Consensual technology

Internet technologies largely run on consensus. This makes the rate of significant change in Internet technologies actually quite low.

In order to make a major change to a technology you need the consensus of quite literally many thousands (or hundreds of thousands) of people.

Before you start thinking too hard about Web 2.0 – well, Web 2.0 isn’t actually a technological change. It was a philosophical change or usage change for technologies that already existed.

Microsoft, as an example, have been trying to produce technological advancements to email – the first significant changes to email protocols since 1982 – and so far lack the clout to accomplish it.

HTTP-ng, intended (nearly two decades ago) to make significant improvements to HTTP likewise couldn’t acquire the necessary buy-in to make it a reality.

IPv6 – now more critical to Internet infrastructure than ever, as the last of our IPv4 addresses are being allocated (as well as being a major performance improvement) – is likewise more than a decade old, without widespread adoption.

The problem – if it is a problem – is that our existing technologies simply just work “well enough”. They could work substantially better, if we were to make major changes to them, but we don’t. About once or twice each decade we make a minor tweak through standards-bodies, and any major changes are eliminated from the mix, because it’s much harder to get buy-in for a major change than it is for an entirely new technology (which isn’t easy in itself).

Much of it is a bootstrap issue. You might have a fantastic new email protocol that vastly improves on current systems and cures many of the ills and issues with the current system. It’s not much good if only sixty or so people actually use it, worldwide. Software makers probably don’t want to support two different systems at once, and with multiple software clients, users aren’t exactly going to buy-in when 99.999% of their email still comes in on one piece of software, and they need to run another one for the occasional message.

It’s like having incompatible telephone standards. Lack of interoperability is fatal unless you can capture a huge chunk of the market at the outset, and that means consensus for the change.

HTML5 might look like a significant step forward technologically. It’s not. Actually, it’s little more than a years-long standardisation process for technologies that have already been a part of your Web experience for quite some time – and the process of completing that standardisation isn’t even finished yet. In a sense it’s all about making some very small changes to existing technologies in order to achieve the sort of consensus needed to ultimately make it all work.

Almost every Internet technology we currently use dates back many years (and often decades), and knowing what we know now we could substantively improve those systems, in management, deployment, operations, performance and efficiency.

We don’t, though, because we just cannot get the widespread consensus to do more than inch along with micrometric changes to technology that barely works “well enough.”

* Footnote: Throughout, I’m referring to technological improvements that are no more costly than the technologies that they are replacing. Either representing the same (or lower) operational and implementation costs.

Tags: , , , , , , , , , , , , ,

Categories: Opinion, Technology.



Got a news tip or a press-release? Send it to [email protected].
Read previous post:
Close