Peter Gafert

Yes, back then I decided to only import local file URIs, because I didn't think anybody would have the use case to import http URIs :wink: Also because I thought...

Thanks for raising this issue! I'd like the docs to be as helpful as possible. My original hope was that the concept described [here](https://www.archunit.org/userguide/html/000_Index.html#_the_lang_api) plus the [ArchUnit Examples](https://github.com/TNG/ArchUnit-Examples) would provide...

Very good points @rweisleder :smiley: The only thing about the central entry point for all predicates / conditions is that I wonder if it would clash somehow :thinking: But in...

On the other hand, `HasName.Predicates` has the advantage that you can apply those to all things with names :thinking: If we have `Predicates.ForClasses....`, etc., we have to duplicate that entrypoint...

Not a full-fledged solution, but I added a lot more docs with links in #912. That should at least improve the discoverability a little by pointing from places that take...

This already came up recently: https://github.com/TNG/ArchUnit/issues/81 I do agree, that there is little value in checking synthetic methods for architecture tests, I'm just not completely sure, at which level this...

Upon closer investigation I have my doubts, that this concrete case is really solved by excluding synthetic members. `Enum.values()` does not count as synthetic (even though the compiler creates it,...

But the access that causes you the problem, is ``` ClassesShouldTest$Foo.values() -> [ClassesShouldTest$Foo;.clone() ``` not the access to the synthetic field `$VALUES` (which is a member of `Foo`). So the...

The information that in this case the method is called on an instance of a synthetic field is definitely not that easily available (even though it is certainly possible). We...

Yes, so I guess that would mean some heavier refactoring, where some sort of 'context' or 'history' object is dragged along, so that one can see, if the last stack...