Thank you! But Crystal's policy is to not have aliases.

Thank you! But Crystal's policy is to not have aliases.

At all? I mean… what about Indexable#[] and Indexable#at? #fetch use the same construction. And it has more implementations (with default value without block, you would not write [1, 2].at(2, 4)). And it's make Indexable (and Array) more similar to Hash (which is also Indexable).

at is a bit of an exception, because [] can't receive a block. And it would be weird to have [] (no block) and at() { ... } (with a block), so at without a block exists too. But the non-block variant could be removed, and maybe even at removed and renamed to fetch, for consistency with Hash.

at is a bit of an exception, because [] can't receive a block. And it would be weird to have [] (no block) and at() { ... } (with a block), so at without a block exists too. But the non-block variant could be removed, and maybe even at removed and renamed to fetch, for consistency with Hash.

Sounds reasonable. I've created a new PR #6296 (from renamed branch).

Just note that this doesn't mean it will get merged. It was just a suggestion to open a discussion. I guess it will be fine because I doubt many are using at ([]? is more common), but since it will be a breaking change there will be more resistance to accept that code.

Just note that this doesn't mean it will get merged. It was just a suggestion to open a discussion. I guess it will be fine because I doubt many are using at ([]? is more common), but since it will be a breaking change there will be more resistance to accept that code.

Thank you, I know. I just tried to rename the branch for this PR and it closed PR. 😅

I'm ready to discuss and change solutions.

I just want to change something like this:

port = (ARGV[0]?.try &.to_i?) || DFLT_TCP_PORT_NUM

to this:

port = ARGV.fetch(0, DFLT_TCP_PORT_NUM)

(Example by @DRVTiny)