Oliver Bell’s Weblog

During the conference I attended last week organized by the Computer Society of Sri Lanka I was asked a number of very thoughtful questions, I doubt that I’m alone in dwelling on questions that are posed to me long beyond the point at which I throw out an initial answer.

One of those questions at the event in Colombo was focused on the role of open standards in the industry today. It is pretty clear that the question has relevance on a number of different levels, starting with some of the current debates (Open XML & ODF anybody?) and carrying on through to the long term implications of the role of standards in our wider and rapidly advancing software industry.

In considering the answer to this question I have been trying to think through a number of related dynamics including (but not limited to!) who is helped by the existence of open standards, evolution of technology, competition between standards and technologies, and what role all of this might play in national government policy.

Of course all four of those strands are woven together, it might help to talk about each one in turn first of all then maybe bring them back together in the form of national policy.

I tend to think about two main groups who are helped by a documented open standard, developers and end users. Under the banner of developers I’m bundling system integrators, independent software vendors, providers of web based services and individual developers together.

For developers the documentation that an open standard provides ensures that it is possible for any one of them to build code that will have a defined place in a the market place, other developers will be able to interact with the tool that is produced. At the same time some strands of who the target user base will be is understood from the outset by looking at who the consumers of that standard are. This means that the software that is developed will reach users, both domestic and overseas, in a form that is immediately useful to them.

For the user open standards assist in a couple of really important ways. First of all they provide a way for users to choose technology that ensures it will address the need that they have, be it a need to communicate with another device or process data that is provided by a different system. Secondly open standards frequently provide choice for the user by broadening the selection of technology that meets their need, this coming back to the right that any user has to select a particular operating system platform, or application to run on that platform.

Thinking about the evolution of technology is important when considering open standards and how they help or hinder innovation. The first area of debate for me here is around the direct role of standards in the evolution of technology, are the standards themselves the point at which technology evolves or does the software evolve first and the standards follow.

Since the delivery of the first IBM PC loaded with the Microsoft licensed PC DOS the model for our industry has been to develop the software first, then the standards follow as the technology evolves, and I personally see a lot of value in the innovation that is enabled by this existing model.

Assuming there is a cycle that begins with propriety implementations that result from research and development investment that eventually work  through to a documented, ratified open standard then I think we can find examples of technologies that are part way through this cycle today.

Instant Messenger technologies might be one such example. Not so long ago we had a small number of free standing networks that came with the invention of the technology, each one experimenting with different functionality, network management techniques, marketing methods etc. Today those networks have documented and shared basic protocol standards that enable users to communicate between the networks and I’m pretty confident that the future will involve and increasing number of standards for instant messaging as user behaviors, functionality requirements, bandwidth implications etc. are better understood and the industry evolves to a point where there is enough understanding of the technology to develop standards in this area.

So when you look at this cycle maybe standards play a vital role in driving the maturity of well defined and understood technology, but in my opinion they may have less of a defined position around emerging technological advances.

I think the final point around evolution is that innovation does not end with the publication of a standard, computer scientists are amazing people and will always be finding new ways to solve technology problems. In some cases this will mean that a standard can be revised to include a new technique, or in many cases it may mean the development of a new standard that supercedes whatever went before it.

Next lets take a look at the debate around the issue of competition between open standards. Different individuals or companies can work on different assumptions around what a user base requires and in many cases will get together with like minded companies and individuals and develop some standards documentation that has mutual advantage to those involved.

Sometimes these ideas compete and we see multiple standards appear at the same time. After various competing efforts are published there tends to be one of two outcomes, either one standard never sees any form of implementation because the other has a really clear advantage in one area, or different segments of markets choose different standards to satisfy different needs that they have.

I think the VHS vs. Betamax example is a really good one here. The general assumption seems to be that VHS won the battle between the two standards and Betamax went away. In reality Betamax lived on (and still may for all I know) in the professional recording industry where it evolved to meet important needs those those constituents had.

When I apply this thinking to some of the current Open XML and ODF discussions I worry about a phrase that I have heard countless times in the last few months which I’ll paraphrase as “why develop a new standard when there is already one in the form of ODF“.

If that were really the right approach standards then there is a good chance that I would be sending mail to colleagues via x400 email and using TP4 to connect my devices at home together, both were existing standards at the time when we all chose SMTP for email and TCP/IP as the basis for the Internet as we know it today.

From a personal point of view I think emerging and competing standards put us in a position where market dynamics can make decisions around the technologies that it wants to use, in some cases similar standards will continue to exist side by side like VHS and Betamax, and in other cases one standard will become pervasive as it did with SMTP over x400 for email transfer.

By the way, I have to offer my apologies to those of you who were implementing mail systems in the mid 90s,  I may have just brougt back the recurring x400 nightmares that you thought were behind you. All those years of therapy were worth it, and hopefully you’ll soon put it all behind you again.

Finally, when I pull all of this together under the headline of national policy we have to look at the effect that adoption of standards can have on a national software industry, on citizens and other technology users, on development of internal government systems, and on the ability for technology that is provided by the commercial base to interact with other technology either domestic or overseas. That is a big set of issues, I think I’ll leave it for a future post…