No one is going to see that unless they go looking for it though
>you shouldn't have used "~", "^"
That is how the npm ecosystem works, and in any case the issue arose from dependencies that in turn depended on underscore and thus are not under my direct control.
> Anyway, with a decent test suite, surely a modicum of (automated) testing has already identified the pinning that should take place?
My particular issue is in a new project I could not install several dependencies as npm downloaded the newest compatible version which was supposedly 1.7.0 - the only way to fix this is to manually create a lock file.
>you depend for free.
True, and I'm tremendously grateful for all the work that has gone into it and it would be a massive shame if that work was for nothing if a continued disregard for semver forces people to switch to lodash.
Do you imagine that the maintainers of all these dependencies are going to catch any fraction of the hell that jashkenas seems to be catching for this? It's just as much their fault as it is his. Why don't they pin and test?
https://news.ycombinator.com/item?id=8244888
Do you mean to imply that breaking-change frequency should stay below some particular maximum value? That argument is completely unrelated to the use or disuse of semver.