Wednesday, April 8, 2009

What’s “Open” Got to Do with It?


In early March, I happened to be in Washington meeting with Ed Markey. It turns out that the incredibly important words that required the $6.6 billion in smart grid demonstration projects to use "open standards and internet protocol" was his amendment! These words were modified in the final Economic Recovery Act by industry lobbyists to include "where available and appropriate."

I was in Markey’s office to explain to him why these same words should be applied to wireless demonstration projects in the transportation sector, in health care digitization efforts, and likely in education, although I don’t know. Markey was excited by my interest, and wondered if I could explain to the layperson why open standards mattered.

A week and a half later, I bumped into a state Secretary of Energy – one of the very people who would get to spend the smart grid demonstration project money. This person didn’t understand the implications of “open standards” and asked me to explain it. Over the course of the last month, I’ve met with high level officials in transportation, energy, and environment positions from several states, none of whom understood the value of openness.

We are about to spend billions and billions of taxpayer dollars on technology infrastructure and many of those advising precisely what to buy have every incentive to say that closed proprietary systems, networks, devices are the best way to go. How does this missed opportunity make you feel?

A friend blogged on this subject and I loved his headline:
Using Public Dollars to Build Proprietary Systems?

Proprietary systems have their own secret languages and secret rules. You can play only if you are invited in (by buying the ratified stuff) and you can only play the games agreed upon (your ideas for new games or new ways to play the old games are unwelcome, unheard, and impossible to incorporate). Examples of closed proprietary systems abound, but a nice irritating example would be how you have to throw away your current cell phone if you want to change carriers.

Open standards mean that different people/companies/devices could, if they wanted to, find common ground.

Here, excerpted from a piece David Reed wrote for The MIT Communications Futures Program Principal Investigator Blog is a nice description of how the internet -- which is an open standard -- works:

“The Internet is a set of agreements among members (who happen to control small, medium, and large networks). The agreement required members to carry each others’ packets, delivering them via best efforts to the hosts at the edge of the network—your laptop, Google’s server…each member of the Internet who contributed to the mutual enterprise gained connectivity disproportionate to the member’s contribution.”

As David puts it, "The Internet is not a technology, but a set of interoperable standards."

Open standards give the ability to evolve over time.

Sure, proprietary systems can evolve, the speed depending entirely on competitive pressures. Most government contracts come with nice long contracts: three, five, ten, and even 99 year terms! Why bother to innovate during the first seven years of a ten-year contract? Steve Crocker, one of the Internet’s founding fathers, wrote a really wonderful piece for the New York Times that describes how the Internet’s open standards were able to evolve over time. As he told me “We had no idea when we started [forty years ago] that this is where we’d end up.” Of course, who among us can predict the future?

Another friend offered a simple test: “If you think this is the final and best version, buy the closed proprietary system. If you think it will continue to evolve over time, go open.”

Open standards invite and encourage participation

From a Steve Crocker email “Open standards become particularly important when they enable new products and services to be built on top of existing ones. Openness is not just about enabling others to build the same products and services and compete directly. It’s also about enabling huge vistas of new inventions that brings the enormous expansion and payoff from new technologies.”

I'll close with Steve's penultimate paragraph from the NYT:
“As we rebuild our economy, I do hope we keep in mind the value of openness, especially in industries that have rarely had it. Whether it’s in health care reform or energy innovation -- [OR smart transportation adds Robin] -- the largest payoffs will come not from what the stimulus package pays for directly, but from the huge vistas we open up for others to explore.”


Some interesting links about open standards not referenced in the above:
In health care and in promoting multimodal transportation.


4 comments:

Anonymous said...

This post is so correct that it's almost scary that it has to be written.
In the time since the stimulus bill was passed there have been varying cries from the energy vendor community that standards don't exist, can't be done, aren't desirable and so on. Time to call BS - anyone who lived through the telecom, IT, Internet or Cable revolutions can testify to the power.
"if available?", please. Is IP networking not available?

Mike Weisman said...

Robin;

I think you are confusing two different concepts, 'open' and 'standards.' Let me explain.

'Open' the way you are using it here means the technology or the engineering is open to see. Anyone can observe it and see it. You imply, as I understand you, that by being open and being able to see the technology, then someone could 'build to it.'

The openenss exposes more than that. It may be that the technology is open, but poor quality. It may be that by being open, we can readily see that, but there is nothing we can do about it. It may be open, but no one else will want to adopt it for various reasons; poor quality, doesn't work well, is buggy, etc.

Using 'standards', which is what David is talking about, is another subject entirely, but it is the subject you are really talking about. Standards mean the particular technology, modality, OS, connection, has been reviewed by a public body. Not public as in government, public as in nudity. It was adopted at a meeting at which a bunch of interested people sat around a table, and everyone could see everyone else.

Standards bodies are much more complicated than that, or course. But standards bodies arrive at a mutually agreed upon and publicly available or published set of terms and criteria that come to be the definition of the standard. Then, since the standard is available, and has been tested, engineered, and published for review, competitors can 'build to it' and have a somewhat level playing field. They can make improvements that can be brought back for further refinement by the standard setting body.

What you call 'open' you mean to say 'standards.' For example, outside the US, mobile phone networks operate on a publish standard, the GSM.

When you buy a CD, it works in your player bcause the CD technology is a published standard, originally developed by Royal Phillips.

When you turn on your new digital TV, it works, no matter who you buy it from, because digital TV is a published international standard.

Compare that to 'HD Radio' or digital radio (two different techs), which, in the US, are broadcast using proprietary standards. Your HD radio won't work anywhere else, and it a few years, when the rest of the world switches off their FM radio for digital, your American FM radio won't work outside the US. The US did not adopt the international standards.

So standards matter. Openness is also generally a good thing, but the point is that the language should say 'published standards.' This way, even competing standards, as long as they are published, can compete for space and innovators can use them.

Robin Chase said...

Mike,

Thanks for the good points and I will try to take more care in my words. One thing I'm concerned about is how to encourage/foster an open standards-making process, so that we can create and adopt new standards.

I think the IETF RFC process has been a pretty open consensus driven
process. IEEE standards seem much less so, being more driven by industry representatives, often adopting privately developed standards (like the CD). How long does it have to take to finalize the 802.11s?

Mike Weisman said...

Robin;

We shouldn't let this devolve into a semantic duel. But the point you make about IEEE is right on. The IEEE 802.11blah process is lengthy and full of industry representatives. But that is because it is occurring in public, in a structured process intended to make sure the final standard is a good one, and the public vetting process takes time.
I met a woman, a non-engineer progressive activist type, who is on the IEEE standards committee, and I had the chance to listen to her describe it from her vantage point.
In my opinion, the consensus part is not as important as the published part. For standards to offer the public benefits of what is essentially a type of monopoly, it must be a public process with a published deliverable.