3 ms·

The problem is the part where you think about the framework instead of what you are building! For 15 years: My bet is on React, see the amount of apps currentl

The problem is the part where you think about the framework instead of what you are building! For 15 years: My bet is on React, see the amount of apps currently build on React, see multiple large companies backing it, see the #huge ecosystem, all the jobs. Even if it goes out of fashion we'll see another 15 years of it. Other mentioning React will be like jQuery supports my point: Out of fashion, but well maintained and still used a lot. (Side note: Rails is also like that imo). Svelte does an interesting thing with the compiler, but otherwise not that revolutionary, and in some parts questionable. In terms of longevity they have a ways to go. But 15 year longevity seems like a strange criteria to me. Is the app you are building really that important that it will need to survive 15 years, but not important enough to update it to newer tech? I think it's also good to realize that all the options are largely similar and only differ on details. It's a different flavor of the same thing. Learn one and you will be able to pick up the others very easily.

jfengel 3y ago
Is the app you are building really that important that it will need to survive 15 years, but not important enough to update it to newer tech? I find that's not uncommon in large corporations. The app does what it does, and nobody particularly wants to rewrite it. It gradually both petrifies and rots, and yet a ton of workflows go through it. For that scenario you want it in a framework that lasts, even if it's not the best one. Some poor schmuck has to keep it limping along, and then pass it off to some other poor schmuck. Eventually the whole thing comes crashing down and you have to deal with Conway's Law -- the whole corporate structure has been built to reflect the scope of that elderly piece of software and everything cobbled to deal with it. You get to re-do it in the latest-and-greatest -- but hopefully one that will last another 15 years.

jfengel 3y ago

Is the app you are building really that important that it will need to survive 15 years, but not important enough to update it to newer tech? I find that's not uncommon in large corporations. The app does what it does, and nobody particularly wants to rewrite it. It gradually both petrifies and rots, and yet a ton of workflows go through it. For that scenario you want it in a framework that lasts, even if it's not the best one. Some poor schmuck has to keep it limping along, and then pass it off to some other poor schmuck. Eventually the whole thing comes crashing down and you have to deal with Conway's Law -- the whole corporate structure has been built to reflect the scope of that elderly piece of software and everything cobbled to deal with it. You get to re-do it in the latest-and-greatest -- but hopefully one that will last another 15 years.