/questions/[id]?site=codegolf instead of
//api.stackexchange.com/questions/[id]?site=codegolf to save a few bytes.
In a recent solution I took this a (small) step further, requiring that it be run at the root level of
api.stackexchange.com and using
questions/[id]?site=codegolf to save another byte.
But that then got me to thinking: if we can require a specific domain then we could, by extension, be able to require a specific path, with the above example becoming simply
[id]/?site=codegolf if run from
Personally, I feel that it's pushing things too far but couldn't come up with a legitimate argument against allowing it so decided to post this question to garner some thoughts from the community and, hopefully, reach a consensus on it.