Time to study the classics: Vintage tech is the future of enterprise IT

Look back in wonder

Opinion Business IT is driven by the need for the new. Not necessarily your business's need, but certainly that of vendors and service providers desperate for new revenue, the dismissal of the old once it's done its real job, and the inevitable prying open of the corporate checkbook.

If it's old, it's obsolete; and if it's obsolete, it needs to die. A remarkable number of those corpses are coming back to life, though, and they have something interesting to say.

There's no very good definition of what makes something officially ancient. Microsoft seems to think it's somewhere around a decade, as it just quietly euthanized 2009's Windows 7 and 2012's Windows 8. Hard to shed many tears here, yet other, far older technologies are driving more than just nostalgia.

The vintage tech itself isn't the point. For sure, there's enough of it about already in the enterprise, slogging away at a task that's too fragile or expensive to update. Type approval, industrial process integration or deceased suppliers can all leave orphaned islands of ancient hardware and fossil software better suited for a museum than a day job. Sometimes a particular piece of tech does a particular job so well that there's no point in adopting later, less suitable developments, but they're still not where you'd go if you were starting from scratch today.

There's another side to vintage tech, though, and it's one of surprising innovation. Recapping a Commodore 64 is a pastime, but reverse engineering a proprietary signal processing board to revive an entire generation of a failed '90s audio format is proper long-trousered engineering. Or take Haiku, which has just seen a significant public beta release – a recreation on modern hardware platforms with modern software tools of BeOS, a quirky, doomed but technically spectacular OS from thirty years ago.

You can find teams reviving and renewing mainframes, minicomputers, software environments, and applications, from the most historically significant to the most obscure. Documentation may be scarce, if it exists at all, with proprietary code bases long since lost, and with long-obsolete components in garishly baroque combinations. Speaking of which, this description is also apt for more than a few legacy enterprise infrastructures still going today.

What's old is new again

Those skills involved in understanding, debugging and renewing vintage tech in the face of complexity and obscurity are entirely applicable to the challenges in dealing with contemporary multi-vendor installations running software stacks with heterogeneous components from many sources, of varying levels of maturity and with varying levels of support and documentation.

Perhaps the most important aspect of familiarity with digital days of yore also stems from that heterogeneity, and it carries forward as you progress up the job stack towards system design and specification.

The biggest risk to enterprise IT is failure in security, with ransomware and data breaches making the headlines every other day. It is such a major structural weakness within the industry that we are almost blind to vendors' chronic inability to live up to their promises.

That's not going to change; it can't be fixed. But it can be managed.

What turns weakness into catastrophe is monoculture. All software has flaws, all systems have weaknesses, and you design around that by having multiple different systems with different weaknesses. If the world's banana crop is a single strain, then a single pathogen can take out the lot.

When everyone relies on the same systems, then attackers will see a huge reward by concentrating on those systems. They work to the same risk/reward equations we all do.

You're not going to harden your IT infrastructure by rewriting Active Directory in FORTH under VMS, much as such an exercise may appeal to the more perversely creative. But the more experience you have of seeing things done differently, of analyzing and understanding systems that don't make the current industry standard assumptions, the better you'll be at assessing and working with different contemporary options. And the past in tech has oh so many failures to learn from too: if BeOS was so cool, why did it fail?

It's the same mindset that provoked the invention of the Raspberry Pi, itself a call-back to the vintage tech of the BBC Micro, that exposure to a rawer, stripped-down and non-standard way of doing things equips you with many more options for problem solving in digital systems.

Vintage technology was once new, and vintage business tech was designed to solve problems by extremely clever people operating under much tighter constraints than today. It contains perspectives and ideas that challenge modern assumptions while explaining where they came from. It is as close as our young field gets to an education in the classics, with many more flashing lights. Any enterprise worth its salt should encourage such knowledge, especially among its newer workers. Innovation is vital: we have 50 years of it to learn from. ®