Thanks for reading and responding.
Standards do not replace competition. They are about cooperation among competitors for the purpose of agreeing on a shared result. If successful, a standard creates a stable point around which to compete. If that stabilizing point occurs, competition simply moves elsewhere. If it doesn't, I expect the likely outcome is competition remains. Perhaps all this means is that competitors are not ready to cooperate.
Successful families of standards will, in time, define the markets they serve. At this point, the risk is high that the competition will migrate into the standard development process.
This is a fascinating space to explore. Someday I'd like to get the opportunity to write about it; but for now, my remaining two articles will remain at the overview level. In the second article, we will look at the standards lifecycle and in the third, we will discuss standardization in the age of technology convergence.
Chair, IEEE Standards Association Corporate Advisory Group
I also agree that the answer is yes.
Sometimes the standards process can be overtly influenced by industry players and there are plenty of examples of companies pushing a line that suits their product offering better or creating a "barrier to entry" for new players by addign extra layers of complexity.
This issue is pretty minor when you consider the disasters that can occur when there is no standardisation. Here are some examples that occur to me:
1. lack of interoperability
2. safety is not ensured
3. environmental pollution including the RF spectrum
4. major players can dominate through misinformation and prevent new entrants from getting traction
5. best practices are not developed or used
6. industry level learning is not captured
7. synergy and cooperation are reduced
8. emerging economies don't get the jump start they need and can repeat the same mistakes
If you take a big picture, long term view, standards are an essential tool for ensuring that industries and markets have the opportunity to reach their maximum potential.
I agree that the answer is yes.
That being said, sometimes standards do end up over complicated. This is a common "barrier to entry" tactic that established companies with representatives serving on the standards committes can use.
However this is a minor issue compared with the disasters that occur through not having standards. These are:
- lack of interoperability
- safety is not ensured
- environmental pollution including the RF spectrum
- major players can dominate through misinformation and prevent new entrants from getting traction
- best practices are not developed or used
- industry level learning is not captured
- synergy and cooperation are reduced
- emerging economies don't get the jump start they need and repeat the same mistakes
So yes, standards are an essential strategy for industries and markets reaching their full potential.
Dueling standards are a problem, yet 2 is better than N >2, and at least it offers some basis for further convergence. IEEE is a great organization for the late stage and finalization of a standard, yet more efficient and practical methods and organizations exist for early stages. UPF/CPF competition and SVA/PSL competition exists because certain companies continues to believe that a proprietary language position provides them a competitive advantage. In contrast to standards is the open source movement in which the release to the community is immediate and it is open to contributions. One example is the Google Wave wave protocol which was established and released coincident with the product announcement. Either way interoperability is crucial for innovation and engineering efficiency so that we can all work on the core technologies.
Sometimes standards can be a joke. In EDA industry it is just always. Look at the CPF vs. UPF fight. What is the standard here? And why is IEEE supporting only UPF (1801)? And, there is still inconsistency in the implementation of these standards by various companies. A recent article by a designer at ST-Ericsson in deepchip.com (http://www.deepchip.com/items/0483-01.html)shows how one company uses "so-called" standard to impose its monopoly on tools. And the interesting fact is IEEE is endorsing this standard. I think in EDA world, standards are used to establish tool monopolies.
I still think standards are important and will generate innovation in the industry. But when the principal arbitration agency such as IEEE does not take active interest in driving the standard, individual companies take advantage, and in case, behind the aegis of IEEE, which is a pity.
The difficulty with this discussion is the complexity of the topic. Almost every one would agree that standards for measurement matter. Standards for similarity of process or product certainly matter - consider the world-wide success of ISO 9000. The article appears to address standards for compatibility, or interface standards. These standards are often a struggle to identify, implement or use due to competing interests.
The rallying cry "Open Standards" is often heard when discussing compatibility standards. There is almost no one who is interested in open measurements standards or open ISO 9000 standards. The reason for this is simple, when the standard controls an interface, all that wish to use the interface must use the standard. If the compatibility standard is controlled in some way, all users are affected by that control.
The current growth of the standardization ecosystem is directly related to the inability to easily resolve the struggle for control of compatible interfaces. Controlled compatible interfaces are the basis of some of the most successful companies on the planet - Intel x86 interfaces, Microsoft Windows APIs. Until divestiture the phone company's' control of their interfaces was one important way they controlled their markets. These commercial successes show that companies who can maintain control of their interfaces can reap enormous rewards. In turn this makes companies work very hard to achieve such control, by controlling the forum where the standardization is taking place, by asserting patents over the interface, and by attempting to dominate the market that supports the interface.
Many different approaches are being attempted to address this issue of commercial organizations dominating their markets. In the past anti-trust law was applied to reduce commercial control of markets. Anti-trust law has not been effective when applied to patents or trade secrets on interfaces. Another legal way forward is developing and testing different licensing regimes to see which are best accepted by the developers, implementers and users.
The technical way forward, where the interface is completely implemented in software, is to implement standardized adaptability mechanisms that select between different potentially compatible interfaces, public or proprietary, allowing the end user to choose which interface is desired. This is made more attractive by the growing ubiquity of the Internet which allows for the downloading of a different interface when desired by the end user. This technical direction also requires models for pricing of interfaces by use, download or support that not yet understood.
All these different legal and technical approaches can be tried because of the very large and expanding standardization ecosystem. Think of all standardization as an evolutionary process. Different standards are created to serve different needs and over long spans of time the standards that serve the most needs are the most successful. The plethora of standardization organization and their compatibility standards is not the problem, it is the only fair way of identifying some solutions.
Do standards matter - of course. I think a more interesting set of questions is how they get made, how open they are, what are the terms of IP licensing, and how much vendor lock-in is implied by them.
My Mom the Radio Star Max MaxfieldPost a comment I've said it before and I'll say it again -- it's a funny old world when you come to think about it. Last Friday lunchtime, for example, I received an email from Tim Levell, the editor for ...
A Book For All Reasons Bernard Cole1 Comment Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...