On Mon, May 20, 2019 at 03:56:14PM +0200, Ævar Arnfjörð Bjarmason wrote: > > -test_expect_failure 'fetch of password-URL clone uses stored auth' ' > > +test_expect_success 'fetch of password-URL clone uses stored auth' ' > > set_askpass wrong && > > git -C clone-auth-none fetch && > > expect_askpass none > > I've only looked at this very briefly, there's a regression here where > you're assuming that having a configured credential helper means it > works. > > I.e. I have a ~/.gitconfig where I point to some-gnome-thing-or-other > what doesn't exist on my VPS in my ~/.gitconfig, cloning just warns > about it being missing, but will store the password in the repo. > > With this you detect that I have the helper, don't store it, but then my > helper doesn't work, whereas this worked before. There are more cases beyond that, too. You might have a helper defined which doesn't actually store passwords, but just sometimes tries to provide one. My thinking was that if you're clueful enough to have configured helpers, you can probably deal with the fallout. But you're right that it may still be a regression in the sense that the user may still have to actually _do_ something to get their fetch to work. I guess a more robust version of this is that _after_ the successful clone, we could ask the credential system "hey, do you have the credential for $URL?". And if it can't answer, then we can take action (whether that action is setting up credential-store and seeding it with the password, or just advising the user about the situation). -Peff
