Roch Devost
Roch Devost
This is causing regressions compared to some optimizations we had in place before. I think this should either be reverted or we need to find how to apply those optimizations...
@BridgeAR Some of your comments around dependencies made me realize that by moving to linking, we lose our only safety net that guarantees that we're not using a dev dependency...
@scagood Did you ever find a workaround for this? We're hitting the same issue.
@BridgeAR Open to other suggestions if you have any.
I guess the logic is that the PR contains a semver major change, so it should be labeled as such. The fact that the change is hidden behind a flag...
@watson So you would keep `semver-minor` but with an added tag? Not a fan of `v6-only-changes` for a few reasons. First of all I don't think it's quite semantically correct...
> After the release of v6, we want to be able to land a PR containing changes both targeted v5 and v6. Isn't that just a normal `semver-patch` or `semver-minor`?
> To some extend yes, but that doesn't allow you to automatically know which release notes to update. Just the ones for v5, just the ones for v6, or both...
@BridgeAR Given how difficult that test file is to work with, and since we don't have access to nice helpers like `semifies` and the current use case has been only...
There is already an API to add exceptions to spans. How is this expected to interact with it? As a user, how to I differentiate between them? Do I need...