I don't have a strong opinion in this matter, so I won't argue. I would suggest, however, to make the communication of the bearer more explicit. The current overloading of the bearer/network_required fields seems hackish to me.
Regards Jörg > -----Ursprüngliche Nachricht----- > Von: Aarno Syvänen [mailto:[EMAIL PROTECTED]] > Gesendet am: Donnerstag, 29. November 2001 12:11 > An: Jörg Pommnitz > Cc: '''[EMAIL PROTECTED]' ' ' > Betreff: Re: recent WAPPush with IPv4 bearer crashes > > Hi Jörg, > > Jörg Pommnitz wrote: > > > > I think the bearer selection needs some thoughts. IMHO > > making this decision in wap_push_ppg is a slight layering > > violation (debatable). IMHO this decision should be made > > in dispatch_datagram(WAPEvent *dgram) in wapbox.c. One problem: > > at this point we lost the information about the type of the > > address. I think this should be made part of the WAPEvent. > > > > What do you think? > > I agree, in principle. However, PPG updates PAP attribute after > it has found an unssupported bearer. (And PI can query it if it > wants.) I think it is better to have all functions manipulating > PAP attribute in same module (data cohesion of modules is one > architectural principles of Kannel.) > > Aarno >