Francesco Casella

The latest [error report](https://libraries.openmodelica.org/branches/heavy_tests/ScalableTestGrids_NB_SymbolicJacobian/files/ScalableTestGrids_NB_SymbolicJacobian_ScalableTestGrids.Models.Type1.Type1_N_1_M_1.err) now shows a failure in RemoveDummies, `NBStrongComponent.makeDAEModeResidualTraverse`: ``` Notification: Performance of Initialization: time 2.124e-06/1.055, allocations: 0 / 1.583 GB, free: 97.36 MB / 1.108 GB Error:...

@kabdelhak is working on a partial re-implementation of DAE mode in the NB, that should hopefully fix this problem.

I did some analysis, the NB does not use tearing and alias elimination, so the initialization system is much larger and apparently converges to the wrong solution. Alias elimination shouldn't...

OK. Are there plans to port tearing algorithms to the NB?

OK. Please also consider implementing the pruning of non-observable variables and equations. That should be easy, following the ideas of my [old paper](http://www.ep.liu.se/ecp/063/088/ecp11063088.pdf). We have something like it in the...

The latest [regression report](https://libraries.openmodelica.org/branches/history/newInst-newBackend/2025-03-28%2004:37:35..2025-03-30%2004:25,00.html) shows improvements on this front. Maybe because of improved homotopy? Now `PowerGrids.Electrical.Test.SynchronousMachine4Windings` passes verification with flying colours, as well as `PowerGrids.Electrical.Test.SynchronousMachine4WindingsAccurate` and `PowerGrids.Electrical.Test.SynchronousMachine4WindingsExact` The three NoLoad...

I include @kabdelhak, @phannebohm, @niklwors, @adriguir and @jph-tavella, who may be interested.

This is a good plan, but it requires some significant resources. We could try to get funded by LBL next year, to improve support of Buildings. My idea was to...

> It might already help to clean up the solver options / reorganize OMEdit menus to highlight the most important choices and seperate them from expert / experimental options --...

Lots of interesting findings while experimenting with the cpp runtime [here](https://github.com/OpenModelica/OpenModelica/pull/7689#issuecomment-881840567) (#7689)