Francesco Casella

@ceraolo you should also test it. Much improved rendering of diagrams now.

@phannebohm if you can manage by Christmas (2021, I mean) it would be still good enough to get into 1.19.0. Possibly as a cherry-picked bugfix if we have already branched...

@phannebohm is on vacation already, so this is moved to 1.20.0

#9224 shows another reason why we should always put $cse variables (not only from records, btw) among the torn variables. @phannebohm do you think you can fix this?

Since selecting $cse variables as iteration variables often leads to convergence failure, I would say that any strategy that leads to convergence success is preferrable to that, even if it...

> Agreed! I'll do that now. Good 😃 > This may not be worth it. We would need to do the tearing, then causalize the torn set to see if...

Thanks @phannebohm! I'm curious to see if this has a positive impact on the library testing regression report. We'll see tomorrow.

The [regression report](https://libraries.openmodelica.org/branches/history/master/2023-01-24%2011:26:49..2023-01-25%2015:46:47.html) has a huge number of improvements because we added reference results to Buildings_9. However, there are a few other improvements, e.g., three more models simulating in Buildings_8....

@phannebohm we probably found one more case that still needs to be addressed. Please get the latest version of the Buildings library on master and check [Buildings.Air.Systems.SingleZone.VAV.Examples.OptimalStart.ConventionalSpring](https://libraries.openmodelica.org/branches/master/Buildings_latest/files/Buildings_latest_Buildings.Air.Systems.SingleZone.VAV.Examples.OptimalStart.ConventionalSpring.sim). The initialization fails...

Adding @Pavanrajhr to the loop