Last week, as is my custom, I supplied you with the top 10 things you don’t need to worry about in the coming year. I hope they made you happy.
For all happiness is worth -- according to a recent study published in The Lancet, that sunny state of mind is overrated. Being happier or even less stressed than average seems to have no effect on longevity.
That's why I have a clear conscience as I drop my annual “other shoe” -- the top 10 things you should worry about in the year to come:
1. Your company's valuation
Do you work at a hot startup? Well, no matter what they told you, very, very few tech companies are worth $1 billion. Being the “next Google” (or Amazon or Facebook or whatever) is a nice pipedream, but that's all it is.
A whole lot of VCs didn’t spread their bets widely enough and piled on too few companies in too few of sectors, so we’ll see more re-valuations in 2016. Some people will get hurt -- but did VCs really think they could breed unicorns by feeding them $100 million? Where would sales revenue come from and in what time period? As for startups targeting “enterprise customers,” enterprises take their time making technology buying decisions. In a competitive marketplace, there's no way your average NoSQL database company is worth $1 billion.
2. Gig economy regulations
As I’ve said before, I have deep misgivings about Uber management and its creep factor, though the average Uber car smells less like vomit and cigars than the typical taxi and always has GPS (at least in the United States). Meanwhile, I had my local cash-only joint’s Vietnamese homestyle special delivered to my office and paid with a credit card using Postmates, and it was everything I imagined it could be.
Unfortunately, regulators don't share my enthusiasm and are looking closely at services like these. Don’t get me wrong, in a wealthy country I think we should all make a living wage -- but quashing the gig economy isn't going make anyone’s life better.
3. The Internet
The cable/telco oligopoly is still pushing its agenda, and I see no reason why President Trump won’t overturn the common carrier FCC ruling. The MPA, RIAA, and BSA are all getting their happy thoughts shoved into trade agreements. The 1 percent has never been comfortable with an open and competitive field or fair use, and they won’t stop in 2016, even though 2015 brought them a series of setbacks. Heck, even the "Happy Birthday" copyright was overturned!
But remember, their side has more lobbyists and lawyers and enough interest to sustain this; when you aren’t looking they’ll pass this stuff, if not directly, then through secret trade agreements like the TPP that get fast-tracked through Congress.
4. Yet another huge cracking scandal
I see no indication that companies are starting to take security seriously instead of using it as a reason to feed their IT bureaucracy. It doesn’t matter that anomaly detection is easily available for everyone. Nor is there reason to believe the U.S. government will stop actively undercutting the security of U.S. firms with their brain-dead war-on-math encryption policies because “terrorism” (or something). Nope, whether it's a “state-sponsored actor” or script kiddies or corporate espionage, we're all p0wn3d.
5. Your job maintaining large server farms at a small company
Almost everyone agrees that the cloud is the destination. If you run a donut joint, why do you have a server? If you run a donut chain, why do you have a server farm? Let someone who runs server farms do that for you.
What people disagree on is how fast this transition will happen, but I'm here to tell you the days of lovingly curating and cabling your own personal server farm are numbered. Retool yourself or redeploy.
6. An eight-week course to become a coder
At one point I thought these were a good idea, but not anymore. You pay a ton of money to take a six- to eight-week course in Ruby or Python or Web design. Next, after you graduate, they take 20 percent of your first-year salary.
Look, it takes a while to get good at anything, and until then you’re not that valuable. I don’t imagine I’d be fluent in Spanish after a six- to eight-week course, so why would you think you’d be fluent in software development? A coding "bootcamp" can get you started, but you’re not actually employable after six to eight weeks. They aren’t worth $6,000 to $30,000 in tuition. Get real.
7. Oracle
In the run-up to the Crimean war, Tsar Nicholas called the Ottoman Empire the "sick man of Europe." But with a little help it managed to beat Russia and hang on long enough to suffer defeat in World War I. Likewise, Oracle ended a triumphant decade with the capture of Sun Microsystems and then ... well ... what the hell has Oracle been doing the last few years? Sun was a difficult beast to gobble, and integrating it hasn’t gone smoothly.
Oracle has missed its earnings goal for the last dozen quarters or so. Meanwhile, it still holds valuable properties like Java and has a war chest to be reckoned with, although it seems to use the latter mainly for “buy and bilk” (buy something you depend on, raise the price). Moreover, it has half the industry by the short and curlies. You’re not going to migrate your mountain of PL/SQL code and big, fat Oracle RDBMS any time soon unless you have to -- so Oracle retains pricing power.
What about new licenses? A few years ago, if you started a company, you assumed you’d build out a big Oracle RDBMS to start. How many people assume that today? If you think NoSQL is a fad, then Microsoft SQL Server is a fine choice and there are plenty of decent open source RDBMS alternatives besides Oracle's MySQL.
Where will Oracle’s growth come from? Lawsuits against people who implement Java APIs is one option, I suppose. It could be bad for business in the long run; as a developer I now despise Oracle because it convinced the court to make interoperability illegal. (Look what a sue-everyone strategy did for SCO.) Of course, Oracle also has a cool way of raising its price on legacy -- see what that did for Novell. Oracle may be the sick man of the tech industry, but we're all potential collateral damage.
8. Hadoop/big data stack differentiation
MapR has Drill. Cloudera has Impala. Everyone now realizes that they love Spark! Meanwhile Cloudera has Sentry, and Hortonworks has Ranger. Cloudera also has the proprietary Navigator, and Hortonworks has Atlas (though very little code behind it as yet). There are real reasons for some of this to exist, such as Impala for MPP or Spark for in-memory analytics and streaming. Yet the vendors are making it hard for you to wrap your head around their offerings. Plus, Cloudera shot the one thing everyone had in common right in the head: HDFS may now on occasion give way to Kudu.
What's the meta-brand for all this stuff if it isn’t Hadoop? What if you want to build an organization-wide analytics core and get one support contract for the whole thing? The biggest companies are executing on “multivendor strategies,” but that's a rich man’s game, like shooting endangered species. Differentiation and rebranding away from “Hadoop” makes people start to wonder exactly what we're talking about.
9. All that JavaScript junk
Watching JavaScript evolve makes me want to kick kids off my lawn. Go to any meetup where they unveil the new thing in the new framework and you’ll be like, “Holy crap, every other language or framework had that over a decade ago.” But the wide-eyed millennials think it's all new. Oh well, Java and C# and so on had their “isn’t this shiny” phase, too, where we reinvented the wheel and called ourselves genius while the uppers stroked their long white beards wondering why we were so proud of ourselves.
Just like over a decade ago, when 1,000 new MVC frameworks arrived every day, now everyone is writing their own JavaScript framework. This time it's reactive -- because concurrent event-driven programming with observers is totally new and no one ever thought of that before. In 2016, we’ll see yet more JavaScript frameworks, they’ll all be “reactive,” and at least one of them will be pretty good.
10. A thousand machine learning frameworks and nothing to do
While the young folks write JavaScript frameworks, more seasoned developers are writing new machine learning frameworks -- and not explaining to the business how to use the algorithms they have. In the past quarter or so, Google, Microsoft, and IBM have all blessed us with their own vanity frameworks. I have 10 equally sized screwdrivers in my toolbox, so maybe I need a bunch of duplicate implementations of k-means clustering (the example everyone gives when they want to pretend they’re a data scientist).
Expect every major tech company to have at least one vanity library and don’t expect very much in the way of explanation about how/when should you use it (which is exactly why doing so is in my marketing plan for 2016).
How'd I do last year?
Unlike other prognosticators, I keep myself honest. I told you to worry about 10 things a year ago. Was all that stress unnecessary? Or did I nail it? Here's my self-assessment, completely free of any conflict of interest:
Your second-tier PaaS provider going away: I wrote that a couple of months after Cloudbees exited the PaaS business and became a Jenkins business. I don't know who uses a second-tier PaaS like Engine Yard, say, or AppFog (now a property of CenturyLink), but no one else seems to have fallen off the edge. On the other hand, this week’s news is all about how Amazon, Google, and Microsoft are going to rule the cloud, including PaaS, so I’ll call this one a win, because I'm doing the judging.
Overfunded startups: I should get an "A" this year for this one alone. Several of the leading NoSQL players suffered write-downs in their valuations, for example. As noted above, we’re not done here.
IBM's cloud plans: I nailed it when I predicted IBM would get its cloud act together this year. Bluemix might not yet be a household name, but it seems to be on the rise, and IBM is actively marketing it now. The company no longer requires you to fax things as part of the sales process like when I first talked about its cloud plans.
Microsoft Surface: The Surface wasn't the disappointment I thought it would be. I was right at the beginning of the year, but toward the end, press reports said the Surface and Surface Pro were outselling the iPad.
Another huge cracking scandal: How's that Ashley Madison account treating you? This week we’re hearing reports of leaked Amazon accounts, and earlier in the month VTech leaked your kids profile. Whether you’re looking for love, a good time, a flatscreen, or a something for the kids, companies are still hiring the cheapest developers in the world to write their software and hoping a downstream bureaucracy and network security will magically protect them. Until this changes, expect more of the same ... this year, too.
The developer shortage: We’re still struggling through the developer shortage, although the squeeze seems less desperate this year than last. I expect this to continue to be a challenge and a strain on the industry.
Your Red Hat stock: If you ignored me about your Red Hat (RHT) stock last year, then you’re 10 points richer for it. We’ll call this a miss. But the company will need to extend its vision if it wishes to continue to compete. No one cares about JavaEE anymore, and AWS is my operating system.
The Internet and the Fourth Amendment: Clearly, the Fourth Amendment is still dead and the Internet is still under attack by companies looking to fence off a franchise.
Leap seconds: Certainly there were bugs, but I don’t know of anything that came crashing to the ground due to leap seconds. We’ll call that a miss.
Your job as a developer: I haven’t seen a huge pullback in bullshit job titles, so I guess your job as a Visioneering Imagineer at some “Uber of ____” is safe for now. Maybe as valuations are refigured this will change. Meanwhile, we’ll call this a miss.
Summing up
I wonder, if I buy a Surface, will it run Ubuntu?
Despite clear misses, I nailed valuations and cracking so wonderfully, I’ll give myself a passing grade on my prognostications.
For this year watch as billion-dollar valuations continue to shrink to hundreds if not tens of millions as VCs gain a more realistic insight into where the industry transformation is going to happen (hint: not storage). Get ready to pay more for delivery as the regulators get their pens out. Love your throttled surveillance state with no fair use and tightly controlled creative expression.
When you buy your wife flowers (this is not sexist) to apologize for your transgressions, remember that your florist account will probably be cracked, too. How are you going to pay for those flowers anyhow, since you’ve been lovingly installing Windows on servers by hand (meat devops) at a small firm since the '90s and your CEO is talking cloud?
You thought about signing up for an eight-week developer bootcamp, but you remembered that your fitness membership didn’t turn you into Taye Diggs -- it only lightened you financially. You have a new Oracle sales team, and they want to talk about “compliance.” Meanwhile, the boss wants you to pick a Hadoop vendor, but your dev team deployed Drill, Impala, and some Ranger thing for security, so no single vendor will support any of it. Your customer support site is down due to a JavaScript bug incurred by some kid writing half the site in Angular 2 and the rest in Ember.js before giving up and creating their own JavaScript framework. Your boss wants to do predictive analytics to determine when there will be outages and he’s pretty sure you can deploy k-means clustering because a data scientist he met on his last plane ride said so.
That is your meaningless life. Then you will die. Enjoy 2016!