[07:29] <lullabud> i just noticed that my /etc/hosts file has "127.0.1.1  hostname" as an entry...  anybody know what's up with 127.0.1.1 ?
[07:29] <lullabud> it also has 127.0.0.1 localhost
[07:29] <lullabud> i've just never see the 127.0.1.1 entry before..
[08:10] <mralphabet> there are several mailing list discussions about it on google
[08:10] <mralphabet> just search for 127.0.1.1 and it will come up
[08:37] <lullabud> mralphabet:  yeah, i found a bunch of "bug #" type of posts.  i guess i was hoping for a technical explanation of exactly why they chose to do that...
[08:54] <mralphabet> the best explanation that I could see was "localhost should be the only thing that resolves to 127.0.01, everything else should move through xyz"
[08:54] <mralphabet> in this case, xyz = 127.0.1.1
[09:01] <lullabud> huh... interesting.  that seems like it's a good idea, actually.
[09:01] <lullabud> thanks mralphabet 
[09:02] <coNP> why is it a good idea?
[09:02] <coNP> I mean I think I missed the point
[09:07] <mralphabet> I believe, and others would be more qualified to answer then I, that if you have a machine that is on any sort of network, you do not want the machine name to resolve to 127.0.0.1 because the secondary machine may get the wrong answer and try to connect to 127.0.0.1, which would be . . . itself . . . it's preventing the snowball effect
[09:12] <lullabud> another thing is that it offers a way to locally test network changes, specifically firewall changes.
[09:13] <lullabud> if you allow connections from localhost as 127.0.0.1 but want to deny others, you could connect to localhost to test your lo, and [hostname]  to test outside access
[09:14] <lullabud> i know solaris uses one hostname per physical interface, so it would make sense that logical interfaces would also use that scheme.  but this is linux...
[09:23] <coNP> thanks
[11:24] <Spritz> anyone using ubuntu/sparc?