FAQ

Why are Nodes readonly?

Making nodes a readonly model allows separate optimizations for the read and write side. While the new CR already offers decent performance we have way more options now and will continue to improve. 

Why have Nodes no methods for traversal to access parents, children and siblings?

Nodes are meant to be lightweight and only store chunks of data. For traversal the Content Subgraph offers a plethora of methods that are type safe and extensively tested.

Why is PostgreSQL not supported?

The PostgreSQL Adapter is in the works and promised great performance but was not finished in time. We focused on the MySql/MariaDB adapters for the 9.0 release as those are clearly the most common ones.

Why is flowQuery find sometimes slow?

The Content Graph read model we now use offers efficient traversal between directly related nodes which is the 99% use case. While many FlowQuery operations have the same performance or are faster the performance of find is not optimal yet. We plan to improve that in future releases if it turns out to be an issue and cannot be avoided. The nice thing is that the separate read model allows us such optimizations in future.

Why are many of the new classes final? 

The Event Sourced Content Repository is all new and shiny. The final classes allow us to do changes that may be necessary in a non breaking way. If you run into issues with missing extensibility please contact us. We will surely add more extension points in future but only once we truly understand the needs.

Why are there only importAll and exportAll commands?

The import and export primary deal with the event stream that is our single source of truth. Because exporting parts of the events requires filtering and more logic we decided to first start with site:importAll / site:exportAll as those cover the cases where systems are upgraded from older versions. In future we plan to add site:import and site:export commands for single sites.

Where is the history module and undo functionality?

While the Event Sourcing already gives us the perfect audit trail and all events that happen in Neos 9.0 are recorded there is no interface to browse the history and undo changes yet. However with Neos 9.0 this has become a possibility that could even be implemented in a community package.

Why do i have to rewrite all Node related PHP code?

While we managed to support or migrate the fusion read side the nature of separating the read from the write side did not allow to do the same for PHP code without breaking the patterns we wanted to enforce.

However the PHP API is for the read and write side offers a much cleaner and easier to use interface. See Content Repository PHP API

Why were Node Signals removed?

Node Signals are replaced by custom Projections, Catch Up Hooks and Command Hooks that offer the same functionality with a cleaner api and reflect the separation of the read and write APIs.  

Why are Node Privileges removed?

Node Privileges that could be configured via Policy.yaml were removed for now to limit the already extensive scope of the Event Sourced Content Repository. We plan to add an overhauled version to future releases that will likely be based on Subtree Tags.

In the meantime you can implement custom privileges.