There are things that last because they were done once and were done, mostly, right. Then there are things that have lasted because they were done once, mostly

There are things that last because they were done once and were done, mostly, right. Then there are things that have lasted because they were done once, mostly wrong, and still provided enough value at the time that they became popular among the group with the power to determine what would be the dominant solution before they could be refined. The flip-side of permanence is path-dependency. So many programs get written for a certain architecture of chip and instruction set, for instance, that to change it to something more efficient would break too many things. To make something that lasts because it is good, that lasts on its own virtues, would be a wonderful thing. Nonetheless, the next generation ought to be able to surpass us - and if they are not able to do so, we have done something wrong. Blanket approval of past solutions simply because they have lasted seems likely to lead one wrong.