After issue #15 (the native bounding methods dispatcher) is in place, the MPA RTC packages and classes could actually drop the *_PwAffine suffix. However, such generic curves in a non-PwAffine shape cannot be used with the PMOO analysis. This is a restriction of the theory, not the NetCal DNC implementation.
For usage with the PMOO analysis, there is the decomposition into token buckets or rate latencies in Curve_MPARTC_PwAffine. Implementing the attempt to do so with generic curves is a larger but not necessarily useful task. The current solution to ensure PwAffine curves from the start is more straight-forward. Then, for the generic curves, the PMOO analysis simply cannot be executed.
Overall a more advanced restructuring is required. It should result in generic classes and classes for creating curves that are restricted to the piecewise affine shape. This class should be closed w.r.t. the min-plus operations. TODO: Check Anne Bouillard's "toolbox"-paper for this.
After issue #15 (the native bounding methods dispatcher) is in place, the MPA RTC packages and classes could actually drop the
*_PwAffinesuffix. However, such generic curves in a non-PwAffineshape cannot be used with the PMOO analysis. This is a restriction of the theory, not the NetCal DNC implementation.For usage with the PMOO analysis, there is the decomposition into token buckets or rate latencies in
Curve_MPARTC_PwAffine. Implementing the attempt to do so with generic curves is a larger but not necessarily useful task. The current solution to ensurePwAffinecurves from the start is more straight-forward. Then, for the generic curves, the PMOO analysis simply cannot be executed.Overall a more advanced restructuring is required. It should result in generic classes and classes for creating curves that are restricted to the piecewise affine shape. This class should be closed w.r.t. the min-plus operations. TODO: Check Anne Bouillard's "toolbox"-paper for this.