Use "true" for force_update in relevant install_from_dist calls in config.rs#2074
Conversation
af62fcd to
6d26dfa
Compare
kinnison
left a comment
There was a problem hiding this comment.
The implementation is good; but are you certain that the tests fail if you don't have the src/config.rs change?
|
@kinnison, I've done some local testing using latest Below seems to validate that Please let me know if there are other ways you'd like me to test to validate more. Thanks. `master` version failed the downloadPR version installed
|
|
Right, but do your test cases behave this way -- an easy way to check that is to simply temporarily change the trues back to falses and run those new tests -- if they fail with |
6d26dfa to
71d1e72
Compare
@kinnison, you were right about that the previous tests were inefficient that there was no difference when reverting the booleans. So I rewrote the test w. I also confirmed that test would fail before the boolean change, and then with the boolean change the test passed as expected. PTAL? Test failed before the boolean change in config.rsTest passed after the boolean change in config.rs |
71d1e72 to
f11e7f1
Compare
kinnison
left a comment
There was a problem hiding this comment.
If you can fix the test case, I think this looks good. Nice job.
|
Since I had some time, I made the change for you and pushed a commit. if you're okay with it, and it passes tests, then please fold it into your commit appropriately and then we'll be good to merge. |
09a24c5 to
0800f41
Compare
@kinnison, ah my bad. Didn't use wildcard w. P.S.: Thanks for your comment that led to the learning about |
This PR resolves #2068 by updating the
force_updateparam fromtofalsetruein relevantinstall_from_dist()calls in config.rs, and provides two tests to mimic the users cases where theseinstall_from_distcalls would occur. (Could update the tests better under review guidance. Thanks in advance.)