Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I empathize with the unique difficulties of the web and I don't doubt that the complexity solves for real problems (e.g., old IE versions); however, I take issue with treating all problems as equal irrespective of how niche those problems are. The C and C++ folks make the same kind of argument to justify their abusive build systems--if the build systems made sane assumptions, then it would exclude certain niche use cases. Of course, we all agree that sane tooling excludes niche use cases, but we disagree about whether excluding ~1% of use cases for a dramatically improved experience for the ~99% is worthwhile (and we may even disagree about some of the qualifiers I used in this sentence).


Webpack sure has problems (for example speed) but most of the more niche things that people do aren't actually supported by it but implemented through the plugin system, at which point any category of niche/sane kind of flies out of the window. If you enable people to do weird things people will do weird and maybe unwise things. The best solution might be to take notes of how people tend to develop js stuff nowadays, scrap javascript completely and use something completely new that doesn't have the issues of javascript. But that seems rather unrealistic at this point.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: