Gáspár Nagy

@trampster could you please send a PR for the updated template?

@304NotModified is this ready for review?

It might be related to the [different async behavior](https://docs.reqnroll.net/latest/guides/migrating-from-specflow.html#complete-changelog-of-specflow-v4) of Reqnroll (and SpecFlow v4). I would suggest to contact https://github.com/FlaUI/FlaUI, they have probably better ideas what special async behavior is...

Closing it as out of scope, but please reopen if there is anything received from FlaUI that we can do.

Did a research: also could not find a generic way to access the output folder (shame...). So I was thinking about this formatter implementation. It could be a pretty good...

> Another thought- I agree that the logger should set a distinct environment variable. Should that value overwrite the configuration for the respective formatter OR be in addition to it?...

> Let's clarify our rules and order of precedence. > > 1. We treat the configuration of each formatter separately. > 2. As we configure a formatter, a configuration provider...

@clrudolphi sorry for not responding earlier. > As the first portion of refactoring this, commit [f99c7d3](https://github.com/reqnroll/Reqnroll/commit/f99c7d3bbf05206c9d20e425b5287b42ebf0b34d) creates two named loggers (one for messages and one for html). Usage: `--logger "formatter-html;outputFilePath=foo.html"...

The failing build is actually just an indicator that this change is breaking for an "average" plugin that uses the AST. Please ignore that test like we did in https://github.com/reqnroll/Reqnroll/pull/233/commits/da0214222a6df161549d13c8b567332683baccab...

This would simplify the things a lot and maybe this is the way to go. The only thing that comes to my mind is when someone uses a generator that...