4 ms·

Misuse aside, cloud computing (storage or otherwise) is in its infancy so everyone is trying their own take. Its real use will become apparent. In the 1980s eve

by g_lined 16y ago

Misuse aside, cloud computing (storage or otherwise) is in its infancy so everyone is trying their own take. Its real use will become apparent. In the 1980s everything had a clock in it because suddenly LCD screens were small enough to put a clock into anything the size of a pen or bigger. Now we've moved on and realised that it's not /clocks/ we wanted to be small, but chips. We'll move on from 'in the cloud' but in its wake will be left something useful.

tzs 16y ago
And as clocks started showing up everywhere, that transition to and from daylight savings time became very annoying as we had to manually adjust a dozen clocks. I was in college shortly after the digitization of everything started (Caltech, class of '82). I took a course in digital logic. I don't remember the name of the textbook, but I remember the last chapter (or perhaps it was an appendix) and the TERRIFIC advice it gave--still relevant today. The title was "The Engineer as Dope Pusher". It talked about how engineers want to do fun things, and so if not careful will end up pushing things into products that don't belong there, because the engineer wants to play with them. For example, suppose you are an engineer designing the timer for a dryer. On the one hand, you can use the common mechanical timer, with the circular dial. This thing is cheap (because they've been manufactured essentially unchanged for decades and so the process is highly optimized), and it is reliable (again, decades of experience have worked all the bugs out), and consumers all know how to use them. But you've been reading about these new-fangled micro-controllers. Those sounds a hell of lot more fun than a 50 year old mechanical timer. If you go with a micro-controller, you'll get to play with something new. You'll get to write software. Maybe display the time remaining with an LED progress bar, and you can do some neat animations on it to signal completion. Maybe implement a bunch of different drying modes. Never mind that your fancy micro-controller based dryer interface offers no advantage at all to the consumer or the manufacturer. Instead, it costs more for the parts. It involves many more parts--the micro-controller logic board, a power supply, the progress display, the input buttons whereas the old mechanical timer was two parts (the timer, and the dial). The parts are more expensive. It is harder to assemble them. The consumer knows how to use the old system--the new system will probably require them to actually read the manual (oh, someone has to write a new manual). The fancy system might also be less reliable (more parts to break, and the manufacturer does not have experience with running digital electronics in the environment of a dryer). Software engineers in particular have a tendency to become dope pushers, from what I've seen.

by g_lined 16y ago

tzs 16y ago

And as clocks started showing up everywhere, that transition to and from daylight savings time became very annoying as we had to manually adjust a dozen clocks. I was in college shortly after the digitization of everything started (Caltech, class of '82). I took a course in digital logic. I don't remember the name of the textbook, but I remember the last chapter (or perhaps it was an appendix) and the TERRIFIC advice it gave--still relevant today. The title was "The Engineer as Dope Pusher". It talked about how engineers want to do fun things, and so if not careful will end up pushing things into products that don't belong there, because the engineer wants to play with them. For example, suppose you are an engineer designing the timer for a dryer. On the one hand, you can use the common mechanical timer, with the circular dial. This thing is cheap (because they've been manufactured essentially unchanged for decades and so the process is highly optimized), and it is reliable (again, decades of experience have worked all the bugs out), and consumers all know how to use them. But you've been reading about these new-fangled micro-controllers. Those sounds a hell of lot more fun than a 50 year old mechanical timer. If you go with a micro-controller, you'll get to play with something new. You'll get to write software. Maybe display the time remaining with an LED progress bar, and you can do some neat animations on it to signal completion. Maybe implement a bunch of different drying modes. Never mind that your fancy micro-controller based dryer interface offers no advantage at all to the consumer or the manufacturer. Instead, it costs more for the parts. It involves many more parts--the micro-controller logic board, a power supply, the progress display, the input buttons whereas the old mechanical timer was two parts (the timer, and the dial). The parts are more expensive. It is harder to assemble them. The consumer knows how to use the old system--the new system will probably require them to actually read the manual (oh, someone has to write a new manual). The fancy system might also be less reliable (more parts to break, and the manufacturer does not have experience with running digital electronics in the environment of a dryer). Software engineers in particular have a tendency to become dope pushers, from what I've seen.