Hi Artyom,

I should emphasise my main use-case is the IRIs - so I implemented these
largely as a container for names for identifiers (in RDF), so I didn't flesh
out the proposal with much beyond RFC 3986 and 3987.

This seems a little remiscent of time-zone data being bundled - presumably
we would have to keep maintaining and updating the external list somehow.

Would this be better as a separate proposal or an auxilliary library?
I guess that it could go alongside IDN or punycode as it concerns additional
interpretation of hostnames (in the HTTP scheme?).

Thanks,
Duncan

> On 29 Jul 2026, at 19:52, Artyom Bologov <xxxxxx@aartaka.me> wrote:
>
> Hi y’all, hi Duncan,
>
> There’s this use-case of displaying the URIs prettily in web browsers
> and other web-related tech. This usually strips off the https://,
> shortens host to just the domain name, and lists path. Such is the
> minimal URI that can be acted on.
>
> However, this requires a way to get the domain from an IRI host. This is
> not trivial, as there are TLDs like .co.uk etc. I suggest adding
> iri-domain procedure that processes the host in such a way. This’d
> require depending on IANA’s list of TLDs
> <https://data.iana.org/TLD/tlds-alpha-by-domain.txt>. But I consider it
> ultimately useful and deserving an extra mile.
>
> iri-tld might be useful too, to quickly get the TLD and act on its
> national / domain-specific (pun not intended) nature.
>
> By the way, how does SRFI 275 deal with IDN / punycode?
>
> Best of luck,
> --
> Artyom Bologov
> https://aartaka.me