We use sourceforge's bug tracking tools. SOSlib bugs can be browsed and
reported there.
Also choose `Closed' in the `Status' menu to look for bugs that have
been fixed in the current development version. You might get hints there
how to solve the problem.
For some serious bugs we provide patches. See below.
If you still have problems see our support
site and follow instructions there.
WRONG RESULTS for variable compartments, i.e. assignment or rate rules for compartments don't work
Units of a species in all equations are usually in concentrations, unless the species flag "HasOnlySubstanceUnits" is set to true or the species' compartment's "SpatialDimension" is set to 0. Models with such species are handled wrongly in 1.6.0.
WRONG RESULTS for missing values in SBML. Values not given in SBML are interpreted as 0 (species, parameters) or 1 (compartments). Bug describtion and a patch file can be found here.
Sensitivities were initialized wrongly, which should be OK for a single run (if the compiler cared), but WILL PRODUCE WRONG results for iterative runs with the same solver instance. No patch file availabe, please use CVS version if you need (iterative runs of) sensitivity analysis or contact the authors (see support site) for instructions.
WRONG RESULTS for piecewise function. Piecewise function was implemented wrongly according to wrong piecewise expressions in the first SBML Semantic Test Suite. No patch available, please use CVS version if you have models with piecewise functions or contact the authors (see support site) for instructions.