Oliver Bell’s weblog

For computing today interoperability is probably as important to the industry as some of the longer running conversations around subjects such as security and reliability. In previous posts I have talked a great deal about some of my own experience and views in this area.

Interoperability itself has evolved for the computing industry and a growing challenge over the last thirty years. At the start of the journey the picture was a much simpler one than we know today with most systems being delivered by a small collection of very large single vendors, starting with the hardware and storage systems and working right the way up through the stack to the applications that directly serviced the needs of users. At this early point in time interoperability was often defined and resolved by a series of infrequent gateways between the vendors themselves, connecting their large managed customer networks together.

The advent of the desktop personal computer and the consumer operating systems from Microsoft, Apple and more recently the open source community brought with them a new world of consumer choice. Today your desktop system will contain an array of fully interoperable technologies from different vendors who are doubtlessly operating on different continents. The processor, the disk array, the operating system, your applications and your user input devices are all highly interchangeable, and regardless of the choices that you make for each of these components you will still be able to communicate with friends, colleagues and a range of information providers over the common network that we know as the Internet using common sets of applications.

Given the magnitude of these changes many will agree that they have happened in a relatively short period of time. Some questions around what interoperability is or how issues should be resolved are still the subject of significant debate within the industry. Sometimes those debates are driven by large multinational companies who have an interest in one model over another, and more frequently they are driven on a technical level by individuals or groups who can see conflicting methods of solving similar problems.

One of the measures of maturity of any area of science in the involvement of academia in managed research around well defined challenges and real world problems. Interoperability is one area where we are just starting to see academic projects focus on the technical, organizational and semantic language questions that interoperability brings. Every study adds to an evolving view of what the industry needs to do from here, along with delivering well researched data and real world case studies.

As a company Microsoft spends a great deal of time engaging with universities around the world, posing questions that help us gain a better understanding of many aspects of the industry including challenges that our customers are facing and how we need to develop our own plans to support a rapidly evolving landscape.

One such piece of research reached its conclusions earlier this month and was jointly published by the Berkman Centre for the Internet and Society at Harvard Law School and the Research Centre for Information Law at  St. Gallen University in Switzerland.

The conclusions are interesting, and I would encourage you to read the report itself which you will find here rather than just relying on my own observations.

The study itself looks at three use cases for interoperability. DRM in the context of online and offline music, identity management systems and the interactions between web services. There of the key findings can be summarized as follows;

First, interoperability does not mean the same thing in every context. Nor is there the same need for interoperability in every situation. Highly secure systems, for example, will probably not have the same interoperability requirements as consumer systems.

Second, there is no one-size-fits-all method to achieving interoperability in the ICT context. Interoperability can be achieved by multiple means including the licensing of intellectual property, product design, collaboration with partners, development of standards (open or defacto), and governmental action. The best path to interoperability depends greatly upon context and which goals matter most, such as prompting further innovation, providing consumer choice or ease of use, and the spurring of competition in the field.

And third, the report comes to the conclusion that the private sector generally should lead interoperability efforts. The public sector should stand by either to lend a supportive hand or to determine if its involvement is warranted. Trying to impose universal answers can produce unintended consequences such as curtailing innovation, limiting consumer choice, or reducing overall competition.

Given many of the policy debates currently underway in this area across Asia, the third finding requires some more investigation particularly in the context of the joint regional objectives that the industry and governments share in the area of economic development and growth.

I have spent most of the last week in the United States, joining the American half of my family for thanksgiving. It was great to see everybody again, and fun to see the smaller members of the family getting a step taller since we last saw them a few months ago.

Most of the weekend however has been eaten up by the long journey home, leaving Maryland at 4:30am to catch a 7am flight from IAD to SFO. Personally I enjoy the long haul flights, they are usually a great time to think, clear down my inbox and work through the constantly growing backlog of reading materials. Very tranquil and productive periods of time.

Today was a little more dramatic than usual. United was my carrier from IAD to SFO, flight number UA 187 (departed November 24).

The approach to SFO was the usual route, over some multicolored fields of some sort that I have been meaning to find out about for years, but never seem to get around to asking anybody who might know.

As we got about a couple of miles out from  the runway a JAL 747 could be seen in the distance running in parallel to our flight, the person sat next to me pointed it out first of all… saying something about the fact that you don’t usually see a 747 so clearly while in flight. It was a fine, clear and sunny San Francisco morning!

Oddly the JAL 747 seemed to get closer over the next couple of minutes, and the view got clearer. I assumed that SFO was running a parallel runway system and that we were about to partake in some sort of spectacular synchronized  landing.

As we got to about 500ft above the ground (I’m neither an altimeter or a pilot so I’m only guessing at the height) it became evident that something else was happening. The JAL flight seemed to be turned slightly in our direction, and myself and the passenger next to me began to wonder if the two planes were not heading for the same runway as apposed to adjacent ones.

A moment later, when the JAL flight was closer than any other plane I have seen from an aircraft window seat the JAL flight banked hard left and climbed out of harms way.

I’m sure nobody was really at risk throughout the five minutes or so that elapsed while all this took place, but I am now curious about what actually happened. As you would expect the United crew didn’t pass any comment on the event. If anybody reading this knows of an archive where such things are documented I would welcome a link to it.

Maybe it was just an everyday occurrence, but it didn’t feel like one. For me at least, it was a little close for comfort!

I have a couple of links for you this morning on the progress of the comment resolution process, as we head towards the DIS29500 Ballot Resolution Meeting at the end of February.

The first is to the web site of Ecma International who currently lead the work on the DIS29500 specification, Rex Jaeschke is the JTC1 project editor and Ecma’s TC45 is actively supporting his work.

From the Ecma International web site;

As outlined in news release dated 18 October 2007, Ecma is maintaining active involvement leading up to the ballot resolution meeting 25-29 February 2008 by supporting the ISO/IEC DIS 29500 Editor, Rex Jaeschke, who has been tasked with producing a proposed disposition of all 3522 comments received during the ballot period, and by maintaining a web portal accessible by National Body members which will host current drafts of the proposed comment dispositions.

The second link is to Brian Jones’s blog. Brian is the Microsoft representative on the TC45 committee and is without doubt the closest person in the company to the work that is currently underway. He talks about how the TC is working through the comments, some of the work that has been done to address the commonality of comments from different countries, and how many dispositions are currently available for review by the various national bodies.

From Brian’s blog;

There are currently 662 responses, and the plan is to provide updates to this list every few weeks. We still have almost 2 months until the deadline, but given that we have a lot of issues to work though, we thought it would be best to provide the responses earlier than the Jan deadline to allow for more time to discuss the issues.

Needless to say, there is still a lot to be done in the lead up to the BRM. For each of the national bodies around the Asia Pacific region though it is probably a good idea to ensure that you have an opportunity to review all of the work coming out of Ecma at the moment in preparation for that meeting.

Earlier this week I spent almost seven hours stranded in Baltimore Airport waiting for a connecting flight, this gave me plenty of time to catch up on reading some of the blogs that I have lost track of during the last few weeks.

One of the posts that caught my eye was an opinion piece on Bob Sutor’s blog, the second in a series he has been writing up on the Open XML ballot resolution process.

Bob is a smart guy and has a lot of good things to say about standards in general, certainly about the way that they relate to the industry today.

In this particular case though he has gone a little beyond the usual stylish and subtle way that he seems to put across IBM’s PR messaging on Open XML. As usual he puts his opinion forward as fact, but in many parts of the post it seems that he may be misunderstanding the well established ISO processes.

Jason Matusow has taken some of the questions raised by Bob and responded accordingly, both posts make for interesting reading.

As is always the case with these types of conversations there seem to be multiple views put forward for the benefit of different audiences. I find that the obvious authority on these types of debate are the national standards organizations who always have a very clear and level headed view of issues that are subject to broad debate on the blogosphere.

When it comes down to it the regulations that govern and direct the actions of JTC1 are generally pretty clear, and I would always suggest reading them for yourself, along with anything you can find posted by anybody involved directly with any standards process, then forming your own view on what needs to be done.

So, if you find yourself stranded in an airport somewhere with free wireless and nothing better to do then I’d suggest having a read of Bob’s views on the BRM process, Jason’s response and finally the BRM FAQ that was put out by JTC1 earlier this month.

Interoperability this, interoperability that… it is clearly an important conversation in our industry today, and one that spans the whole range of vendors, technology areas and end users.

Frequently conversations around interoperability get pulled in the direction of the technology that is in use, but while there is clearly work to be done here it is important to recognize that solving this problem also means tackling a number of other dimensions of the issue.

In 2004 the IDA in Europe issued their first version of the European Interoperability Framework, several IF documents had proceeded it but as far as I remember this was one of the first documents to look beyond the technical issues of interoperability and begin to examine other challenges that would need to be dealt with to provide truly interoperable systems across agencies in Europe.

On a very simplistic level the document broke the interoperability challenge down into three distinct layers;

Technical Interoperability - lets move on…

Semantic interoperability - this involves looking carefully at the language that is in use between agencies and business processes and lays down a goal of harmonizing elements where it made sense to do so. The goal being that a descriptor that might be used to define a part of a process within one agency would be understood to mean exactly the same thing in another agency. If I am tagged as an individual receiving child support benefit in one system then all other systems should know exactly what that means.

Organizational Interoperability - this section talks about something that is clearly a huge challenge for any organization, looking at how business processes intersect and what type of organizational changes need to be made to ensure that those processes work seamlessly together, at least when viewed by the consumer of any resulting service.

 So, with the history lesson over lets get back to the point of this post. Over recent months I have been rolling the IDA model around in my mind, and keeping an eye open for examples of interoperability where only the semantic and organizational layers of the problem have been dealt with to fix a particular business challenge.

Two weeks ago I was at a Microsoft global get together in Thailand and happened to bump into another blogging friend of mine, a chap that some of you will know by the name of Doug Mahugh. Doug has been in the industry for a number of years and has seen several sides of it, as a result he is a great guy to pose questions like this to, and invariably he will always have a smart answer to hand. This time was no exception.

Doug pointed me to an example on ushero.org that talked about the way that the emergency services work together when they are jointly searching disaster zones for signs of survivors or those in need of help. During the conversation Doug talked pretty extensively about the work that took place straight after Hurricane Katrina in New Orleans, and presented a very simple model designed by the US emergency services to allow several different services to work together without duplicating efforts, and at the same time providing very clear communication about the status of the searches between teams on the ground.

As I say, the model is a straight forward one. As a team enters a building they draw a simple stripe on the outside by the door from top right to bottom left, this tells any other team that comes by that the building is currently being checked. When the team is done they draw a second stripe from top left to bottom right that completes an X and then use the four quadrants of the graphic to convey important information about what they found in the building.

Click on the thumbnail above to enlarge the picture and take a look for yourself.

In my view there are lot of lessons to be learned from this by the information technology community. As a community we get very involved in the technical challenges of interoperability when sometimes, as in the case above, a simple change in the design of the business process or the language used to communicate between teams might have been all that it would take to solve the problem that we’re faced with.

Making changes to business process can frequently meet resistance due to a need to involve senior management or other cross sections of the organization, but as the team looking at interoperability of government services in the European Union had already worked out some years ago this is sometimes the quickest and most cost effective way of tackling a given problem.

If you are aware of more examples like this then I would love to hear about them!

We cannot dismiss the technical work that need to be done, and is being done in the area of interoperable systems. 

While it might sound obvious, we often seem to miss the fact that there is great benefit for all involved to look at these issues in the context of the wider business challenge that is being solved.

Those of you following the progress of DIS29500 (Open XML) through ISO might be interested to read an FAQ that was posted by JTC1 on November 15th. You will find it at this link.

There are discussions in the blogosphere about the process leading up to and during the BRM in Geneva, some correct and some of them a little misleading. JTC1’s own FAQ should clear up any remaining questions and I would suggest that it is the best source as a primary reference on the subject.

You might have caught a story in the press today about a translator project that Microsoft and the Digital Accessible Information SYstem (DAISY) Consortium are establishing on SourceForge that will provide both some new functionality in Microsoft Office along with an offline capability to convert Open XML files into DAISY format for use with a range of accessible technologies.

Here is a link to one of the stories from this morning from eWeek, the quote below is from Reed Shaffner, the Program Manager in the Office group who has been working on this from the Microsoft side;

The project is being hosted on SourceForge, with the first beta code expected by early next year and release by March 2008, Shaffner said, noting that the plug-in will work with all Word documents created with Office XP, Office 2003 and the current Office 2007.

“Essentially what will happen is that the plug-in will convert an Open XML file to an intermediate Daisy XML file in the Talk Book format. Customers can then use one of many tools, which are already available, to create a bunch of different accessible outputs, be it Braille or a really rich audio file that allows them to navigate by heading or page number and navigate tables with much more detail than they would typically be able to,” he said.

George Kerscher, secretary general of the DAISY Consortium, explains the DAISY translator plug-in project this way:

“Microsoft’s announcement is monumental in greatly facilitating the availability of text in DAISY books. It provides a clear, production path for organizations and universities who will be able to use the Microsoft plug-in to move into DAISY XML. Putting tools in the hands of people who create content is a giant step toward creating equal access to information … It’s going to move DAISY … from the niche of the libraries for the blind community into the mainstream.”

This project provides both a capability to convert Open XML based documents to the intermediate DAISY format, but will also deliver an open source based reference implementation that any other developer can use when building their own similar translators.

Views and understanding of every topic evolve over time, one of the more “interesting” elements of working for a company as large as Microsoft is that it can take a long time for change to be recognized in the wider marketplace. While internally it is often possible to point to the exact moment when a decision is made it is frequently impossible to highlight the moment in time when the industry accepted that decision as part of the way that the company works.

One example of this is probably our relationship with the open source community. On a regular basis I end up in conversations where there is a general assumption that open source and Microsoft are head to head competitors who cannot face being in the same room together, or partnering to solve customer and industry problems.

If you look back seven or eight years that was probably a very true statement, today however it is a very different story.

In reality most open source development is good for Microsoft, and there are a number of examples of where we are fully supportive of this work. Open source development on the Windows platform brings a number of benefits to our customers, it revolves around developers who are using tools to support the Microsoft platform and in turn build a wider array of application and solution choices for everybody involved.

Open source in itself is not competitive to Microsoft, where there is a clear competitive environment is between products such as Microsoft Windows and Linux or Microsoft Office and OpenOffice.org.

This is especially important when looking at government technology and procurement policy. Today in Asia I seem to encounter two flavours of policy being considered in this area. The first flavour involves governments who are looking at policy to mandate the use of Linux as an operating system within agencies, the second often involves governments putting a pervasive mandate for the use of open source software across all levels of the software stack.

In my opinion neither of these options serve government employees, service providers, citizens or businesses well. Delivery of any computing system in today’s world will involve the selection of best of breed software across the board. This is important due to the differing use cases that platforms are designed against and wide array of choices made by software users both at home and abroad that governments need to interact with.

Microsoft’s commitment to working with the open source community is somewhat diverse, and is designed to deliver an environment where technology interoperability issues are resolved between vendors rather than leaving them for customers to resolve, as the industry as a whole seems to have done in previous years.

Some examples;

• Previous posts have discussed the “Windows Principles” which lay down a very high level framework that is designed to offer unprecedented levels of predictability for any vendor wanting to partner with Microsoft, or build applications on the Windows platform.
• “Interoperability by Design”. In 2003 Bill Gates wrote a public memo with this title, the goal was (and remains) to push the product groups to think hard about interoperability as they go though internal design processes, we’re starting to see the fruits of this work in products such as Office 2007, Windows Vista and the upcoming release of Windows Server 2008.
• Building community. There are a few great examples here, starting with the Interoperability Executive Council where we bring some of the top CIOs in the world to Microsoft’s corporate headquarters about twice a year to discuss the issues that they have with interoperability, some of these CIOs are customers that do a lot of work with Microsoft and some are quite the opposite. Concrete solutions come out of these meetings and again there are examples of product design changes that have resulted from these discussions.
• The second example of building community are the relationships that we’re building with companies like Novell, Linspire, Xandros etc. These partnerships involve the sharing of API information, intellectual property and sometimes code to ensure that interoperability issues get resolved in products before they ship.
• Finally, having worked for the company for almost a decade and a half now I have watched a significant shift in the way that Microsoft thinks about the world of standards. In the mid 90s most standards activity was the responsibility of individual program managers and product groups. Sometimes this worked and sometimes it really did not. Today we have a core team committed to our activity in the standards world, looking both at how we apply the expertise that we have in the existing standards community along with how we standardize some of our own technologies.

Beyond interoperability between open source and Microsoft platforms there are a number of evolving activities that look at how Microsoft builds a closer working relationship and deeper direct involvement in the open source community.

The Solutions Sharing Network (SSN) was probably one of the first examples of this, and directly relates to sharing open source applications between government customers. This was a project that started with a couple of cities in Europe in early 2002, then was officially released for general use in November of the following year. The goal at the time was to provide a safe and secure environment where customers could build communities of practice to share both business practice and solution code. Today the network is supported by a central site called SolShare, and a network of private sites that are in use by many governments around the world.

CodePlex is a more generic project designed to encourage the sharing of projects between developers, today there are almost two thousand live projects running on this shared environment, including many from Asia, a couple of which I have highlighted in previous posts.

A further piece of the jigsaw involves the work that Bill Hilf’’s team have been leading in Redmond, running an interoperability lab where there is a phenomenal level of testing and learning about how Windows works alongside traditionally competitive offerings from the open source community and other vendors.

Finally, out of almost 150,000 projects on SourceForge, one of the premier open source project sites, over 52% are multi-platform and include binaries that will run on the Microsoft Windows platform. From time to time Microsoft has been directly involved with creating projects of our own on SourceForge, most recently including the translator project that is looking at conversion of documents between Open XML and ODF.

So, what is the point I am making here? It is a simple one. Solutions to any technology or technology policy issues will often involve diverse choices from a variety of vendors, and when formulating government technology or procurement policy it is important that every door is left open to allow agencies to choose the best of breed technology to build their solution and interact with their constituents.

The industry as a whole has made some great strides over the last decade to tackle complex interoperability issues, and the open source community is obviously a key component of this. An open source component of a national technology policy can bring some real value, but only alongside all of the other technology that is available in the market.

Some of you might remember this post from a while ago, a few short weeks later and I’m a committed user of Windows Live Writer for preparing offline blog entries.

This morning I noticed that it has now moved beyond beta and has become a release product. If you’re using the beta, or would like to take a look Writer, then you are able to download the release version from here.

Some of the marketing blurb from the download site;

Publish to most major blog services. Including Windows Live Spaces, SharePoint, Wordpress, Blogger and many others. Don’t have a blog yet? Writer will help you set one up!

Create a compelling blog easily. Insert photos and videos, maps, tags and lots of other cool content, all in a snap. Powerful editing features include tables, spell checker, and quick hyperlinks.

Preview before you post. Writer shows you what your blog entry will look like before you post it, using your blog’s templates and formatting.

Compose your entries offline. Publish them later, when you get back online.

Enjoy!