Arpad Borsos

> besides a suspect on breakpad own `` header I don’t think it brings its own string type. It is indeed super weird. Also, why do you get `"\034\222\002"` for...

@michirod have you continued your investigation? It is super weird that this ends up in a c++ std function. Maybe some kind of ABI mismatch problem? We do see some...

The problem with iOS and Android breakpad right now is that neither of them runs our integration test suite. However I have integrated the iOS part into a swift application...

This does look like a `capture_event`, right? That does not automatically attach a stacktrace. You would have to use `sentry_event_value_add_stacktrace` for that. https://github.com/getsentry/sentry-native/blob/1b151d10fdb4b1fa2e4440cff4a436d6186fc862/include/sentry.h#L373-L380 The more interesting bit however would be...

The limits imposed on envelopes should be quite sufficient actually: https://develop.sentry.dev/sdk/envelopes/#size-limits We do have some customers that hit those, but they usually deal with a large number of threads.

You have to use `MACOSX_DEPLOYMENT_TARGET` environment var, as explained in https://cmake.org/cmake/help/latest/variable/CMAKE_OSX_DEPLOYMENT_TARGET.html The docs say the attribute is names as `MACOSX` even if you are dealing with ios. Also see https://stackoverflow.com/a/34208904...

> Is there a way to debug this while being attached to Xcode? Or does that cause side effects that make testing it difficult? When I was testing, the debugger...

Android indeed is weird related to a couple of things: * Because of the diversity, we simply don’t have debug info for a lot of the symbols. * Also, some...

It would be interesting to see the output of `/proc/XXX/maps` related to ` libUE4.so`, I wonder why that is split in two.

If its from a device you control, connect via `adb` and copy that for the process in question to a file. From within the process, you can read from `/proc/self/maps`....