At 11:56 PM +0900 4/8/02, JINMEI Tatuya / =?ISO-2022-JP?B?GyRCP0BMQEMjOkgbKEI=?= wrote: > >>>>> On Sat, 6 Apr 2002 09:28:39 -0800, > >>>>> Toerless Eckert <[EMAIL PROTECTED]> said: > > > Well, i've just seen bad and worse ip multicast applications over the > > last 10 years, programmers have been ignorant about the issues at hand > > and people who deploy the applications have been ignorant about the > > network issues they cause. > >Out of curiosity, do you have concrete examples of the bad >applications? I tend to agree with you, but I'm still not fully >convinced that application programmers are that silly and we need to >add the complicated intelligence to the kernel. > >I'd also like to hear from others. We cannot make a decision just >from the "local" conversation...
Jinmei-san, There is at least some precedent for having different default behavior for IP multicast than unicast, in order to reduce the chance of accidental "misuse" of multicast: in the IPv4 multicast socket support, we had TTL default to 1 for multicast, so that someone wanting to send multi-hop multicasts had to explicitly say so. I think Toerless's suggestion is in the same spirit, ensuring that multicast MTU discovery doesn't happen unless explicitly requested. I don't think it takes that much complicated intelligence in the kernel to have different defaults for multicast and unicast. It's not so much judging application programmers as "silly" as reminding them of consequences they may have forgotten or never realized. Steve -------------------------------------------------------------------- IETF IPng Working Group Mailing List IPng Home Page: http://playground.sun.com/ipng FTP archive: ftp://playground.sun.com/pub/ipng Direct all administrative requests to [EMAIL PROTECTED] --------------------------------------------------------------------
