-
Hi everyone, I know that Sentry ships it's own Android Proguard/R8 rules. Most of them are defined in the sentry-android-core/proguard-rules.pro. Previously we added the rules recommended by Google: Required configuration when using R8, ProGuard, and DexGuard I see that Sentry already includes the first rule (so we can remove it in our own proguard rules file): But I am wondering if we should keep the following rule recommended by Google: I found out that this line was present in the old Sentry Gradle Plugin: https://github.com/getsentry/sentry-java/pull/790/files Any opinion or explanation why this rule is not present anymore? Thanks |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 2 replies
-
I guess this rule is necessary, because there's no server-side deobfuscation of the exception titles on Firebase. This is mostly relevant if you have custom exception types that you'd like to keep from being obfuscated (all default types like IllegalArgumentException are part of Android/JDK already). Sentry tries to deobfuscate this server-side but there are some rough edges, so sometimes we also have to provide rules like this ourselves:
|
Beta Was this translation helpful? Give feedback.
I guess this rule is necessary, because there's no server-side deobfuscation of the exception titles on Firebase. This is mostly relevant if you have custom exception types that you'd like to keep from being obfuscated (all default types like IllegalArgumentException are part of Android/JDK already). Sentry tries to deobfuscate this server-side but there are some rough edges, so sometimes we also have to provide rules like this ourselves:
-keepnames
should be enough in this case as we only are interested in the name of the class.