The problem
I want the logs during unit testing to be mostly silent, so that when a message is printed it is a salient signal that there might be something worth looking at. Right now, the unit tests spit out quite a lot of mostly noisy information, which has and can continue to lead to missed signals about issues.
The solution
Fix or disable the source of the noisy logs emitted during testing.
Arguably we should run Lit in production mode during testing, as it will stand a chance of catching more issues than if we run using dev mode. If we deem dev mode appropriate, we should filter out this log.
This is likely due to Yarn resolutions not using an identical (major?) version of Lit for everything, which may get fixed naturally by #634.
Element test-pharos-* scheduled an update (generally because a property was set) after an update completed, causing a new update to be scheduled. This is inefficient and should be avoided unless the next update can only be scheduled as a side effect of the previous update. See https://lit.dev/msg/change-in-update for more information.
This one is a trickier component-level issue for several components, and we would potentially get mild performance benefits by fixing this at the source in addition to removing the log noise.
🚧 404 network requests on Webkit:
lib/styles/icons/fake.js
This is an expected error, as many tests use a fake icon which is not expected to exist. We should take care when silencing this only to silence the fake icons, lest we start silently failing to get real icons at some point.
The problem I want the logs during unit testing to be mostly silent, so that when a message is printed it is a salient signal that there might be something worth looking at. Right now, the unit tests spit out quite a lot of mostly noisy information, which has and can continue to lead to missed signals about issues.
The solution Fix or disable the source of the noisy logs emitted during testing.
Additional information
Most of the logs are one of four things:
Arguably we should run Lit in production mode during testing, as it will stand a chance of catching more issues than if we run using dev mode. If we deem dev mode appropriate, we should filter out this log.
This is likely due to Yarn resolutions not using an identical (major?) version of Lit for everything, which may get fixed naturally by #634.
This one is a trickier component-level issue for several components, and we would potentially get mild performance benefits by fixing this at the source in addition to removing the log noise.
This is an expected error, as many tests use a fake icon which is not expected to exist. We should take care when silencing this only to silence the fake icons, lest we start silently failing to get real icons at some point.