4 ms·
I like this a lot. I would say that going with one of the existing standard libraries is a start - lodash is a good example. Another thing is anyone can roll m
by hateful 6y ago
I like this a lot. I would say that going with one of the existing standard libraries is a start - lodash is a good example.
Another thing is anyone can roll most of these tools into standard libraries if the source license allows them to. So maybe some "rollup" packages are in order. You would get more than you need, but if you setup any sort of tree-shaking (which has been the focus of many build systems lately) then you can only bundle what you used in the end. One other thing that may be an issue is if you have two major libraries and each one is missing one function that you need - then you need both...
- trollied 6y agoThe point of having a standard library is that it'd actually ship with the interpreter, so you wouldn't have to "get more than you need" - it'd already be there without having to download. Obviously there would be a period of time (years) whereby you would have to do a conditional download, but the longer term aim would be that it's already provided as part of the base language.