Closed ace-n closed 6 years ago
Thanks! Published as 0.9.2
It is great that this landed, but there is no CI in this repo, and the changes over in google-auth-library have been quite substantial. It would be good to add a CI, or possibly move the functionality from here to google-auth-library itself. /cc @JustinBeckwith
I ran locally before publishing. I have an issue to open CI here: https://github.com/stephenplusplus/google-auto-auth/issues/32
But, generally agree: if google-auth-library can make auth easier, like this library, then we should blow this one up.
(A discussion on this should probably happen in a different issue, depending on the direction you want to take: "let's move functionality" vs "we need CI")
I think a CI should exist regardless. There is already an issue open for it: https://github.com/stephenplusplus/google-auto-auth/issues/32
@ace-n @ofrobots I had to revert. Would anyone be willing to send another PR, given the changes noted in #35? As I recall, I initially tested against the 1.0.0 pre-release, and things worked fine without changes. It would probably be quicker for someone more familiar with 1.2.0 to take a shot at this.
That might be something for @ofrobots to look into - as I'm not very familiar with these libraries, and simply massaged them until npm test
passed. (Another reason why we should have CI + solid testing, as @ofrobots mentioned above.)
We all want CI, we just need to get around to adding it :)
RE: solid testing, we could definitely use some real integration tests.
As title.
The primary motive is to to pull in a
google-auth-library
bugfix for Container Builder, which is integrated in the new release ofgoogle-auth-library
.Note: ideally, this would be a patch or minor version bump. If this change involves a major version bump, we'll have to update any dependents in order to fix the Container Builder issues.