I know lots of folks would like to pretend that the latest fancy scripting lingo is going to be as longstanding as COBOL, but let's be real - if your day job is still maintaining business apps you wrote 10 years ago in something like Silverlight, either your company is rich enough to keep stuff long past the refactor expiration date, or too stupid to realize that they're spending more on maintenance than they would on a rewrite.

I, too, read Guns, Germs, and Steel a number of years ago. The difference is I paid attention when more contemporary historians started calling Jared Diamond's work a load of shit ~15 years ago. It was a fascinating hypothesis, but at this point it's mostly historically-flavored fiction than anything else.

Then what's the point? All of these gorgeous synthetic benchmarks fall flat on their faces when it comes to light that Intel was lazy yet again with some part of their core design requiring a predictive pipeline to be disabled for security reasons. If you're ok with a significant chance of becoming a gigantic fireball, there are redeeming qualities to a Pinto performance too!

Unmitigated hatred of M$ with arguments that haven't been updated in the decade since they've been relevant? Yup, must be a 4 or 5 digit SID.

Most of the 'it wasn't that important' comments seem to be coming from higher IDs - I don't know many graybeards who would argue that USB was one of the most important innovations in widespread acceptance of home-scale computing, right up there with the Mouse Cursor-driven GUI. USB extended the hardware-level customization of your system beyond opening the box up and using screwdrivers and dip switches, which made it more accessible for everyone on the user end. Having a 'real' standardized interface and software design meant that whoever was making your device was obliged to at least try to make it work, or face backlash. USB was *great.*

"students who are willing to work a nine-month contract for $14 an hour in exchange for the chance to work on a triple-A franchise." The big issue is that the people who trumpet how much cost savings they realize by hiring inexperienced devs are silent when it comes to admitting that overruns and obscene crunch are often cause by inexperienced devs. If there was a consequence to shitty project management, there would be improvement, but there isn't, so there's not.

It's not just local environments that get handicapped by the lowest-common denominator tech hurdles - I support a major enterprise-grade compliance application that has components of the core code that require 32bit OS environments, so we have security exceptions to keep 2008SP1-32 running in a freaking bank.

I happily second this. More wireless peripherals should have cost $10 more and have a BT radio built in, it's so much nicer to not need to swap a dongle around between my laptop and server racks at work.

To be fair, Apple has developed a rather shitty reputation in the supply-chain universe for doing just that. Becoming %60+ of a vendor's business for a few quarters (if you are lucky!) just to pull out because someone else can save you $.001 per part is devastating to the supplier unless Apple footed the bill for infrastructure improvements to scale up, and even then, you can put a successful business on life support by disrupting enough of their durable business with your temporary gigantic order.

You reference Germany, Japan, and China in the same breath, even though they represent 3 vastly different points on the spectrum for this sort of thing. German suppliers are happy to tell you to get bent if you suggest they radically change what is profitable for them just to bump up 2 quarters of revenue (that's a uniquely American perspective, really...) Japan is likely to use a banking environment favorable to domestic business to underwrite it instead of either party actually fronting cash, and China is just going to tell you what to do.

Nuclear power has been a political issue and not a financial one for decades now. If Gates can convince the Hill to unfreeze effectively 30 years of bullshit that has hamstrung the US civilian nuclear program, I'd be happy to jump in the boat with my own money (albeit, much, MUCH less than he has.) Until then, you could promise a bajillion dollars and not move the needle when it comes to a viable solution for getting us off the fossil teat.

Ah, the old 'Avaya' strategy. Bold move Cotton, let's see how it works out for them...

Except the rich-ass people who buy this stuff aren't bothering to add a bunch of component audio to it anymore. Think about it, these people are the reason we have crappy wireless audio mandated on most phones.

It's frustrating (but not entirely surprising) that Evolution, which is a great example of statistical probability in real-world action, is fundamentally misunderstood by so many people. Evolution != Design. Sure, it came up with some great solutions, but it also drug a bunch of (effectively) useless baggage along for the ride because the mutation didn't happen to reduce survival and reproduction in a particular environment.

Literally the first comment I can see to you completely misinterprets how probability and natural selection work to argue your point. I feel your pain.

Wait, are you excusing the NY Subway because it's older? Because I know that at work, we spend a ton more to maintain the old system that services 8,000 users than the new system that services 1,000. When Seattle has a century-old system on which they've ignored basic maintenance for multiple decades despite the userbase increasing 100-fold, then I'd imagine people would get a bit irritated. Yet, around the globe, systems that are virtually the same age as NY's result in happy riders who cannot say enough good things about them.

NY is reaping the cost of deferred maintenance, and doing so in about the most visible manner one could design.

I've enjoyed the Hub as well, but as with most home automation, it works best when you keep it simple, thoughtfully designed, and FAST. I'm frustrated every time I have to log into to Logitech for making changes, and it's noticeably less responsive if I am not on the local network. If they break the super-basic functionality I have it for in the name of adding Alexa or some shit, I'm out.