Tomas Grosup

I think the existence of side effects is not a showstopper as long as the different execution order is documented and cannot break existing code without people opting into it....

Indeed. The F# compiler currently does not implement any checks needed for this to work correctly, so it would allow you to author unsupported code. If you can do those...

We agreed with Vlad to close this. Such change would be too dangerous.

This is a good idea, editor could then put all such occurences into a task list. I have used the "rocket" icon to indicate preference for "nyi" until a better...

I think this PR is good to attach for a new fslang-design RFC, please follow the instructions leading to a template here: https://github.com/fsharp/fslang-design#the-process There are sections in the template to...

Is there a language where programmers can do folds over collections into a single value, without using a `fold`-like function call? The closest similarity I could find is using `...

I get the benefits and how it fits pragmatic programming, my question was really after if this was being solved in any **other programming language** or its library ecosytem. Also...

> Racket's for/fold: https://docs.racket-lang.org/reference/for.html#%28form._%28%28lib._racket%2Fprivate%2Fbase..rkt%29._for%2Ffold%29%29. This is a nice reference, thanks. I like how Racket's syntactic forms allow them to build upon a basic construct with a consistent naming scheme. It...

I keep thinking about a syntax that would not feel enforced for a singular use case. For limited scenarios where one can pass built-in functions/operators, the sample provided by @brianrourkeboll...

Reading the proposed lists, is there any function where the expectation would be to utilize inner structure of the Set and gain a lot better perf characteristics compared to using...