I doesn't have to stop the waves, it only has to convince the homeowner that the waves are being stopped.

That must be the source of the selection pressure that is causing cell towers to evolve into pseudo-trees.

(One day I hope to come across one while it is bearing fruit -- I assume that is how new cell phone varieties are cultivated)

If you're suggesting that capitalism is always a zero-sum, winner-take-all game, where only a single company can survive in any given market, then perhaps it is you who doesn't know how capitalism works.

Much like the English language, which is also quite useful and therefore widely used. Being useful over a wide variety of scenarios and being bloated-and-complex are often two sides of the same coin.

If one wanted, one could design a scanner whose only data-output path was an audible tone. You could download a list of license numbers into it (using a unidirectional data transfer, e.g. via a serial port with the device's TX pin removed), and then it would beep if it saw one of the plates in the list, and that's all it would do.

To hack it to output a list of license plates it had scanned that day would both require hardware and software modifications -- not impossible, but inconvenient enough that it's unlikely most police departments would be capable of doing it.

Perhaps it should be?

Is making the hyper loop even faster considered desirable? Unless they can find a way to make the path of the track/pipe perfectly straight, the passengers will experience g-forces proportional to the velocity of the vehicle whenever they go around a curve. Too much speed could make the ride rather uncomfortable. (You can reduce that by making the track straiter, of course, but that reduces your flexibility in placing the track)

It's good that you're confident in your company's abilities; but in order to win the contract, you'd have to gain the confidence of the decision-makers in the school district as well. As any company involved in outsourcing over the last decade can tell you, a cheaper up-front quote is no bargain if the delivered product is screwed up.

This is why it's so critical to include scheduled malfunctions in your control logic. That way you'll get called in every 6-9 months to "fix" the system, which you'll be able to do very quickly since it is just a matter of resetting the timer for next time. You'll make a few hundred dollars each time, and everybody will recall you fondly as the indispensable genius who is the only person who knows just how to keep the system running. Just be sure to randomize the timeouts a bit so that nobody catches on ;^)

(disclaimer: I'm joking; I don't really advocate doing this)

All well and good, if the failures are in fact randomly distributed.

OTOH, if the failures are caused by particular entry conditions (aging/temperature/pressure/whatever) and all of the valves are experiencing those entry conditions simultaneously, then the likelihood of all valves failing simultaneously may be much higher.

It is pointless, in a single-user environment, where all the software was written by the same guy, with no networking, who never makes coding errors.

Admittedly that's a pretty specific use case, but it works for him. It reminds me of coding on the Amiga, back in the day. Bad pointer dereference? Oops, total OS crash and reboot, and if you're really lucky now your filesystem is corrupted and you've lost the source code you were working on. That'll teach ya... ;)

What Clippy gets is a good swift kick in the pants. Siri should count herself lucky to only get 76.5% of that.

It's a cost of doing business? If you think Apple is bad, try developing games for Nintendo.

I know a number of software developers who think along these same lines... ;)

Also, please do not vomit on the upholstery.