Bulletin 28 December 2018. What ho 2019: predictions of a bygone age
The more that things change…
Why are bricks the size they are? Frankly, I don’t know, but I bet it is down to a number of factors which had to be weighed up one against another. Not least ease of laying, production quality and, probably, aesthetics. The result is a reasonably standard set of dimensions which can be seen the world over.
In a similar way, many of the things we do with technology work best when they are at a certain scale. Code modularity, for example, balances internal and external complexity, probability of dependency, functional specialisation and generalisation. The result is similar, whether you are working with old-school programming functions or all-new microservices.
The same notions apply for business processes, user stories, data models and, probably, any type of abstraction. We could call it Goldilocks’ First Law in that it feels just right, balancing the needs of cohesion and coupling. It can be equally straightforward to get wrong, as I have found out across various software audits and consulting assignments.
What has this got to do with predictions, I hear you say? It’s a good question, not least in that I didn’t mean this bulletin to head in the direction it has. The plan was to make some general point about right-ness, then use that to set a few expectations for the coming year.
Having derailed myself (prediction zero: this won’t be the last time), I could nonetheless continue regardless. But heck, in for a penny. Instead, here’s a few ancient truths dressed up as analytical wisdom:
**1. Something that wasn’t quite possible before becomes possible.**Possibly because it is cheaper to do, or processing/networking/storage passes a threshold. In this camp are various applications of AI, IoT and other data-intensive use cases.
2. Something gets standardised and everyone adopts it. Standards often happen because everything is using a certain thing anyway, or it passes an adoption threshold of some sort. Suddenly we go into this doublethink stage, where people act like they’d always been doing it. Cue: Kubernetes and microservices.
3. A consensus is reached that if something is going to happen, it needs to be managed. Generally this will be because that new new technology has proved itself in pilot studies but crashed and burned when adoption went enterprise-wide. From the failures fo early adopters come frameworks and tools that were always necessary, in (ahem) hindsight. P.S. Hello, Value Stream Management.
4. Something that everyone said was pointless becomes both successful and mundane. You know, that new kind of tech it seems that everyone is banging on about, yet most seem to say it’s over-hyped nonsense? Well, whaddyaknow, there was a use for it after all, but it’s not actually that interesting. Those who said it was cool are now, frustratedly, trying to say, “I told you so,” but, frankly, nobody cares. And Blockchain.
5. Something which looked like it has done its time proves to still be useful. While many applications end up in the elephants’ graveyard, the capabilities that they are built of tend to last a bit, or a lot, longer. Cue yet another article (I should know, I’ve written plenty) titled, “Tech XYZ is dead: long live tech XYZ,” which points out the flaws in suggesting the demise of, well, anything. So, no, serverless models will not wash away everything that has come before them, sorry.
There we have it: now we can all write a few predictions. Or alternatively, we can finish the roast, have a final, wistful look back at 2018 and wish each other all the best for the new year, whatever may come to pass. Merry Christmas, and may health, happiness and perhaps a little value-based positivity come your way.
This has been by 52nd newsletter, which means I lasted the year without running out of things to talk about. Hurrah, huzzah and thank you all for reading this far, and for this long.
All the best, Jon
P.S. A few more thoughts from me about the nature of prediction happen to be the next section of Smart Shift, so here it is. Not sure what happened to the formatting, but scroll down and it’s there.
