[00:06:30] --- fujiwara has left
[01:24:08] --- dthaler has joined
[01:25:37] --- dthaler has left: Replaced by new connection
[02:32:13] --- gas has joined
[02:33:01] --- gas has left
[19:49:56] --- yangwoo has joined
[19:54:43] --- frodek has joined
[19:58:09] --- mrex has joined
[19:59:29] --- tlyu@jis.mit.edu has joined
[20:01:50] --- john.zhao has joined
[20:02:14] --- ruri_hiromi@chat.gizmoproject.com has joined
[20:03:43] --- petithug has joined
[20:04:04] --- petithug has left
[20:05:31] --- arifumi has joined
[20:07:05] --- fneves has joined
[20:08:14] --- taiji-k has joined
[20:08:59] --- raeburn@mit.edu has joined
[20:09:34] --- bhoeneis has joined
[20:12:07] --- ruri_hiromi@chat.gizmoproject.com has left
[20:15:11] --- mohsen has joined
[20:15:27] --- slm has joined
[20:16:49] --- slm has left
[20:22:04] --- franck.martin has joined
[20:22:07] --- Simon Leinen has joined
[20:28:48] --- fujiwara has joined
[20:31:27] --- Simon Leinen has left
[20:33:27] --- iljitsch has joined
[20:34:14] --- nm has joined
[20:36:06] <franck.martin> there was a question yesterday about can you use code in RFC for an implementation...
[20:41:08] --- eric has joined
[20:41:29] --- weiler has joined
[20:43:18] --- Simon Leinen has joined
[20:47:59] --- dthaler has joined
[20:48:32] --- nm has left: Replaced by new connection
[20:48:33] --- nm has joined
[20:50:57] --- ryu has joined
[20:52:07] --- john.zhao has left: Computer went to sleep
[20:53:43] --- raeburn@mit.edu has left
[20:54:52] --- eric has left
[20:55:55] --- eric has joined
[20:57:33] --- john.zhao has joined
[20:59:31] --- peetu has joined
[21:01:32] --- TJ has joined
[21:08:11] --- tomp has joined
[21:08:53] <franck.martin> with all the probe packets happening generaly, most likely to wake up the PC often?
[21:10:17] --- weiler has left
[21:11:53] <iljitsch> you would probably want to set some rules for what gets to wake you up
[21:12:30] <iljitsch> I recently looked at wake-on-lan which you can do across the internet with one little problem: you need directed broadcasts. these are almost never supported
[21:13:58] <franck.martin> iljitsch: a psp can wake up a ps3 but not sure how it works...
[21:14:35] <franck.martin> and they faced false wake up that they fixed
[21:16:17] <iljitsch> franck: see http://arstechnica.com/journals/apple.ars/2007/09/21/wake-on-lan-dont-let-sleeping-macs-lie [http://arstechnica.com/journals/apple.ars/2007/09/21/wake-on-lan-dont-let-sleeping-macs-lie]
[21:16:49] <dthaler> wake-on-lan for PCs doesn't need directed broadcasts
[21:17:00] --- yangwoo has left: Computer went to sleep
[21:17:17] <iljitsch> dthaler: then how do you make the NIC see the packet if it can't respond to ARPs?
[21:17:47] <dthaler> cards support a fixed number of patterns. One of the patterns is for arp with your address in it.
[21:18:12] <iljitsch> so the card implements ARP?
[21:18:15] <dthaler> (or rather, one of the patterns the OS sets, it's not part of the firmware or hardware)
[21:18:22] <dthaler> no, the card just implements pattern matching
[21:18:35] --- nm has left
[21:18:46] --- eric has left
[21:18:52] --- john.zhao has left: Computer went to sleep
[21:18:52] <dthaler> Bruce is proposing more than wake-on-lan, where the card DOES support arp
[21:19:01] <iljitsch> the trick is generating a broadcast packet with the necessary pattern (to wake up the computer) in it from over the internet
[21:19:24] <dthaler> the trick is generating a packet (not broadcast) with a patterm that will wake the computer
[21:19:40] <dthaler> sending a packet to the machine's address results in the router arp'ing which wakes the machine
[21:20:02] <iljitsch> when the ip address is gone from the arp cache of the router broadcasts are the only packets that will do this
[21:20:08] <dthaler> another pattern is directed MAC (so if the router has an arp cache entry and just forwards, it still wakes the machine)
[21:20:35] <dthaler> routers generate arps, so the remote end need not do broadcasts or know the machine is asleep
[21:21:56] <franck.martin> dthaler: easy but the machine will be always awake due to the "noise" on the Internet?
[21:22:22] <iljitsch> I'm not sure if you're talking about what's deployed today. I haven't been able to find any way to do this without generating some knd of broadcast explicitly
[21:23:16] --- ryu has left: Replaced by new connection
[21:23:34] --- ryu has joined
[21:24:54] --- nm has joined
[21:26:50] --- yangwoo has joined
[21:29:50] <yangwoo> My laptop with Vista wakes up in a few seconds after going to sleep if LAN is connected and WoL is on.
[21:30:00] <yangwoo> There seems problem in "patterns".
[21:31:49] --- john.zhao has joined
[21:32:24] <franck.martin> is Lakshminath here?
[21:33:27] --- weiler has joined
[21:33:34] --- nm has left
[21:35:57] --- jimsch1 has joined
[21:36:08] --- tomp has left
[21:37:44] --- weiler has left
[21:38:40] <franck.martin> seems to me that WoL needs to be in rl2n
[21:40:16] --- dthaler has left
[21:42:22] --- peetu has left
[21:52:22] <franck.martin> is SIP a successful protocol?
[21:53:55] --- arifumi has left: Disconnected
[21:54:00] <TJ> It is extremely successful in terms of the # of pages of drafts generated.
[22:00:19] <franck.martin> but what I know it is more successful than h323 ;)
[22:02:26] --- bhoeneis has left: Replaced by new connection
[22:03:13] --- bhoeneis has joined
[22:04:08] <bhoeneis> We'll see once H.325 is out...;-)
[22:07:04] <iljitsch> the real fun starts when we've gotten rid of TDM
[22:07:29] <iljitsch> as someone who has configured a lot of BGP I'm not a huge fan of VoIP
[22:09:59] --- frodek has left
[22:10:01] <franck.martin> H.325 seems fun, I like the part in the report "Who modify SIP?", like why we did not see it coming...
[22:10:26] <franck.martin> iljitsch: I don't understand your comment re BGP/VoIP
[22:11:28] <franck.martin> in the H.325 I like also "regaining the lead.."
[22:11:34] <iljitsch> IP routing isn't the most stable thing in the world, I don't want to import those issues into my telephony
[22:13:09] <franck.martin> iljitsch: understood, but where I leave I need to try 3 times at least to make a phone call on my mobile, also it is very difficult for US operators to call me to join a conference call.... I think reliability of the phone network is overstated
[22:13:34] <mrex> KILL XML
[22:14:33] <mrex> (that could save LOTS of energy)
[22:14:37] <iljitsch> franck: it hasn't been getting much better lately
[22:14:40] --- john.zhao has left: Computer went to sleep
[22:14:43] --- TJ has left: Computer went to sleep
[22:14:48] --- fneves has left
[22:15:01] <iljitsch> but I think it's sad that we're tearing down a perfectly good infrastructure just because it saves us one or two cents a minute
[22:15:06] --- yangwoo has left: Computer went to sleep
[22:15:18] <iljitsch> goes both ways: telcos shouldn't keep their prices as high as they do
[22:17:04] --- iljitsch has left
[22:17:42] --- tlyu@jis.mit.edu has left
[22:17:59] --- Simon Leinen has left
[22:18:12] --- bhoeneis has left
[22:31:52] --- mohsen has left
[22:33:19] --- ryu has left: Disconnected
[22:54:48] --- franck.martin has left: Replaced by new connection
[23:12:52] --- fujiwara has left
[23:18:11] --- taiji-k has left