i am not looking for building repairs at the university of rhode island. i am trying to maintain my uris and see what insights other people have had if any

anyway my core intuition is that not only is naming things hard, but also every name is an abstraction and every abstraction is a cost. you could feasibly run an entire service dedicated to maintaining records of which names map to which things. this is basically dns but you don’t generally want to give every resource its own fqdn. so http(s) will more often be used, but people just use locations as names more often than not, which breaks when the location changes. and that’s a maintenance cost.

what i mean by this is that in some ways the primary role of something like http is not to return a document… that is secondary functionality. the primary role of http is to respond to messages, the majority of which are requests to fetch a representation of something which is most often a document. but the mapping of “name to thing” is an entire field of its own. give someone their own domain, point to an http server, and you’ve so far solved very little…

what i want is to be able to maintain a mapping of uris to responses, and have that be portable across various backends as essentially a personal lookup table. a sort of easily user configurable http server, one that goes all-in on ease of config. apache htaccess files are cool but can we do better. nginx gets more widely used but doesn’t allow for as much flexibility. i feel like there’s something there with this idea

the best way for a name to last ~forever is to make maintenance of names as easy as possible within whatever glorified lookup table you are using