Internet Explorer 8 - time to upgrade

Businesses should upgrade to Internet Explorer 8 to remove the standards compliance problems with old versions.

The saga of Internet Explorer (IE) and standards starts in the late 1990s, when web standards were not well established, and there were significant differences between browsers. At the time, Netscape Navigator had an 80% market share. To compete with Netscape, Microsoft promoted IE (some would say unfairly) by distributing it free with Windows. Microsoft also owned the popular web authoring tool FrontPage and worked with the then-dominant ISP AOL to ensure AOL users had a version of IE. Microsoft was very successful: by 2002 IE had around 96% of the market share. IE was so dominant, it could get away with not following official web standards closely.

From around 2004, there has been a growth in other browsers, such as Mozilla Firefox, Apple Safari, Opera, and more recently, Google Chrome. IE's market share has now declined to around 65%. Web standards have also advanced significantly. These other browsers follow official standards closely. They all display web pages similarly to each other, but differently from IE.

The differences between IE and other browsers are a big headache for web developers. They have to resort to tedious work-arounds to get websites to look the same in all browsers. Many problems — such as unreadable text or jumbled layouts — are caused by the need to write different code for IE and standards-compliant browsers.

Web standards support has been improving in IE, but there are still problems in IE7. However, IE8 was released in March and follows official standards closely. You can create websites that look the same in IE8 and in other browsers, without complexity and messy work-arounds. This simplifies web development and improves reliability and user experience.

Microsoft have handled backwards-compatibility well in IE8. IE8 can display websites designed for earlier versions of IE, either by setting options in the browser, or by using additional metadata added to the website.

Most home users of Windows will accept automatic updates and migrate quickly to IE8. IE8 usage has grown rapidly to around 15% as of the end of June, and will pass the usage of IE6 any day now.

Businesses tend to be more conservative. They assess change more carefully, and need to allocate resources to manage the upgrade. I guess that most IE6 usage is business-related, and it is businesses not home users that will cling to IE7.

The standards compliance advantages of IE8 are substantial. The case for upgrading to IE8 is stronger than the case for IE7. We should:

• Set a policy that encourages standards-compliant websites and browsers, and discourages features and bugs that are specific to earlier versions of IE.
• Consider the long-term advantages of IE8 to balance the cost and risk of upgrade.
• Explain the situation to our business colleagues, and encourage them to accept minor problems during the transition to standards-compliant websites and browsers.

If we all do this, we will help overcome one of IT's less glorious episodes, and reap the rewards of easier web development and a more reliable user experience.

© Copyright 2009 Minimal IT Ltd. See the Minimal IT website for the original newsletter and copyright information.

Opera Unite - Dawn of a new era?

Although still in its infancy, Opera Unite could start a big shift in how we use IT.

Some years ago, I wrote about a second PC revolution in which PC users create outward-facing services and offer them to the world as easily as they currently create spreadsheets.

That vision has come one step closer with Opera Unite.

Opera Unite, currently in beta testing, is an add-in to version 10 of the Opera browser. It allows you to serve web-based applications from your PC to anyone on the Internet, or your colleague at the next desk, whatever browser they run. It currently supports only a handful of services such as a web server, file sharing, media player and chat. New services can be written in JavaScript. Opera hosts a site for users to upload and share services.

Opera handles the networking. Instead of other people accessing your PC directly, they access a public URL (something like http://computer.yourname.operaunite.com/web_server), which the Opera servers then forward to your PC. You can use the service to run web-based applications without configuring servers, domain names and firewalls, even from behind a corporate firewall.

From our technical perspective, it's easy to dismiss Opera Unite. It isn't hard to set up a web server on a PC, or to use an addressing service such as DynDNS.com to provide an address to it. And of course running web applications from inside a browser on a PC is architecturally inept.

But we dismiss this new direction at our peril. There are many reasons why Opera Unite is significant:

• Don't underestimate the barriers to setting up outward-facing services: technical knowledge, permissions, fear of looking silly, the IT department's work request processes. Opera Unite does away with all of this, and could unleash a flood of suppressed demand for IT.
• People like having their own stuff. There are lots of social websites and hosted business services where people can share things, but there is something deeply ingrained about having things under our complete control. It is one of the reasons PCs are so popular.
• The installable service model could grow hugely. There has a been a lot of interest in Apple iPhone applications and in Facebook applications. There could soon be a huge library of Opera Unite applications ready for anyone to run and offer to the world. (A good starting point would be services which allow collaboration on office documents.)
• The networking is significant. Anyone with Opera Unite can set up a file share or website to colleagues and contacts within or beyond the organisation. It will be easier to use Opera Unite than to follow the official procedures defined by the IT department.

If Opera Unite is successful, and if similar features are built in to more popular browsers, we in IT risk losing control as users abandon centrally controlled IT and take matters into their own hands. We can see this as a threat, or see this as an opportunity, but I do not think we should ignore it.

© Copyright 2009 Minimal IT Ltd. See the Minimal IT website for the original newsletter and copyright information.

What's really going on 4: putting it all together

The combination of shared understanding of IT, long-term proactive management and evaluative theory is a powerful tool for addressing the problems and opportunities of IT.

Over the past three weeks I have written about how management of people, process and technology is not sufficient, and many IT problems and opportunities are not addressed.

I have suggested that structuring management around whole systems provides a useful common viewpoint. The remaining parts of this mental jigsaw are long-term proactive management and evaluative theory.

We know that it is cheaper and easier to be proactive: to prevent problems rather than fix them, and to grasp opportunities for improvement in a timely fashion.

To give some IT examples, it is cheaper to manage IT systems so that they are useful for longer, rather than to let systems decline quickly and replace them frequently. It is easy to keep systems well-documented, testable and well-maintained, but very hard to recover them once they start to decline. It is easy to plug occasional vulnerabilities in otherwise secure systems, but difficult to implement comprehensive security on insecure systems.

In IT, we find long-term proactive management very hard. It is hard to make sense of everything that is going on across all the systems, and to identify, prioritise and justify the interventions that are needed. Proactive management is at best piecemeal. Inevitably we do not address all of the issues, and costs and risks increase dramatically.

This is where evaluative theory really helps.

You can base management on one of two types of theory. You can use an explanatory theory which describes how the people, processes and technology must be arranged, and then execute on that management. This is effective for common tasks such as project delivery and day-to-day operations. However, this approach requires both management effort to define and ongoing resourcing to execute. It is not suitable for the occasional, discretionary, responsive activities that proactive management demands.

An evaluative theory does not explain how things will be done, but describes what a good thing looks like. It is much easier to define an evaluative theory, and you do not need to do very much in order to manage. Assessment against an evaluative theory quickly identifies and prioritises what needs to be done. Because it looks at outcomes rather than the detail of implementation, an evaluative theory is much easier for a non-specialist to understand.

The combination of evaluative theory and a system view is very powerful. Instead of IT being a fragmented and mysterious complex of people, processes and technology, IT becomes a set systems, each with a set of understandable outcome measures or "qualities" (which is why we call this approach system quality management). This gives insight to make effective, justifiable recommendations for proactive management that can make a significant difference to IT.

What's really going on in IT? We don't have the management disciplines that we need to manage IT effectively for the long term. Our management of people, processes and technology is fragmented and inward-looking. We can fix this. Combining a system view (that everyone can understand) with evaluation (that identifies what needs to be done) gives the clarity and justification that we need to manage IT effectively for the long term.

© Copyright 2009 Minimal IT Ltd. See the Minimal IT website for the original newsletter and copyright information.

What's really going on 3: thinking of IT differently

Despite the effort put into managing people, processes and technology, IT management is chronically fragmented and inward-looking. We need to think of IT differently.

We need to understand clearly the purpose of IT and how IT delivers value.

The purpose of IT is to provide useful information to the business. IT delivers value because it allows this information to be preserved through time (storage), derived from other information (transformation) or moved from place to place (communication). Higher-order value can be built on top of this, such as decision support, process automation, or enabling business change, but these can all be traced back to these basic capabilities.

We also need a simple structure for thinking about IT that counterbalances the complexities of people, process and technology. Our requirements for this are:

• Easy to understand.
• Shows the connection between what we do and the value that IT delivers.
• Provides a common point of reference between IT and business.
• Provides a common point of reference between different IT specialists.

Value is delivered from IT by the automation of the storage, transformation and communication of information. To get the structure we need, we can split IT down into understandable and manageable units around this information automation capability.

That is what a business application is - a related set of IT capability. We can use this as the basic granularity for our understandable IT structure.

But we need to think much further than just business applications. When we talk of applications, we think of software, as distinct from hardware or infrastructure or support processes. We want to bring all the people, processes and technology of IT into our definition.

For this reason, use the term "system". A system is an interrelated set of people, processes and technology that does something useful for the business. The granularity we need is the same as business applications. But we need a much broader scope, and our definition of system include the business usage of the application, the software and hardware on which it runs, the service and support processes around it, and all the people needed to deliver the IT capability.

Using this definition, "system" meets our requirements. It is understandable and recognisable by both business and IT. It is connected to value because we can consider the value derived from the information that it stores, processes and communicates. We can understand all the people, processes and technology that go to make up the system.

Next week I will cover how we can use this way of thinking to overcome the problems caused by fragmented and inward-looking IT management.

© Copyright 2009 Minimal IT Ltd. See the Minimal IT website for the original newsletter and copyright information.

What's really going on 2: why we do not succeed

Consistent, long-term success in IT is impossible because IT management is fragmented and inward-looking.

In IT, we put a lot of effort into managing people, process and technology, but we still have lots of problems and there is huge room for improvement. Why is this?

The problem has its roots in the complexity of IT management. The intricacies and connections between the different groups of people, different processes and different technologies have become so complicated that we no longer understand them fully. We can not see clearly the connection between what we do and business value. Nobody outside IT understands IT, and few of us within IT understand more than our own little corner.

The way we talk demonstrates this. If you ask nearly any IT manager what they do, or what's important, or where they feel the organisation should go in the future, this is expressed in terms of their view of the people, processes and technology of IT. We say that delivering IT projects on time and to budget is vital. We say that enterprise architecture is important. We want shared infrastructure services. We want reusable components. But all these are just parts of how we do IT, and are not the purpose of IT. We have forgotten, if we ever knew, what the purpose of IT is.

This fragmented, internal focus is at the heart of IT's problems.

We find joined-up management impossible. If, for example, we want to reduce cost, we use simplistic management approaches, such as arbitrarily reducing staff or delaying upgrades, which might well increase cost. We can not control IT in any meaningful way. We can not make a true business case for maintaining IT, so systems inevitably slide into unsupportable legacy. We are not confident that we can work with the IT that we have, and so we gamble fortunes on massive, risky, and often unsuccessful replacement projects.

IT is not the only complicated profession. The medical profession has significantly more sub-professions, more complicated procedures, and uses a lot of technology. Why is the medical profession so much better at keeping people going than we are at keeping IT going?

The underlying reason is that the medical profession has a simple and obvious focus for their work: patients. This provides a common basis for measuring value, a common point of reference for all medical professionals, a common focus for all processes, a reason for all use of technology. A cardiologist and oncologist can both discuss the same patient, and have meaningful discussions about the cost of patient care with a hospital manager.

In IT, we do not have such a clear focus for value delivery, around which we can structure our people, processes and technology. We follow, and try to improve, our management of people, process and technology, and hope that by doing so diligently IT will somehow work. But without a consistent value-related focus this inevitably leads to fragmented and inward-looking management that can not succeed.

Next week I will cover how we can succeed if we think of IT differently.

© Copyright 2009 Minimal IT Ltd. See the Minimal IT website for the original newsletter and copyright information.