Re: Discourse forum Lassi Kortela 26 Nov 2020 19:38 UTC

> It might be an interesting project for an economist to find a way to
> quantify the preciousness of a subdomain (given they're not
> up for auction), but I think we can all agree there needs to be *some*
> bar set as to what gets given a domain and what doesn't, and we probably
> won't be able to agree on a written definition of where that bar is :-)
>  From my perspective, I think the biggest "value" at stake is real estate
> on the front page of; if a subdomain points at a dead
> project, nobody will know unless they go looking under the covers, but a
> link to a dead project on the front page is poor.

Good point. Agreed that we should prune languishing subdomains from the
front page on some schedule.
currently has a `display?` property for each project. If set to #f, the
project is hidden from the front page listing, but the DNS keeps working.

> Based purely on gut feel, I suggest that subdomains be given to:
> 1) Any existing thing that meets the basic standards wrt. privacy and
> relevance to Scheme and so on, and is showing some community traction.
> 2) Any new thing that there's pretty broad consensus on the need for.
> Any proposed new thing that *doesn't* get broad consensus so is rejected
> by (2) isn't murdered by the tyranny of the vocal majority crushing
> innovation - it just needs to start on a domain and prove
> its worth and thus become eligible under (1).
> Anything that dies can be removed from the list (and hence from the
> front page), but there should probably be some grace period between
> removal from the front page and removal from DNS, to allow for any
> remaining users to migrate away, while not pointing any new visitors at it.
> Converting those hand-wavy definitions of "some community traction" and
> "dies" and "pretty broad consensus" is left as an exercise for the reader.

All of those are good ideas.

Perhaps we could use sub-sub-domains for experimental stuff in some
cases. etc