Life is tough, then you die, so please enjoy Arby’s. In the meantime, why worry?
Kidding! We all know there’s plenty to worry about. So as my annual public service, I offer you you 10 things you definitely don’t need to worry about in 2016. This year, as a bonus, I’ve tacked on a report card assessing last year’s predictions.
1. The Dell-EMC merger
Big mergers take time. However, this is a merger of the walking dead in the era of containers, cloud, and cheap SSDs. The next generation of large data centers uses something that looks more like local storage and distributed file systems. If this trend catches on in corporate environments, EMC’s main SAN business is in trouble. With increased reliance on cloud and centralization, there’s less reason to fill stacks of servers in little corporate data centers around the country.
Also, have you ever been on the phone with EMC while three or four salespeople from different divisions or subsidiaries pursue different goals at once? Now another voice will join the party!
2. Apple doing anything important
I don’t see Apple Cars coming out in 2016. While Tim Cook found time to block me on Twitter for retweeting a report about Apple’s labor policies, he hasn’t had the time to drive much innovation. The Apple Watch sold OK, I guess. People who buy at least one of whatever Apple sells bought an Apple Watch (I know Galen has one), but it hasn’t fundamentally changed the market like the iPad or iPhone. Mainly, it inspired Fitbit and others to add more mobile integration to their product lines. But it's a watch -- how much can you do with a tiny screen on your wrist?
The next game-changer will be some kind of VR device that doesn’t make you look like a Glasshole, a self-driving electric car, or a drone delivery service that lets you try on clothes and return them if they don’t fit. Apple won't bring you any of that in 2016. Instead, expect more stories saying Apple under Cook is a snoozefest.
3. Apache Spark losing steam
Is there anything hotter than Apache Spark right now? It's lightning fast compared to Hadoop. One of my company’s projects recently took a customer batch process from six hours to less than two minutes. If the data source would have allowed it, the client could have streamed and gone instantaneous. Spark is that fast. It's still maturing, but with full backing from Cloudera and IBM, as well as Hortonworks getting onboard, you can bet it will grow up quickly. Being able to distribute a collection across the cluster with a single
sc.parallelize() is a fundamentally good idea.
4. Oracle’s cloud
This one makes its second annual appearance, thanks to the company's new cloud announcements. Ultimately, the question is why? Why would you use an Oracle cloud offering? It isn’t like Oracle is nice to its customers, has great service, offers the best technology, or prices aggressively. Nor is there an ecosystem to pull you in (as Salesforce or Microsoft have). If you’re willing to consider a cloud, then you’re willing to consider an IT migration project -- and if you’re willing to do that, then why not cut your losses and move to newer technology, better pricing, and a better solution? For Oracle to have a play here, it will have to do something it hasn’t done in years: Develop an actual product and a compelling reason to buy it.
5. Regaining your right to privacy
The Kabuki theater where the NSA pretends to stop spying on us (when it's actually redefining common English words so that they can continue to spy on us) continues. The spooks will keep trying to weaken encryption so that it’s easily broken by terrorists, spies, and law enforcement alike. The avalanche has already started, and it’s too late for the pebbles to vote. The deep state runs thicker than blood or democracy, so I expect no change here. Do you?
6. MongoDB losing its NoSQL dominance
Hopping aboard the Mongo train has always been easy for developers. The difficulty has always been the locking model and the operational tools around it. In the past year, the company fixed the locking model with WiredTiger; if it improves the operational model beyond the current vaporous offering, it’ll really be a force to reckon with. While there are great differentiated and overlapping uses for MongoDB's nearest competitor, Couchbase, MongoDB still commands most of the market and remains the first database that comes to mind with NoSQL. Yes, some of the buzz has worn off, but each year has brought greater maturity and greater adoption in the enterprise.
7. AWS slipping from No. 1
"Cloud" generally means Amazon Web Services. In companies that decided to "go cloud," nobody ever got fired for choosing AWS. Google and Microsoft have great offerings, but they can't match the AWS ecosystem. AWS is becoming like a public utility. Talk about barriers to entry: Who is going to build as many data centers distributed around the globe? Real dominance isn’t being the top pick among many -- real dominance is when the people assume by default. Amazon holds a position similar to that of Microsoft Windows from the mid-1990s to mid-2000s. Many third parties may support other clouds, but they damn sure are going to support AWS, which happens to be my operating system of choice.
8. Docker fading away
Containers (aka Solaris Zones) are a fundamentally good idea -- and a far, far better idea than emulator-style virtualization ever was. Combine containers with an inheritable package scheme, and gosh that's nice. But now we must address the "s" word (security) and the "m" word (manageability). I expect big developments here in 2016. Success is Docker’s to lose.
9. Oracle’s big data/NoSQL plan
There's no compelling reason to buy Oracle's weird, hardly used Hadoop distribution and key-value store, especially when you can purchase an equivalent technology for pennies on Oracle's dollar. Sure, Oracle's stuff will “sell well” in bundles -- where Oracle sets a price and you agree that the thing you wanted is $X and you'll take five other things you have no intention of using in exchange for that price. I suppose “because we got it for ‘free’ with our site license” is a reason to go with Oracle, but when you find out it was “free as in kittens” you'll pay less in care and feeding alone by going with a market leader like Cloudera, Hortonworks, or MapR.
10. Your job as a developer
If recent history is any guide we should get at least one or two more solid years out of the current boom before the millennials learn what happens to freelancers and job-hoppers in a down economy, when everyone stops jumping around shouting “Developers! developers! developers!” and recruiters stop chasing you and start “keeping you on file” (file 13, that is). I expect 2016 to be another good year for recruiter spam.
How’d I do last year?
Aside from Oracle’s cloud, Oracle's big data/NoSQL play, and developer jobs (all of which I obviously did pretty well on or they wouldn’t be making a second appearance this year), how’d I do last time?
Pivotal HD/Greenplum: I totally missed that one. [You should have listened to me. -- Ed.] Pivotal performed what many called a face-saving exit from the big data market in order to focus on Cloud Foundry, which Pivotal seems to consider the most profitable open source project ever. I wonder, due to the weirdness of how the EMC federation works: Did people buy Cloud Foundry, deploy it, and start using it? Or did they agree to a certain spend with EMC -- and allow part of the sum to be allocated to CloudFoundry? A year later, I still haven’t seen a lot of private PaaS -- or PaaS at all.
Hadoop going away: Score one for me. HDFS is here to stay. I think YARN is, too, at least for 2016 -- but we’ll see some changes in the coming year. On the one hand, big data analytics will be owned by Apache Spark. On the other hand, we'll see more Impala and Kylin will emerge for data warehouses. Cloudera has a new file system called Kudu for such use cases, so HDFS is less of a given. I guess the real questions will be “What is Hadoop?” and “Where does this leave the big distributions centered around Hadoop?” Was the Linux distribution model the right choice? I was correct in 2015. For 2016, I’m glad I work on solutions and am not tied to one vision of big data technologies.
NFC, ApplePay, whatever: Can I pay for gas, groceries or restaurants with my phone yet? No. I pretty much buy the rest on Amazon. Like I said, this hasn’t happened.
Netflix’s survival: I cut the cord and now get only three channels: Netflix, Hulu, and Amazon. I hear the NFL and other major sportsing outlets are offering their own streaming services. Meanwhile, TWC/Comcast was nixed. The battle to prevent innovation in network speeds and preserve the cable/telco oligopoly continues, but as I predicted, Netflix remains ascendant.
Amazon Phone: Ever seen one? Did I nail that or what?
Microsoft’s staying power: Here I both hit and missed. Windows 10 happened. It bricked my wife’s laptop, but my eldest loves it. The reviews are mixed. Meanwhile, Microsoft is trying to tell everyone it loves open source -- while at the same time threatening to sue various people using Android or Linux in their products for patent infringement. Meanwhile, Azure hasn't become dominant (or even a close contender) or done too poorly. I’m still bullish on Microsoft, but a little less so than I was last year.
Smartwatches: Some people bought them, but the reception was relatively lukewarm, so I was half right. The Apple Watch is in that phase where some of the people who own them act like Appholes and tell you how special they are. I shared an Uber with an Apphole whose watch kept dinging -- he’d pull out his phone, do something, put it back in his pocket, and 10 seconds later, repeat. This was for a 45-minute ride to the airport in NYC. What did this really bring to his life?
My final grade
For 2015 I give myself a B: 10 points off for the big Pivotal miss, but maybe a couple points back since Cloud Foundry didn't exactly take over the world (whereas Docker will likely do so). Take away a few more points because smartwatches are doing better than expected (mainly as health monitors), and Microsoft is not doing quite as well as I expected.
There's a cliche pitch that begins, “If you have a functional crystal ball...” As well as I did with last year’s predictions, I feel far more confident about 2016. Don’t worry, we got this.
Stay tuned for the annual list of things to worry about.