivering geospatial web services, while simultaneously increasing the reach and functionality of enterprise systems and controlling costs.

This paper compares the relative strengths and weaknesses of closed source geospatial web services software, open source (unsupported) alternatives, and commercial open source — namely the OpenGeo Suite. It aims to help managers see that enterprises can develop and deploy more web-based geospatial applications at lower costs using commercial open source software. Our conclusion is that a strategy to both redirect software maintenance spending to commercial open source and invest the savings into increasing the functionality of open source software will result in superior results.

2. Relative Value in Software Systems

To illustrate the relative values of closed source, open source and commercial open source (specifically, the OpenGeo Suite), this paper posits that the value of any software system can be estimated based on:

1. Its functionality and power
2. How much it costs and
3. How much control it offers software users

We propose that the value of software systems to enterprises in a market can be expressed with a simple equation, where V = Value, F = Functionality and Power, O = Operational Costs, C = Control. That relationship can be expressed as:

V = (F/O) * C

It is beyond the scope of this paper to definitively break down these value elements into all of their subcomponents, or to measure precise values for them. However, it is quite possible to see how relative changes in each drives value and therefore to compare alternative software systems on value or V. In this way enterprises can make judgements about which are better for their operations. Let’s examine the elements of the equation.

Value to enterprises is represented by V, such as how much a desired object or condition is worth relative to other objects or conditions.The Value of software systems for enterprises increases as software functionality and utility increases relative to the overall costs of the software, and as enterprises achieve more control.

Functionality and Power, or F is the value driver representing a comprehensive meaning of overall software utility. It includes elements of reliability like system up-time and “buggy-ness,” as well as elements of “scale” such as numbers of users or amounts of data able to be served at some speed of operation and per some unit of hardware or processing power. It also contains aspects of “power” — the number or percentage of maximum functional operations enabled, as well as the number or percentage of eligible data or collaborators able to be incorporated into applications. A more detailed articulation and weighting of these sub-elements of Functional Power will be developed in a future paper.

Organizational Costs, O, include all elements driving costs, including software costs (licenses and maintenance contracts), hardware and infrastructure costs, and the personnel costs to evaluate, test, deploy, operate and maintain systems.

For the purposes of this paper, we will focus on a key driver of Control: choice. Without exactly determining the value of Control, it is easy to assume that when there are two alternatives for software providers, user Control is better than when there is only one. This is likely to be even significant when one supplier is open source, which mitigates oligopolistic effects. For simplicity sake let’s say that the value of C only varies by how many choices are available; let’s say that when there are two choices for geospatial web services software, Control (C) = 1. In this case, users have a choice between two alternatives, and therefore they have a better chance of obtaining the full value of F/O from one or both of them.

However, when there is only one alternative (that is to say, no alternative at all), then C will logically be less than when there are two choices (again, assuming C = 1 when there are two choices). So, with only one alternative the value of C is a fraction of one and actually lowers Value. When the number of alternatives rises, C improves — probably sharply at first and then much more gradually — up to a point where the marginal value of more Control or Choice is no longer helpful. See Figure 1 for a graph of this type of growth.

Figure 1

A Binary Log Function. This is a good model of how a rising number of software alternatives might influence Control. In the case of software value, C would be “stepped’ down to < 1 when the number of choices equals one (no choice).
Source

In a future paper we will explore this relationship more closely, looking at functions that help describe it (such as step functions incorporating the binary logarithm), and at notions of oligopoly and pricing.

3. Harms of Closed Source Software and Impacts on Value

In a recent blog post, open source advocate Eric Raymond describes the “harms of closed source software.” His helpful list of harms provide a guide to the value elements for software, and to contrasting open and closed source software alternatives. These closed source harms include: Reliability, Unhackability, Agency, Lock-In, Amnesia and the lack of Network Externalities.

Reliability Harm refers to the fact that the secret nature of closed source software can lead to poor engineering and less reliability than open source alternatives, which frequently benefit from the combined efforts of large communities of skilled developers.

“Unhackability” Harm refers to enterprises losing the option to modify software to suit their own schedule or needs, or hire anyone of their choice to do the work. This is a harm that grows with complexity — it varies in importance according to the expected value of customization — and with time sensitivity (how valuable it would be to be able to prioritize changes in software). With open source, enterprises are free to improve reliability and repair deficiencies on their own terms and at their own pace, not those of a closed source software vendor.

Agency Harm is another major deficiency of closed source software, where the closed, owned-and-controlled software situation creates an asymmetrical power relationship between users and the entities or people who control the software code. Owners of closed source IP rights can, and do use this asymmetry to limit choice and increase prices. Importantly, this harm vastly increases with scope of the relationship between the parties — the greater the depth or penetration of usage the software has across an enterprise, the greater the negative consequences of Agency Harm. No such asymmetry exists with open source software — all parties come to the table with equal access to the software code, assuring software users that pricing is either fair or it’s free!

Lock-In Harm is a familiar concept, where the closed source of the software constrains enterprises to use the distribution and service outlets set by the entity controlling the code. This increases transition costs to abandon the software by making escape from other harms more difficult. In contrast, open source software offers enterprises complete choice to hire their own experts to manage and service the code, or to employ any third party they choose. The most extreme form of lock-in harm is Amnesia Harm, which happens when closed source software (or its developers) go away (from acquisitions, companies going out of business, etc.). When this occurs entire capabilities can be forgotten or disappear with their authors. With open source the risk of source code knowledge being lost with the demise of a software authoring entity is mitigated.

Finally, Raymond notes the problem with Network Externalities. Software products have "positive network externalities” when, as more people use them, the value to any individual rises. Unfortunately, with closed source these “positive” network externalities act as lock-in harm; they make it harder and more expensive to transition out to better and cheaper alternatives. On the other hand, with open source, the more users and contributors, the more everyone benefits from bug spotting, software contributions, and more.

4. The Value of Supported Open Source Software Compared to Alternatives

The value factors F, O and C (along with the above mentioned “harms”) provide a framework for comparing various web mapping software sources. By looking at how factors work for closed source, open source and commercial open source, one can begin to get a sense of relative weight for each.

• Functional Power: Reliability and “hackability” provide a boost to the utility of open source software compared to closed source. The deployment of open source for the FCC National Broadband Map application clearly showed the power of commercial open source to both scale and stand up to usage volumes that traditionally choke closed source systems. That said, closed source providers have worked hard to offer the broadest possible functionality in terms of numbers and depth of functions, since the agency harm, lock-in and positive network externality effects mean that broad usage gives them asymmetric power to increase prices or increase profits through lower quality at the same prices. In the geospatial field we have seen closed source software pushed into wide use with a tremendous range of functionality — but with a mixed result due to low reliability and unhackability.
• Operational Costs: Overall, the organizational costs factor tends to favor open source over closed source, and commercial open source over unsupported. Because open source and commercial open source systems have no licensing costs, they start out with a cost advantage over closed source systems (note the experience of NYC DOITT, New York City’s IT infrastructure department). This advantage is likely to widen with the expansion of infrastructure needed to handle big data — resulting in increased licensing for closed source systems. Both commercial open source and closed source have explicit maintenance contract costs, but while unsupported open source does not, this simply shifts these costs over to end users. In recent years, reliability has been a strength of open source geospatial software, with closed source lagging on this factor. Also, because open source software is “hackable,” enterprises can put resources to work to modify or fix it on their own schedules, potentially lowering costs. And, because there is no lock-in and no agency-based intellectual property with open source, enterprises are free to seek the most cost-effective solution — either internally or through their choice of third party services. Additionally, as more enterprises use an open source stack and its functionality and reliability increases (through added bug spotting and funded software development), enterprises can actually find their operational costs going down due to a virtuous cycle of greater and greater contributions back to the core software.
• Control: Open source — particularly when there is a commercial open source alternative — offers a clear advantage in the area of control. Where there is closed source there is no user control, and control of Functional Power and Organization Costs shifts from enterprises to the closed source provider. No such asymmetry exists with open source. The addition of a commercial open source provider increases Control by increasing choice, as enterprises can choose between developing their own solutions on top of closed source or unsupported open source (Do It Yourself). An apt analogy used to compare commercial or “professional” open source to unsupported open source is that of the beekeeper, who mediates the “inconvenience” of dealing with bees in order for consumers to enjoy honey.
  

  Convenience and ease of use (lower Operational Costs) can result with the emergence of a commercial open source alternative, like the OpenGeo Suite, which generally has what users need out-of-the-box (as seen in the case of SP-Ausnet, an Australian utility).

  With open source, the software source code is open and there is no risk of unhackability, lock-in or amnesia harm. Enterprises are free to manage their own timetables for software life span, migration and/or obsolescence. The enterprise has options for modifications, setting priorities, service providers and more; open source provides substantially greater control than closed source alternatives, and commercial open source adds onto that.

Like Beekeepers with bees and honey, commercial open source firms can ease access to open source software for enterprises not accustomed to working with open source communities. Image Source

5. The Value of the OpenGeo Suite Versus Alternatives

The formula [V = (F/O) * C] provides a mechanism to assess the value of commercial open source (the OpenGeo Suite) versus closed and unsupported open source alternatives. As the equation must stay in balance, increasing F (Functional Power) or C (Control and Choice) or decreasing O (Operational Costs) will result in greater V (Value). Table 1 outlines the relative position of closed source, (unsupported) open source and commercial open source (the OpenGeo Suite) on the value elements outlined above.

Table 1

6. Comparing Value for Software Alternatives

Based on the relative positions outlined in Table 1 for factors, F, O and C, the relative values of Closed Source, unsupported open source and commercial open source (the OpenGeo Suite) can be determined (see Table 2).

Table 2

For Closed Source the ratio of F/O works out to be fairly poor, as medium high Functional Power is diminished by high Organizational Costs.Then that is compounded by the poor score on Control (no choice, a lot of unhackability/vendor lock-in, etc.) to end up at a Value position estimated at Medium Low. For unsupported open source all factors are Medium, for an overall Value of Medium. And for the the commercial open source alternative, the OpenGeo Suite the added F (Functional Power) provided by OpenGeo combines with reduced O (Organizational Costs) from expert, enterprise-class support to create an outstanding F/O ratio.

Changing Relative Value

Over time, as either open or closed source changes in Functional Power, Operational Costs or Control elements, the measures of relative value will shift. As we discuss software and its utility, it seems the primary fields of contention should be Functional Power and Organizational Costs. Can closed source alternatives make things work so well and be so convenient that enterprises have no reason to switch over to open source? In other words, can the F/O ratio be driven so high by closed source providers that their relative value stays attractive compared to open source, given the low level of Control closed source offers?

Past performance indicates that both open and closed source geospatial software will continue to increase in Functional Power, but the lead held by closed source is likely to shrink. Some think that an insurmountable organizational cost advantage can be established by closed source through cloud and software-as-a-service (SaaS) offerings, but this does not seem likely. And, even with SaaS, enterprises and their developers need, and want to avoid “unhackability” harm to the greatest extent possible. So, we are likely to continue to see a steady relative increase of the F/O ratio in favor of open source.

The factor of Control remains almost beyond the power of closed source to change, the very nature of proprietary closed source offerings result in a very low C. Open Source is the polar opposite: the opportunity for enterprises to bring open source maintenance and advancement in house will not change. The key here will be whether open source alternatives remain robust. There’s a good deal of evidence that open source geospatial will remain a strong alternative, including:

• The large size and vitality of open source geospatial communities
• The attention open source is receiving by those faced with shrinking budgets,
• The mission of OpenGeo to expand the reach of open source, supported and unsupported open source

7. A Call to Action

What are enterprises to do with this perspective? The call to action is two-fold:

1. Enterprises should transition their web mapping and geospatial services to open source, particularly commercial open source alternatives (the OpenGeo Suite) to the greatest extent possible. By doing so, organizations can actually increase the functionality, reliability and scalability of their web deployments while simultaneously reducing costs.
2. Invest a portion of their budget savings in extending open source technology (in other words, increase F) in order to further increase the value available to their enterprises and to contribute to the positive network externalities that exists in open source software.

When enterprises choose the OpenGeo Suite, they get greater value from the open source geospatial software OpenGeo supports than from closed source or unsupported open source options. For every dollar or euro or pound or yen they spend on software, they are able to do more, faster, on cheaper hardware, and with greater control with commercial open source.

8. About Us

OpenGeo is working to build the best web-based geospatial technology. The company brings the best practices of open source software to geospatial organizations around the world by providing enterprises with supported, tested, and integrated open source solutions to build the Geospatial Web. OpenGeo also supports open source communities by employing key developers of PostGIS, GeoServer, and OpenLayers. Since 2002, the company has provided successful consulting services and products to clients like the World Bank, US Department of State, NYC DoITT, Ordnance Survey Great Britain, SFMTA, Portland TriMet, MassGIS, GeoScience Australia, NOAA and the Federal Communications Commission. OpenGeo is the geospatial division of OpenPlans, a New York-based 501(c)(3) non-profit that informs and engages communities through journalism and open source software. All of OpenGeo's revenue has been and will continue to be re-invested into innovative and useful software in support of the OpenPlans mission.

© 2012 OpenGeo.
Redistributable under the Creative Commons Attribution-Share Alike license.