There’s consumer-oriented application development, and there’s business-oriented application development. Within the business side, there’s end-user application development, and there’s infrastructure application development. Then there’s that weird, in-between, oddball, bongwater-swilling world called “DevOps”.
The latter of which started out unofficially as one definition of (more or less) gluing infrastructure pieces together, under the hood, and then I blinked, and opened my eyes to see it had been gang-raped by a room full of PMP-certed BA’s with intravenous metric-itis while some PR folks held the Go-Pro and came up with the name “DevOps”. Now it has a more respected, semi-official definition, which (more or less) involves spinning up or scaling out data center services in support of cloud services. That sounds impressive, and I didn’t even plan on saying that.
Is “scalable” a bonafide English or Anglish word? My spellchecker doesn’t like it. (doesn’t like DevOps, either). Pffft.
(long inhale + exhale…)
More than once I’ve brought up the tired story of developing web console interfaces for things like Active Directory and Configuration Manager. I hear myself saying the words and all I can think about is how I must sound like that old senile relative at the cookout who keeps telling everyone the same old story about meeting someone famous, once. With each new person walking by, the story repeats, and the nearby guests are contemplating who will be the first to put a bag over my head.
Whether it be WWA, CMweb, DPMS, ADWA, CMINFO, or APPINFO, or whatever the acronym ended up. The same soup with new ingredients.
I’ve had a lot of those discussions. At first, they were exciting. Why? Because it felt like someone gave it credence and value. Eventually however, I’ve learned it’s not that different from parents hearing other parents comment that their tiara toddler will be the next Kim K. while waiting for the judges scores inside the shopping mall.
More times than I can recall, the other person would say something like, “you should market that!” or “you should sell that!“. I’m thankful for the kind thoughts and words, but the reality is 180 degrees out. Here’s why:
- Most web apps which are tied to on-premises infrastructure services are intranet, rather than cloud-hosted, by nature.
- AD already has a web interface, sort of. It’s called Azure.
- ConfigMgr doesn’t have one, but Intune does, and my money is on the two becoming one in the not-too distant future.
- Most of those projects I was involved with were born from internal needs, rather than an outright proclamation like “let’s get someone to build this!“.
That said, if you (or I) tried to market something like this to other enterprise-scale customers, the reaction isn’t very friendly. I know, because I actually gave it a try. Granted, I come from limited means, so my “try” wasn’t like someone with VC backing would “try”. I got the reaction of someone watching a live cow being lowered into a running wood-chipper. Like I said, it wasn’t very friendly.
Being the inquisitive type that I am, I would usually strike up a follow-on conversation about “why not?” without layering any weird vibe, just to learn as much as possible about their perspective. It typically fell into one or both of the following reasons:
A – Too difficult to assess the value without having an operational “demo”.
B – Tired of dealing with small developers with cats running around their desk during phone calls.
Given that a “demo” version was easy enough to build, that still left the second hurdle to overcome, and none were willing to go there unless my name was Mark Zuckerberg, Johnny Ive or one of the Google guys. Little old me wasn’t enough to get their check books to magically appear.
Their first priority, understandably, is to identify an off-the-shelf solution. After that, it’s on to finding an off-the-shelf solution that allows for customization. Then after that it’s an off-the-shelf solution which they can bend their internal processes to suit (basically, re-shaping the bolts to fit a new wrench, a la SAP).
Is this set of priorities inherently “bad”? No. I’d probably do the same thing.
The only way to provide any assurance of success for a new infrastructure product offering within such a narrow market is to position it from a market-related reputable vendor. The guy writing code from his house in Virginia (cough cough) is about as likely to make a dent in that situation as Red Sox have at winning the 2015 World Series.
The new era is focused on cloud and mobile. Whether it’s business or consumer, it’s cloud and mobile. I don’t see a niche for that sort of thing on the horizon. Anyhow, I digress (and digress often). This probably sounds like self-pity, but it’s really not. I’m very pragmatic about all of this. It’s time to archive that code and push on to other things.