Marc Henry's part / requirements:

In order to get the NodeIdentity/NodeAdress change finished (see #4868) we must rewrite the write sides projection access. Because of the dilemma here: #4868 (comment)
(The dilemma is that the Node read model has to know its WorkspaceName somehow)

We discussed that the high level api of the content graph will eventually look like findNodeAggregateById(WorkspaceName $ws, $id); and also getSubgraph will accept the WorkspaceName instead of a content stream id.

So why can we not just always work with WorkspaceName's also on the write side?
Because the first thing we do is resolve the current content stream we want to operate on via requireContentStreamId which might return, if in a rebase closure, a globally temporary mock instead of the actual one for the WorkspaceName.

Thus we need to have a low level api where we not only pass the ContentStreamId but also the WorkspaceName.

The WorkspaceName is currently not yet part of any node / node aggregate read model, but will be with the introduction of the NodeIdentity/NodeAdress ON the node. So this part has to be forward implemented without being used by the NodeFactory ... that change is my part.

During this the subtask #4942 also needs to be solved, as the 2 commands dont contain WorkspaceName's yet and we would need them to pass to the lower level methods.

Feel free to edit this issue and add your insights.

Slack threads !!!:

• https://neos-project.slack.com/archives/C04PYL8H3/p1711793524319239?thread_ts=1711791103.956609&cid=C04PYL8H3
• https://neos-project.slack.com/archives/C04PYL8H3/p1711349027838899
• https://neos-project.slack.com/archives/C04PYL8H3/p1711111681225429