Reunification

By Andrew Clifford

For such a young subject, IT is disproportionately haunted by its history.

The first stored program computers came out in the late 1940s. They were very primitive by modern standards but they had one important characteristic that has been lost from later computers. Although the inputs, outputs and programs that they ran were near-unintelligible machine code, they were all in the same unintelligible machine code. They were saved to paper tape in the same way; they shared the same tools and skills.

Since those early days, computers have grown hugely more capable. But as each new capability has come along, from symbolic assemblers onwards, IT has grown more fragmented. In the effort to make things more efficient and easier to use we have divided IT into more and more parts.

In a modern IT architecture, you're likely to have a database management system, file servers, web servers, a content management system, application execution environments, browser-based scripts, and so on. And each has its own concepts and representations: SQL for the database, HTML, CSS and JavaScript for the web pages, and core code written in one or many high-level languages.

Each part has a different set of concepts, a different set of tools, and a different skill set. And as well as mastering separate data, web and programming constructs, you need to integrate them into a meaningful system.

When looked at individually, each new component looks like a good idea, but we have ended up with a much more complicated environment. We are haunted by the legacy of progress, and have not gone back to consider if it can be simplified. I am not suggesting we should go back to machine code and paper tape, but we should use our ingenuity to reduce the number of concepts, technologies and representations. Computers are simply machines that automate the storage, movement and transformation of information; it should not be so hard.

Unification has been one of the driving forces of the Metrici platform. Within Metrici there is no distinction between data, application, content or anything much else. It is all just information to be stored, moved or transformed.

To see what I mean, set up a simple web page in Metrici. Look at getting started for instructions, but in summary:

• Register at www.metrici.com.
• When your account has been created, use the New button to create a new Page.
• Give your page a name and some content, and click Save.
• Click on the page title to view your page.

What have you just done? Obviously you have created some web content – a web page with its own address. But the text you entered is stored in a database. And the actions you have used – New, Save, etc – are the actions of an application.

Within Metrici, everything is the same. The code that creates your account, the record of your account, your account home page, the New dialog, the web page you create, are all represented in the same way and are all used in the same way. It is not an integrated environment that combines multiple tools, it is a fundamentally unified environment, making it hugely simpler.

Next week I will describe the concepts behind Metrici's unified approach.

Subscription