Refine model code generation during build
The automated workflow for generating code from the TDL meta-model defaults to delegate-based operation implementation which can be slow and flaky.
There is no easy way to change that. Custom generator building on the one provided by Xtext may be needed.
issue