The Future of Software (Architecture)

If you hang out on the internet for long enough you begin to pick up memes about what the future (of virtually anything) might be like. Bearing in mind this cautionary quote from Jay Rosen:

Nothing is the future of anything. And so every one of your “No, X is not the future of Y” articles is as witless as the originating hype.

Here is a meme I am detecting on the future of software based on a number of factors each of which has two opposing forces, either one of which may win out.  Individually each one of these factors is not particularly new, however together they may well be part of a perfect storm that is about to mark a radical shift in our industry. First, here are the five factors.

Peer to peer rather than hierarchical.
Chris Ames over at 8bit recently posted this article on The Future of Software which really caught my attention. Once upon a time (i.e. prior to about 2006) software was essentially delivered as packages. You either took the whole enchilada or none of it at all. You bought, for example, the Microsoft Office suite and all it’s components and they all played together in the way Microsoft demanded. There was a hierarchy within that suite that allowed all the parts to work together and provided a common look and feel but it was essentially a monolithic package, you bought all the features even though you may only use 10% of them. As Ames points out however, the future is loosely coupled with specialty components (you can call them ‘apps’ if you wish) providing relatively simple functions (that is, do one thing but do it really, really well) that also provide well defined programming interfaces.

This is really peer-to-peer. Tasks are partitioned between the applications and no one application has any greater privilege than another. More significantly different providers can develop and distribute these applications so no one vendor has a dominance on the market. This also allows newer and smaller specialist companies to build and develop such components/applications.

A craft rather than an engineering discipline.
Since I started in this industry back in the early 80’s (and probably way before that) the argument has raged as to whether the ‘industry’ of software development is an engineering discipline or a craft (or even an art). When I was at university (late 70’s) software engineering did not exist as a field of study in any significant way and computer science was in its infancy. Whilst the industry has gone through several reincarnations over the years and multiple attempts at enforcing an engineering discipline the somewhat chaotic nature of the internet, where let’s face it a lot of software development is done, makes it feel more like a craft than ever before. This is fed by the fact that essentially anyone with a computer and a good idea (and a lot of stamina) can, it seems, set up shop and make, if not a fortune, at least a fairly decent living.

Free rather than paid for.
Probably the greatest threat any of us feels at present (at least those of us whose work in creating digitised information in one form or another) is the fact that it seems someone, somewhere is prepared to do what you do, if not for free, then at least for a very small amount of money. Open source software is a good example. At the point of procurement (PoP) it is ‘free’ (and yes I know total cost of ownership is another matter) something that many organisations are recognising and taking advantage of, including the UK government whose stated strategy is to use open source code by default. At the same time many new companies, funded by venture capitalists hungry to be finding the next Facebook or Twitter, are pouring, mainly dollars, into new ventures which, at least on the face of it are offering a plethora of new capabilities for nothing. Check out these guys if you want a summary of the services/capabilities out there and how to join them all together.

Distributed in the cloud rather than packaged as an application.
I wanted to try and not get into talking about technology here, especially the so called SMAC (Social, Mobile, Analytics and Cloud) paradigm. Cloud, both as a technology and a concept, is too important to ignore for these purposes however. In many ways software, at least software distribution, has come full circle with the advent of cloud. Once, all software was only available in a centrally managed data centre. The personal computer temporarily set it free but cloud has now bought software back under some form of central control, albeit in a much more greatly available form. Of course there are lots of questions around cloud computing still (Is it secure? What happens if the cloud provider goes out of business? What happens if my data gets into the wrong hands?). I think cloud is a technology that is here to stay and definitely a part of my meme.

Developed in a cooperative, agile way rather than a collaborative, process driven (waterfall) one.
Here’s a nice quote from the web anthropologist, futurist and author Stowe Boyd that perfectly captures this: 

“In the collaborative business, people affiliate with coworkers around shared business culture and an approved strategic plan to which they subordinate their personal aims. But in a cooperative business, people affiliate with coworkers around a shared business ethos, and each is pursuing their own personal aims to which they subordinate business strategy. So, cooperatives are first and foremost organized around cooperation as a set of principles that circumscribe the nature of loose connection, while collaboratives are organized around belonging to a collective, based on tight connection. Loose, laissez-faire rules like ‘First, do no harm’, ‘Do unto others’, and ‘Hear everyone’s opinion before binding commitments’ are the sort of rules (unsurprisingly) that define the ethos of cooperative work, and which come before the needs and ends of any specific project.”

Check out the Valve model as an example of a cooperative.

So, if you were thinking of starting (or re-starting) a career in software (engineering, development, architecture, etc) what does this meme mean to you? Here are a few thoughts:

  • Whilst we should not write off the large software product vendors and package suppliers there is no doubt they are going to be in for a rough ride over the next few years whilst they adjust their business models to take into account the pressures from open source and the distribution mechanism bought on by the cloud. If you want to be a part of solving those conundrums then spending time working for such companies will be an “interesting” place to be.
  • If you like the idea of cooperation and the concept of a shared business ethos, rather than collaboration, then searching out, or better still starting, a company that adheres to such a model might be the way to go. It will be interesting to see how well the concept of the cooperative scales though.
  • It seems as if we are finally nearing the nirvana of true componentisation (that is software as units of functionality with a well defined interface). The promised simplicity that can be offered through API’s as shown in the diagram above is pretty much with us (remember the components shown in that diagram are all from different vendors). Individuals as well as small-medium enterprises (SME’s) that can take advantage of this paradigm are set to benefit from a new component gold rush.
  • On the craft/art versus engineering discipline of software, the need for systems that are highly resilient and available 99.9999% of the time has never been greater as these systems run more and more of our lives. Traditionally these have been systems developed by core teams following well tried and tested engineering disciplines. However as the internet has enabled teams to become more distributed where such systems are developed becomes less of an issue and these two paradigms need not be mutually exclusive Open source software is clearly a good example of applications that are developed locally but to high degrees of ‘workmanship’. The problem with such software tends not to be in initial development but in ongoing maintenance and upgrades which is where larger companies can step in and provide that extra level of assurance.

If these five factors really are part of a meme (that is a cultural unit for carrying ideas) then I expect it to evolve and mutate. Please feel free to be involved in that evolutionary process.

One thought on “The Future of Software (Architecture)

  1. Now, that quote from Jay Rosen is funny.. Let's face it: Who can really predict what the future will bring? Big data may be close to finding an answer to that, but we're not there yet. I love the insights on this post though, so thanks!

Leave a Reply to Shaleen Shah Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s