Hacker News Reader

Comment by pjmlp | Hacker News Reader

Currently I just want some form of genericity, regardless which one, but I fear we will go through yet another cycle of community feedback and by August 2021 still not there.

However not to be just negative, I will certainly take the opportunity to provide my positive feedback as well, so looking forward to play with go2go.

View on HN

I'm sympathetic to being somewhat impatient; I'd really like to have generics in Go years ago too. But I'm also glad they didn't just plow ahead with the first draft of the proposal that came out; the way contracts worked was horrible, and the slow deliberative process means that we don't actually have to deal with those mistakes. Unfortunately, good design takes time.

Me too. A bad generics implementation is profoundly worse than no implementation at all. This angers people who have minimal experience with Go and it certainly seems counterintuitive until you understand that the way the community writes their code affects your experience whether it’s just a different paradigm or a conflagration of paradigms, whether it’s a little abstraction or gratuitous abstraction, etc.

Yeah, glad that work is being done. But surprised to see this status update be that nothing substantial has changed in the last year.

They've reworked the draft, written a generics-capable type checker, a translator tool that lets you use the draft in the current version of Go, and published a generics-capable version of the playground. How is none of that substantial?

The Featherweight Go paper also just recently came out, which is an important step:

https://arxiv.org/abs/2005.11710

Reply on news.ycombinator.com