You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
If there's an asset for the current RID as determined by the host (so on windows it would be win10-x64) and there's no RID graph but it's instructed to read the RID graph, the asset should still resolve.
Currently the test never exercises this case, since the CurrentRid will be the current build target RID (so on windows it would be win-x64), which is currently always portable RID, and thus never the one the host determines on its own.
It might be that on source-built runtime, where the build target is non-portable, it could happen that the host determined RID and the build target RID matches, in which case the test would fail.
Maybe not for this change, but this test misses an interesting case.
If there's an asset for the current RID as determined by the host (so on windows it would be win10-x64) and there's no RID graph but it's instructed to read the RID graph, the asset should still resolve.
Currently the test never exercises this case, since the CurrentRid will be the current build target RID (so on windows it would be win-x64), which is currently always portable RID, and thus never the one the host determines on its own.
It might be that on source-built runtime, where the build target is non-portable, it could happen that the host determined RID and the build target RID matches, in which case the test would fail.
Originally posted by @vitek-karas in #84100 (comment)
The text was updated successfully, but these errors were encountered: