Konstantin Ryabitsev <konstantin@xxxxxxxxxxxxxxxxxxx> writes: > On Thu, Nov 15, 2018 at 07:54:32PM +0100, Ævar Arnfjörð Bjarmason wrote: >> > I think that if we use the "principle of least surprise," insteadOf >> > rules shouldn't be applied for git-request-pull URLs. >> >> I haven't used request-pull so I don't have much of an opinion on this, >> but do you think the same applies to 'git remote get-url <remote>'? >> >> I.e. should it also show the original unmunged URL, or the munged one as >> it does now? > > I don't know, maybe both? As opposed to git-request-pull, this is not > exposing the insteadOf URL to someone other than the person who set it > up, so even if it does return the munged URL, it wouldn't be unexpected. Yeah, I think the local "git remote" should show the rewritten URL (because that information is not available otherwise) and it is OK to optionally show the original. Also, I think it would be nice if request-pull gave external-facing names. After all, insteadOf address is for your own use (e.g. maybe pointing at corporate mirror), and not for the intended recipient of request-pull. In short, I think I agree with things you are saying in this exchange.
