Jerry Fishenden's Weblog Archives - Jan 2005

January 21 2005

I mentioned on January 3 that identity management is a hot topic. I’ve been enjoying reading Kim Cameron’s postings at www.identityblog.com along with other related sites that you can find indexed off Kim’s page. Identity management may not sound like the hottest topic on the planet, but for anyone using the Internet it’s clear that today’s approach is unsustainable. The idea I should have a separate logon credential for every site I use is unscalable – and painful. And I don’t like the way banks all use the same “secrets” for their services: it’s inherently insecure and the number of people who now know my mother’s maiden name and my first/last school etc must number in the tens at least. We need a much better, more trustworthy system than this. Kim’s ideas strike me as the best contribution to this debate that I’ve seen in a long time.

January 20 2005

Microsoft has launched a paid-subscription version of Outlook (the Office product that provides e-mail, calendar, contacts etc). This is the first time that part of Office has been made available as a subscription service. “Microsoft Office Outlook Live” includes a subscription version of Outlook 2003 to connect with Hotmail or MSN e-mail accounts. For $59 a year, customers will get an e-mail account with 2GB of storage and the ability to send individual messages with up to 20MB of attachments. Customers can also check multiple e-mail accounts, including corporate accounts that are managed through an Exchange server.

This type of subscription based model could be one part of the answer to the so-called ‘digital divide’. While most surveys suggest the ‘digital divide’ is more myth than reality (since most people who want to can find some way to access the Internet, even if it is to ask a friend or make use of a work computer), if computers and software services can be provided much more in the way we think of mobile telephones or TV subscription services, it would doubtless lower the barrier to entry. What we also need is movement on the hardware front – so that a basic PC could be offered at a suitable price point that it can compete in the domestic space in terms of acquisition cost and provide subscription software services. If someone could acquire and plug-in a basic PC (or alternative form factor, such as a more advanced set-top box) and have it automatically connected to software subscription services, it would potentially enable a whole new segment of the population to participate more easily.

Many businesses have used software subscription models for years to manage their costs in an orderly fashion and to be able to take advantage of new software as and when it becomes available. Consumers tend to buy a PC with bundled software and run that until its hardware is deemed to have reached end of life. A wider type of subscription service in the consumer space might be a good way for those of us who use PCs a great deal at home to cost-effectively be able to upgrade and add to our software portfolio as and when we like. If the OEM hardware companies also partnered in this, it would provide a welcome subscription service that enables home users to continue to update their system and take advantage of new software and hardware – but in a cost effective and more consumer friendly way.

January 18 2005

Lee Gomes in the Wall Street Journal (Word Processors Offer Too Much Fancy Stuff; Basics Are What's Key, Jan 17) mentions the various approaches of word processing software over the years, emphasising some of the features he appreciated in WordStar (which preceded both WordPerfect and Word).

It made me think, and not for the first time, that for all the benefits of word processing that in some ways we have not moved on from the days of typewriters in terms of our comprehension and thinking about what a document is. Most documents remain static: they are fixed set of words developed at a moment in time – an approach which Adobe’s PDF format probably represents most clearly. Whilst this suits some types of environment (notably legal), equally there is another model which has not yet really developed: this is the idea of the dynamic document. If you think of the best Web sites, they don’t use static Web pages. They are data-driven and draw pages and their content based on an interaction between the user, their business logic and the data they contain.

But why is this model only used on the Web? Why is it not more prevalent within companies, business environments and government organisations? Take for example fairly standard documents, such as monthly sales reports, or reports on the number of benefits claimants seen and successfully processed by, for example, a DWP office. Currently the prevailing model is for someone to laboriously compile the figures, create a new word processing document and then put those figures into the document. It’s not only a time-consuming process, but the document itself is a static snapshot at the moment in time it was created – and will often be out of date by the time it’s been through its whole creation cycle.

Instead, imagine a document that dynamically updates itself: as it is opened it populates the document with the latest information drawn from the underlying databases. Readers of the document are no longer merely passive, but can interact with the data.

This is the type of experience we enjoy on the best Web sites: why not in our use of documents? Microsoft Office formats such as Word and Excel have support to make this a simple reality (one of the reasons why XML and Web Services functionality is in the product). In all the debate about interoperability of word processing document formats, one major point seems to be overlooked: the whole debate always centres on presentation and ensuring consistency of mark-up such as bold, italics and so on. But the industry (or at least some of it) has long since moved on from there: today’s documents and office tools are about XML in terms of data interoperability and open access to information systems using Web services.

But for all the posturing and noise of the current debate, the one thing the proposed new “open” document formats neglects entirely is the type of smart data interoperability that the likes of Word provides. If all this debate is about is moving documents in and out of different word processing packages, we don’t seem to have advanced much in the last twenty years: I can recall doing that before, during and after the WordStar era.

January 6 2005

Security is clearly going to remain a hot topic – from items such as phishing, pharming, viruses and Trojans right the way through to the bigger security picture, including protection of critical national information infrastructure, electronic border control and national identity cards.

The statistics from uniras show a far more interesting picture about computer security than the mainstream media might lead you to believe. Last year a delegate at a security conference mentioned to me he thought Open Source had not only come out of the bazaar and become completely commercialised, but that it would overtake Windows in turns of security vulnerabilities. Looking at the uniras statistics, you can only conclude he was right. These official figures show that vulnerabilities and security related problems on Linux distributions and Open Source applications are running far higher than Microsoft Windows related problems. This underlines the point that security is an industry-wide problem we need to collectively address.

I still look forward to the day when we can move beyond this unnecessary binary debate about one software development versus another: software is a broad ecosystem and the consumer has choice.

January 3 2005

No New Year’s resolutions, but some predictions. This year, user identity in an online world and identity management in general will be increasingly centre-stage. Traditionally identity management (and closely related online uses, such as single sign-on) has been focused at best on the enterprise space, scaling to maybe several hundred thousand users. But in the government space, even the smallest countries are often looking at millions of users – and in the large countries, in excess of one billion. There’s been little debate between technologists, politicians, privacy activities and citizens about what sort of model is best suited to meeting both the concerns of the citizen with the needs of the state. The UK national identity card project does not yet seem to have provoked much in the way of a public, informed, inclusive public debate (although sites such as www.idealgovernment.com are doing their best to encourage us). 2005 looks likely to be a year when this will change – and the debate will enter the mainstream.

There are few IT projects anywhere in the world that have tried to tackle the scale of identity management that the national ID card project implies. The UK Government Gateway project, which will celebrate its fourth birthday later in January, has proven one of the few scalable, and robust, projects in this space. Part of the reason for its success is its federated structure and openly accessible interfaces – an architecture that appears to run counter to the implied highly centralised technical model that may arise around the National ID Card project. It will be interesting to see how this develops. Like many other successful government projects, the Government Gateway receives little media attention. Government IT projects that are failing apparently make for much better entertainment. And so a perception grows up that government does not do IT well. In truth, government generally does IT very well – and on a unique scale that many in the private sector would find hard to emulate. Even egg, which claims to be the most successful and largest online bank, only has 3.6m users: still small scale in a UK context. And smaller than the Government Gateway. And far, far smaller than many of the systems that DWP, IR and others use day in, day out.