Martin Ficzel

@hlubek Your opinion would be good on this aswell

I know we discussed this and ended up not doing this now as we were not sure wether this would a good ides to have as an api. We thought...

That is an accident … imho it would make sense for join to assume untyped keys to be join aswell

That inheritence mainly happened without much intention. Refactoring to traits would be fine, however i would also like to see a concept of defaultType for untyped nodes in fusion as...

Regarding this issue i wonder wether it would still help to configure wether a route/resolve should be cached and how long. Maybe we can add the foilowing settings to the...

@bwaidelich maybe we drop the case of disabling routing via false but still allow to set a cache lifetime. Because entries without lifetime are scary ``` - routeCacheLifetime: null|int(default null)...

> My concerns would still apply and I wonder about the usecase for this? Maybe I'm wrong but if you create cache entries from "unbound" values (i.e. external sources) cache...

Agree that appendExceedingArguments is probably a good idea with bad consequences. Since the uriBuilder as it is relies on it we need some form of patch. Maybe we should deprecate...

BTW: I tried to extract a common base class from the resource and package stream wrapper but this ended in a big mess because of many methods that are not...

@mhsdesign i think it is no problem that resource:// slightly overlaps with package:// as both can do specific things the other cannot. Would still recommend resource for all use cases...