Improve tests by checking that the entire sequence ran #2

New issue

Open

opened 2016-05-17 15:32:07 +12:00 by mikedilger · 0 comments

mikedilger commented

2016-05-17 15:32:07 +12:00

(Migrated from github.com)

Right now, the tests just look at the final HTTP status code. But we can get false positives if the sequence failed to run and we happen to get the same status code.

The plan is to store in the client and server bits that we turn on when their various bits of code have run. The test then needs to check those bits are all true.

Right now, the tests just look at the final HTTP status code. But we can get false positives if the sequence failed to run and we happen to get the same status code. The plan is to store in the client and server bits that we turn on when their various bits of code have run. The test then needs to check those bits are all true.