it was anyway dysfunctional and mainly slightly reducing performance.

Just use enableDebugDataLoss() to enable debug output.

This reverts commit 4f2b35be.

Was actually used and tested.

It already exists for ArrayAccessor::readAsync()...

ScalarPollInput and ArrayPollInput: Removed doReadTransfer, which does not make sense in the interface

added missing implementations of writeDestructively() - the inherited ones from the DeviceAccess classes are not doing the right thing...

bidirectional variables: pass on version numbers in modules from the last blocking read on to any write operations, to resolve conflicts of concurrent variable updates properly

Added preliminary support for bidirectional variables. There are still some known issues, e.g. concurrent updates can still lead to race conditions. This will be addressed soon.

added move-assignment operators for the accessors, which is required to use move-assignment operators of modules containing accessors

refactor ArrayAccessor and ScalarAccessor: split off common code and put into common base class InversionOfControlAccessor

allow determining a fully qualified name for application variables (i.e. including qualified name of owning module)

use the new readLatest(), which has been added to the latest DeviceAccess version, to implement the naturally expected behaviour when using a push-type variable with a poll-type accessor

added preliminary support for tags to application variables which can be searched for and used to generate virtual modules

fixed replacing Array and ScalarAccessors, which was broken after moving to the NDRegisterAccessorBridge-based accessors

fixed poll-type inputs using no longer non-blocking read after the switch to the NDRegisterAccessorBridge-based accessors