Identifiers starting with ":" Jeronimo Pellegrini (24 Jun 2026 00:17 UTC)
Re: Identifiers starting with ":" Peter McGoron (24 Jun 2026 20:34 UTC)
Re: Identifiers starting with ":" Artyom Bologov (24 Jun 2026 20:40 UTC)
Re: Identifiers starting with ":" John Cowan (25 Jun 2026 05:54 UTC)
Re: Identifiers starting with ":" Shiro Kawai (25 Jun 2026 06:21 UTC)
Re: Identifiers starting with ":" Peter McGoron (25 Jun 2026 15:01 UTC)
Re: Identifiers starting with ":" John Cowan (25 Jun 2026 18:59 UTC)
Re: Identifiers starting with ":" Peter McGoron (26 Jun 2026 19:02 UTC)

Re: Identifiers starting with ":" Peter McGoron 25 Jun 2026 14:59 UTC

I'm leaning towards exporting identifiers without prefixes (i.e. what
was in SRFI 144 `fl+` and is now `:+` will become `+`). The reason for
this is not really because of the keywords issue, but because it would
allow one to do `(import (srfi 276 ⟨representation⟩) fl)` and easily
have code portable to R6RS and SRFI 144 without a lot of manual renaming.

There is still the issue of escaped identifiers names (i.e. `|1/e|` for
that constant), because SRFI 144 prefixes those with `-` and escaped
identifiers are not standard in R6RS. I could move those to a sublibrary
`(srfi 276 ⟨representation⟩ constant)` so they can be prefixed differently.

Or I could just keep the `:` prefix. Any implementation that treats
identifiers that start with `:` specially needs to have a way to escape
that anyways. Programmers should always import these libraries with a
prefix.

-- Peter McGoron