> Locking onto USB-C now could be a very "640k ought to be enough for everybody" moment.
I agree in general; but at some point don't we hit diminishing returns? If I were to say, "18.4 exabytes of RAM ought to be enough for everybody", that'd be basically correct, even future-proofing for the craziest sci-fi holodeck use cases we can imagine, right?
I dunno. Having lived through the days where having 128kB of RAM was an eye popping upgrade, you could easily have made the argument back then that 2GB of RAM was future proof for as long as the eye can see.
And in fact, they did. By the time I was reaching adulthood, that was that hard limit built into Win95, give or take boot errors if you had a quarter of that (which was gigantic at the time; 64MB was more than enough to boot it).
Fast forward another 25 years to today, and laptops routinely have 16GB of RAM, and higher end Mac Pros can apparently get a whopping 1.5TB of RAM.
Put another way, the high-end went from .1MB to 1TB in 40 years or so.
I get that we're reaching the physical limits of miniaturization, but I'd still be surprised if my grand kids don't have exabytes of RAM or equivalent on whatever the hell their main devices will be by the time they reach adulthood.
I agree in general; but at some point don't we hit diminishing returns? If I were to say, "18.4 exabytes of RAM ought to be enough for everybody", that'd be basically correct, even future-proofing for the craziest sci-fi holodeck use cases we can imagine, right?
But the legislation doesn't lock into something that might happen in the future. It locks the future to what we have today.
As much as I am for interoperability, a legislation directing a specific technology is one way how we end up with ridiculous standards.
Remember, we still have laws on where we can or cannot tie a horse in town on the books a century after they were irrelevant. Laws move in centuries, while technology moves in years. They are not compliant with each other.
I agree; if this law were put to a vote, I would vote against it. I just think it's conceivable that we hit a point where there's very little end-user benefit to additional improvements, and we achieve an equilibrium that essentially solves the given problem[0]. The catch is: we don't necessarily know in advance at what point that will actually be the case (hence the irony of the 640k line, which I'm sure sounded very sensible at the time, just as the 18.4 EB thing sounds very sensible to me now).
[0] One of the reasons that the loss of the headphone jack is so exasperating, is that's an example of an ideal universal connector; it solved the problem so well, there was little reason to improve on it, either in quality or function.
Well, there's the rub: 32 bits for IPv4 was clearly inadequate. IPv6, on the other hand, gives us with 128 bits = 3.4 x 10^38 addresses, which seems from our vantage point like it should cover us until the end of time, even if we colonized the whole galaxy. Is that actually the case? Time will tell. ;)
But that's just the point. When they decided on 32 bits, then it was more than sufficient for the intended use case. If we had stuck to universities and government uses, then we probably still wouldn't have run out.
Things change or evolve, and often in ways not foreseen.
I agree in general; but at some point don't we hit diminishing returns? If I were to say, "18.4 exabytes of RAM ought to be enough for everybody", that'd be basically correct, even future-proofing for the craziest sci-fi holodeck use cases we can imagine, right?