From Francis.Dupont at inria.fr Tue Jan 11 13:53:47 1994 From: Francis.Dupont at inria.fr (Francis Dupont) Date: Tue, 11 Jan 1994 13:53:47 +0100 Subject: DNS WG at the 17th RIPE meeting Message-ID: <199401111253.AA05510@givry.inria.fr> I am collecting the ideas for the DNS WG agenda... First we have the still open points: - a second root name server in Europe (nothing new ?) - IPng support (new proposal and code for TUBA/CATNIP/NSAP, perhaps something for SIPP ?) - last BIND software (version 4.9.2 is in final beta-test) - DNS for Windows NT (I know only one thing: WNT has a resolver :-) New points: - mandatory field zone-c RIPE DB problem (cf Marten's E-mail) You can propose elements for the agenda (before next week please). Thanks Francis.Dupont at inria.fr From Francis.Dupont at inria.fr Wed Jan 19 12:50:28 1994 From: Francis.Dupont at inria.fr (Francis Dupont) Date: Wed, 19 Jan 1994 12:50:28 +0100 Subject: DNS WG at the 17th RIPE meeting Message-ID: <199401191150.AA08817@givry.inria.fr> Here is the final draft agenda for the DNS WG meeting : First we have the still open points: - a second root name server in Europe (nothing new ?) - IPng support (new proposal and code for TUBA/CATNIP/NSAP, perhaps something for SIPP ?) - last BIND software (version 4.9.2 is in final beta-test) - DNS for Windows NT (I know only one thing: WNT has a resolver :-) New points: - mandatory field zone-c RIPE DB problem (cf Marten's E-mail) We need 1.0 hour. Thanks Francis.Dupont at inria.fr From erik-jan.bos at SURFnet.nl Wed Jan 26 20:54:02 1994 From: erik-jan.bos at SURFnet.nl (Erik-Jan Bos) Date: Wed, 26 Jan 1994 20:54:02 +0100 Subject: FRWD: Re: Root nameservers... Message-ID: <"survis.sur.348:26.00.94.19.54.04"@surfnet.nl> DNS-WG, I just found the message below in News. Input for the next DNS-WG meeting? --- Start of News article From vixie at vix.com Tue Jan 25 08:37:33 1994 From: vixie at vix.com (Paul A Vixie) Date: 25 Jan 94 08:37:33 Subject: Root nameservers... Message-ID: One of the reasons why putting root servers outside the US has been avoided is that the most common DNS implementation -- BIND -- has for a long time had bugs such that it wouldn't use "close" servers over "distant" ones. Since the U.S. was the center of most of the world's Internet connectivity (and still is) it was easier to just put the root servers where the traffic ``probably had to go anyway most of the time'' than to fix BIND. BIND is fixed. I think some of the math is still "wrong", since I didn't pick up all of the patches I got from Phil Almquist in this area and I know he spent a whole lot of time on it. But I did fix a lot of the sorting behaviour in 4.9 (and a few more things will be fixed in 4.9.2). If all of the root servers that run BIND use 4.9 or later, and most of the large institutional forwarding servers that run BIND use 4.9 or later, it will no longer matter where the root servers are. Note: I'm just another BIND hacker, I don't make decisions about root servers. But if those who *do* make these decisions want to spread the root service load a bit more, BIND is finally able to do the right thing about it. -- Paul Vixie Redwood City, CA Also: , , decwrl!vixie!paul , , <{bind-workers,objectivism}-request at vix.com> --- End of News article __ Erik-Jan.