No problem. Just making sure this was the expected behavior. I don't want to put in a bunch of ntoh* calls only to have to pull them out again when I next decide to sync ;-)
Glen Natasha Gude wrote: > Hey Glen, > > We keep C++ Flow objects in network byte order, and it is expected for > the Python SWIG one to be in network byte order as well. I didn't > realize the old one converted it to host order before sending it on to > Python - sorry about that confusion, but for now the Python version > has to match the C++ one because we pass the flow back to C++ in some > cases (thus the need to make it a SWIG object to begin with). > Natasha > > Glen Gibb wrote: >> Hi Natasha, >> >> I'm working on bringing the NOX we're using for mobilevms up to a >> slightly more recent version. The changes I pulled in included your >> changes to Flow objects so that they're SWIG objects instead of >> dictionaries. >> >> I'm noticing now that the nw_src and nw_dst attributes seem to be in >> network order not host order. Is this the expected behavior? I've >> forked from 77a260b181f67c98ba35d0fd77ad322462d43763 -- if there have >> been any patches to change this behavior since this commit can you >> let me know so that I can cherry pick it. >> >> Thanks, >> Glen > > _______________________________________________ nox-dev mailing list [email protected] http://noxrepo.org/mailman/listinfo/nox-dev_noxrepo.org
