Crash reporting and RelWithDebInfo

[JulianGro]

Idea behind this PR is to bring back crash reporting, but in a way that yields actually useful crash reports.
In the past, we would collect crash reports, but without debug information and with the builds fully optimized. This would essentially only tell us that someone was crashing.
Now I want to create RelWithDebInfo builds, automatically upload the debug information to Sentry, and strip the debug symbols in the later packaging step.

With Dale's changes, this also switches us from using Crashpad directly to using Sentry-native which takes the maintenance burden of trying to build Google software away.

Conan recipe change: overte-org/overte-conan-recipes#59

I tested the Sentry uploading locally using act, since Pull Requests cannot usually access secrets.

The size difference between Release and RelWithDebInfo/Debug builds stems from Qt. Apparently, when building from source, the libraries Qt depends on are statically linked.

[JulianGro]

@vatral can we change some of Sentry-Native's behaviour?
For one, it appears to periodically check in with Sentry when Crash reporting is enabled, even when we haven't crashed. The user agrees to send crash reports automatically, not necessarily to telemetry.

[sentry] DEBUG sending envelope
[sentry] DEBUG submitting task to background worker thread
[sentry] DEBUG executing task on worker thread
* Reusing existing https: connection with host o4504831972343808.ingest.us.sentry.io
> POST /api/4504832427950080/envelope/ HTTP/1.1
Host: o4504831972343808.ingest.us.sentry.io
User-Agent: sentry.native/0.14.2
Accept: */*
x-sentry-auth:Sentry sentry_key=f511de295975461b8f92a36f4a4a4f32, sentry_version=7, sentry_client=sentry.native/0.14.2
content-type:application/x-sentry-envelope
content-length:1824

* upload completely sent off: 1824 bytes
< HTTP/1.1 200 OK
< server: nginx
< date: Mon, 31 Aug 2026 16:59:41 GMT
< content-type: application/json
< Content-Length: 2
< vary: origin, access-control-request-method, access-control-request-headers
< access-control-allow-origin: *
< access-control-expose-headers: x-sentry-error,x-sentry-rate-limits,retry-after
< cross-origin-resource-policy: cross-origin
< strict-transport-security: max-age=31536000; includeSubDomains; preload
< via: 1.1 google
< Alt-Svc: h3=":443"; ma=2592000
< 
{}* Connection #0 to host o4504831972343808.ingest.us.sentry.io:443 left intact

Then there are quite a lot of log messages coming from Sentry. I feel like it is too chatty.
It probably shouldn't print the backtrace URL either:

[sentry] DEBUG using minidump URL "https://o4504831972343808.ingest.us.sentry.io:443/api/4504832427950080/minidump/?sentry_client=sentry.native/0.14.2&sentry_key=f511de295975461b8f92a36f4a4a4f32"