💾 Archived View for gemi.dev › gemini-mailing-list › 000737.gmi captured on 2024-05-26 at 16:35:40. Gemini links have been rewritten to link to archived content
⬅️ Previous capture (2023-12-28)
-=-=-=-=-=-=-
The specification has URLs for link lines. Is that an intentional choice over a broader URI category? I'm thinking about using text/gemini outside Gemini protocol and would likely use some form of URNs for links. There is also a lovely urn:isbn: namespace that's left behind. I'm new around here and am sorry if that has been discussed somewhere already. p.s. please keep me cc'd.
On Mon, 22 Feb 2021 at 14:01, Vasilii Kolobkov <vasilii at orangeshoelaces.net> wrote: > > The specification has URLs for link lines. Is that an intentional > choice over a broader URI category? > Whilst the spec says URLs, people do put URIs such as `mailto:` in links occasionally, and it's often supported. Whether it's international or not I don't know.
On 2/22/21 9:06 AM, Vasilii Kolobkov wrote: > The specification has URLs for link lines. Is that an intentional > choice over a broader URI category? > > I'm thinking about using text/gemini outside Gemini protocol and > would likely use some form of URNs for links. There is also a > lovely urn:isbn: namespace that's left behind. I cannot read Solderpunk's mind of course, but I strongly suspect the exclusion is unintentional, because I can't think of anything wrong with URNs in gemtext. As Oliver Simmons says, it is already being done. It would be nearly impossible to keep URNs out anyway: URNs and URLs use the same URI syntax, so one cannot generally tell whether a URI is a URN or a URL without understanding the scheme. I guess it should maybe be changed to say "URI" in the spec. -- pjvm
---
Previous Thread: [Spec] <META> in the response header is too vague
Next Thread: [SPEC] Are (simple) commenting systems explicitly unsupported by gemini?