Firebase Crashlytics considers different crashes the same issue

[Alex]

In my opinion, this is a bug of the Web UI (or the underlying backend). The problem is that two different (from my perspective as the developer) bugs/crashes are considered the same because the very top of the backtrace is the same. This means that when I fix an issue and close it, it pops back up as a "regression" while in reality it's a new bug never seen before, and now when scrolling through the list of crash reports I need to keep track of which is for which of the two (or more) different issues, some of which have already been fixed earlier.

Here is the illustration - two clearly different crashes that are considered the same because in both cases it's SIGABRT in `std::recursive_mutex::lock()`.

https://imgur.com/a/p57xilS