Dietmar Warning

Tales are nice, but tales don't compile. If this ghost of SPICE past is worth resurrecting, how about sketching the actual algorithm instead of just the eulogy? Even a napkin-level outline of the mismatch method would beat another round of nostalgia for Brand X's crypt.

VBIC: let xrb/xrc fall back to the extrinsic resistances as well

VBIC: map xrb/xrc to the intrinsic resistances

Hijacking this thread may follow in some confusing. But anyway - there is no instance parameter for beta, only currents cb and cc can accessed. So the good old method by using the current in voltage sources is an adequate approach as shown in the attachement. Mismatch for bjt is another problem.

VBIC: delete excess-phase nodes properly on re-setup

DIO: fix crash when a model has rth0 but the instance is not thermal

Arpad, to prevent overlapping: Was my commit 2ccc0e Revert "omit copy action - work with snp2va.cpp" from 2026-09-05 20:28:18 in your repository?

osdi: free the absdelay buffers in OSDIunsetup