> we don't need 500 crappy, slightly different ways of doing the same thing, we need maybe 5, battle tested and hardened bullet proof solutions that are well engineered.
As much as I'd like to do that, I think we're some ways off from the sorts of stability that allow for that. Sure, we're all neophiles, so there's a fair bit of people using new tools for the fun of it. But I think a much bigger problem is the extent to which our raw materials keep changing under us.
For example, the smartphone isn't even 10 years old. We're all still replacing them every couple of years because the new ones keep getting better. Processing power, memory, power characteristics, sensors, network, ports: they all keep changing. The number of processors in my house is easily 5x what it was 10 years ago. The range of capabilities is larger, too: my phone is way smarter, but my lightbulbs are quite dumb.
Even if that were stable, our conceptions of computing are still changing. Most of the code we write is in languages that are 20 years old, but we've learned a lot since those languages were created. We talk about "virtual servers", which as a mental model is much like "horseless carriage" or "radio with pictures": it's a sign that our understanding hasn't caught up with reality.
We also have an enormous amount of business churn. A ten-year-old tech company seems old. Venture capitalists invested $58 billion last year, mostly in service of disrupting some existing order. Few will see a point in investing for the long term when they aren't sure if there will be one.
Even given all that, I think we underinvest in the long term, for reasons I'm sure we'd agree on. But I am often forced to admit that the level of craftsmanship I personally want is at odds with the current practical reality of software.
And I'm not alone. Back in 1992, Steve Wozniak was hoping for the end of Moore's Law. Bob Cringely writes: "But while the rest of the computing world waits worriedly for that moment when the lines etched on silicon wafers get so thin that they are equal to the wavelength of the light that traces them—the technical dead end for photolithography—Steve Wozniak looks forward to it. 'I can’t wait,' he said, 'because that’s when software tools can finally start to mature.'" [1]
The rise of mobile and cloud computing ruined that plan, of course. But perhaps just about the time you and I retire, the technological base will be stable enough for people to do some serious, long-term building.
[1] http://www.cringely.com/2013/04/03/accidental-empires-chapte...