@42bloom I don't see why 6-week cycle is blamed for amount of features. Frequency != velocity. If Rust had 6-month release cycle, you'd just get 4 times more features per release.
Rust isn't developing features in 6 weeks. It's taking *years* to develop features, and when each feature is ready, it gets stabilized and released in the next scheduled release, whenever that happens.
Rust could release new versions every day, or once per year, and it wouldn't change how features are developed.
@kornel I have no idea which one is the cause and which one is the consequence. But I'm persuaded that short release cycles and features bloat are connected, and if nothing is done, may lead to Rust's collapse (because everything collapses when complexity becomes uncontrolled).
@42bloom I can tell you for a fact that they're not connected. Search for "tracking issue" in Rust's GitHub. You will find that many of them are years old, some connected to RFCs that are still years older, and borne out of pre-RFC discussions that are again years earlier.
Here's a list of unbaked Rust features:
@42bloom For example, poll-based async model has been proposed in 2015. In 2016 there has been proposal for the Future trait and a prototype. 2017 started big-scale real-world implementations and testing. In 2018 there was work to polish the spec and nail down the syntax details, and it finally released in 2019.
That was many years of work of many teams, with burnout-inducing amount of bikeshedding. Nothing was fast about it.
Server run by the main developers of the project It is not focused on any particular niche interest - everyone is welcome as long as you follow our code of conduct!