Gary Nyquist wrote on 09/30/2014 07:21 PM: > Hi, > > I need to build an "ad hoc conferencing system" (where the calls are > dialed-out from the system) using SEMS. > Have figured out how to implement it using DSM script. > Wondering if it will be substantially better if the same functionality is > writen as a C++ application. > Basically, want to know the relative performances of the "DSM version" > vis-a-vis "C++ version". > Does anyone have a ballpark figure? if you are doing conferencing, processing of the audio in the conference takes generally much more than processing of the signaling, where DSM is involved. just run some tests with e.g. valgrind, and you'll notice it's only a few % more.
Stefan _______________________________________________ Sems mailing list [email protected] http://lists.iptel.org/mailman/listinfo/sems
