Double check if craft is already able to auto create the first entry in release registry

@adinauer @kamilogorek what's the outcome of that problem, do we still need to do the release registry dance manually or is it fine now?
getsentry/sentry-release-registry#85

Yes, it's still "broken"

What's the reason to not just automatically enable it for all file io?

You can do that with the following code:

IOOverrides.global = SentryIoOverrides(HubAdapter(), options);

File f = File('path/to/file'); // file is SentryFile by default

class SentryIoOverrides extends IOOverrides {
  final Hub _hub;
  final SentryOptions _options;

  SentryIoOverrides(this._hub, this._options);

  @override
  File createFile(String path) {
    if (!_options.isTracingEnabled()) {
      return super.createFile(path);
    }
    return SentryFile(
      super.createFile(path),
      _hub,
      _options,
    );
  }
}

I did see that and it is indeed awesome, but its also misusing the mocks API, as an SDK, we need to be a bit more conservative about that, I guess.
I'm afraid that this could break the app's behavior that relies on the File type or something like that, or, that some files take more time than usual and you don't want to trace them for some reason (overhead, long transactions, etc), an all in solution could be a deal breaker for some users.
I'd prefer to do this next after gathering some feedback on the current implementation.
What do you think?

Yeah, that sounds reasonable

I was mostly asking because IOOverrides allow you to also measure the performance, if you don't otherwise control the file io yourself. Like for example with Hive, sembast or similar.

Yep, I believe it still makes sense to have it, its a hook that makes it possible since Dart does not have reflection nor monkey patching, not sure if it'd be ever enabled by default tho.
At least not during the first version.

L: What if we are overriding the content of the file do we want to know the length before or after?
If I'm looking correctly at the moment we would get the length before.

The data can only be read before, because after, the file can be deleted, renamed, moved, so it's a trade-off we made.
Maybe I should point that out in the docs.