The software wizard legacy lives on

Desktop wizards reimagined for the cloud data center era.
27 May 2019

25 years later, phishing attacks remains an effective cybercrime technique. Source: Unsplash

Back in the 1990s, we were all still getting used to Windows 3.1x and all the desktop-based efficiencies that would come with it.

As Windows developed throughout the end of the last millennium, Microsoft introduced so-called ‘wizards’ to help guide us through various installation, configuration and setup procedures.

Rather than asking a user to script their way through hard code on the ‘command line’, it was obviously easier to guide a user through various, almost flow-chart-like, options to change the way a particular machine worked.

Although wizards are a mostly forgotten piece of computing history (along with the Clippy office assistant, thankfully), their legacy remains both in the desktop applications we all still use… and in the core building block systems that programmers and developers interact with every day in order to create the networks that we now depend upon.

Many of the systems we interact with today do still have a good degree of wizard-like wizardry about them, but this is now brought about through what— in more contemporary terms— are known as templates and prefabricated controls.

Wizard legacy

Part of the reason for needing templates is speed. In the world of always-on cloud computing, we need to be able to change applications at lightning speed… and this has given rise to the terms Continuous Integration (CI) and Continuous Delivery (CD).

Programmers rely upon CI and CD techniques in order to be able to get ‘live’ applications like social media platforms or online apps to update quickly.

Sometimes now called Pipeline Creation Wizards, these are software tools that don’t do everything for the programmer, but they do shoulder the heavy lifting tasks.

Where the shape of a dataflow, or a user interface type, or a database update (and so on) can be largely classified into standardized shapes, then we can let automation technology handle that part as we just focus on the nuances and tweaks and anomalies that will always be part of the job.

From grunt-work to tweaks & nuance

That same leap from grunt-work to tweaks and nuances is being delivered to business users. Major database companies are attempting to elevate the functions of their Enterprise Resource Planning (ERP) systems and expose more non-technical staff to their mechanics.

Throughout the birth of cloud computing, we have assumed that the only people getting their hands dirty with managing data warehousing tasks are the administrators and systems people. But massive chunky data warehouses are now segmenting parts of their information out into what we call ‘data marts’.

For the record, data marts are smaller and more specialized systems of data, which sit as decentralized units to serve specific business units (finance, CRM functions, specific sales projects, or more general Human Resources functions, etc.) in any given organization.

If we take the grunt-work out of data warehousing, then we give business users smaller, specialized sets of information and present them with productized workshop and prototyping tools… then we’re basically creating democratized access to deep-level enterprise data, which can lead to faster business decisions being made… in theory at least.

It is, if you will, desktop wizards reimagined for the cloud data center era.

But even with templates and automated interfaces, data warehousing still sounds scary… so you can expect vendors to use warm fluffy terms like business warehousing, insight engines, market predictor analysis tools and more besides.

This trend is happening and we’re continuing to see productized template controls appearing at all levels of the software stack from the backend all the way up to the user interface. We don’t even need to mention the fact that artificial intelligence (AI) and machine learning (ML) is helping to drive this effort, both will obviously allow us to track, monitor and service users’ needs in more intelligently-template-based systems.

I don’t want your freedom

So is there a downside?

Some industry commentators will always argue that with abstracted automation there is always a loss of freedom and control. But if the use case has been proven enough times, this argument is tough to uphold.

Wizards (software-based ones) are still with us and the influence of AI is going to drive even more automation into the software we use every day. Do you believe in magic?