Ian Manske

> Can we provide some recommendations on the plethora of `munch_it_somehow_into_a_string` methods, when to use what? I was able to remove yet another one of the `to_{kind}_string` functions. Otherwise, I...

Sure thing, here are all the numbers (on my hardware). Top is main (bed236362a180a73893f3e15e4e9eefa8c222ebf), bottom is this PR. Table ``` benchmarks fastest │ slowest │ median │ mean │ samples...

> Can they be isolated into a separate PR in an easy way or will this become a furball here? Actually, it looks like it might not be too bad...

Ok, I've split off #12782 and #12799 to make this PR slightly smaller. I've also removed the exit code changes. Otherwise, I just have to add doc comments and maybe...

Oh that's cool, definitely something we could try in the future. In the mean time, the bytestream PR should allow us to run the regex on each line of the...

> Conceptually I think that the CI part should only go with the code itself being removed. I don't want some baggage in the repo that may not compile anymore....

> we may want to deprecate the closure arg? I would be in favor of deprecating the closure arg. We would have to figure out another way to implement the...

@fdncred Are you referring to the file redirections (`o>`, etc.), the pipe redirections (`e>|`, etc.), or both?

IMO, all redirections should either work with internal streams and values, or they all should not work. Having some work and others not work is inconsistent. Following from this, the...

> which I assume will stop working with this PR since http get isn't external. That is correct. `http get` outputs a stream of values or a byte stream, and...