> The first server to reply is used, others DHCPOFFER paquet(s) > are discarded > after. > Great, thanks. Both you and Peder agree so I'll do that. I suppose the spec probably said to do that if I dug in deep enough. > I can't tell for R5's net_server, but I saw BONE syslog > messages reporting > duplicates DHCPOFFER rejected packets. > Interesting, I'll have to remember to log that info. > Do you have already a plan on OBOS DHCP implementation design? > If yes, could you elaborate here (userland? stack protocol module?) > If not, maybe we can give you some hint. > > Feel free to post any question. > No, I have no plan on integration. My goals right now are simply to get an executable binary that can communicate with a DHCP server correctly. I am happy if you guys just tell me how to integrate it when the time comes and I'll do what you tell me. ---=+=--- In other more general news, I have the week of May 10-14 scheduled as vacation and I plan to split my time between this DHCP implementation and my other open source project which is a game. (Don't worry, when the game is released there will be a BeOS client.) So I aim to have evidence of some success or spectacular failure in two weeks. BTW, do you have any plans to mark DHCP as "assigned" on the team page? :-) --David