The Toml library seems abandoned. Toml is a relatively popular format for configuration files; I think it’s used by several projects (including one of mine). Should we, as a community, fork the library to keep it alive?
I have YAML on the roadmap for my Orsetto hobby project, and I suppose I should add TOML too. I’ll be watching this thread to get a feel for which is likely to be the higher community priority. My own preference is for YAML first, then TOML later, but I’m open to reconsidering if there are significant problems with YAML that I don’t know about because I generally try to avoid human-readable configuration files altogether in my designs.
Forking and maintaining Toml would be great. I currently pin the repository, and it still works fine, even in 4.07, but Toml should be maintained. It’s a popular format for human-consumable configs.
I think it’s a good idea, Toml looks like one of the least worst config languages out there and having a well-maintained version might encourage people to pick it over less desirable languages like YAML for their programs.
My policy is always to make a good faith effort to contact the maintainer first, and if they’re unreachable, forking seems fine. (If they are reachable, often they’ll hand something over. If they’re reachable and don’t want to maintain the thing but unwilling to hand it over, I think forking is still morally fine.)
One option I’ve suggested before: an orphaned ocaml packages project on github, which would maintain interesting pieces of software that no longer seemed to have homes. (If people like the idea, I will set one up.)
You mean, a github organization (say, “ocaml-community”?) that would gather “official” forks of dead projects? That sounds like a good idea to me, if done in coordination with opam maintainers.