[03:29:28] Cory von Wallenstein joins the room [03:29:36] Cory von Wallenstein leaves the room [04:15:20] Matthew joins the room [05:41:23] Matthew leaves the room [07:47:36] Matthew joins the room [07:55:28] Matthew leaves the room [14:29:40] Matthew joins the room [15:16:18] Matthew leaves the room [15:40:33] Matthew joins the room [15:48:16] Matthew leaves the room [15:48:45] Matthew joins the room [16:03:40] TONGZHOU joins the room [16:06:43] TONGZHOU leaves the room [16:27:05] weiler joins the room [17:10:43] weiler leaves the room [18:33:53] Matthew leaves the room [19:04:54] tony joins the room [19:05:04] tony leaves the room [19:07:02] jennifergichure joins the room [19:08:03] jennifergichure leaves the room [19:25:51] ogud joins the room [19:27:09] ogud has set the subject to: IETF-77 DNSEXT meeting starts at 13:00 Pacific/22:00UTC [19:37:26] jelte joins the room [19:42:39] wouter joins the room [19:47:33] hey wouter [19:47:51] Hi [19:49:28] Patrik Wallström joins the room [19:51:20] healthyao2000 joins the room [19:52:15] Hi Wouter [19:52:44] hsalgado joins the room [19:53:40] i'm glad the meeting isn't at 15:00 or 17:40 :) [19:53:52] danny joins the room [19:53:53] sebastian.castro joins the room [19:55:26] could someone please say hi into a mic for reception quality check and local volume normalization? [19:55:45] before it starts :) [19:56:05] jo, ungefär så… dnsext-audio verkar lira på bra, men det är bara mummel så här långt :) [19:56:09] oops [19:57:15] hardaker joins the room [19:58:16] niall joins the room [19:58:19] I'm hearing quite well [19:58:55] but can you make out the individual voices in the din? [19:59:10] yes, I recognize some of the voices as well.. [19:59:12] geoff joins the room [19:59:20] i think i heard olafur say '..this turns it on' a minute ago :) [19:59:32] or some discussion to that effect [20:00:12] Ted joins the room [20:00:45] yone joins the room [20:01:12] Rickard Bellgrim joins the room [20:01:22] Ted leaves the room [20:01:51] koji joins the room [20:02:00] Frederico Neves joins the room [20:02:45] resnick joins the room [20:03:01] resnick leaves the room [20:03:17] hardie joins the room [20:03:32] matthijs joins the room [20:03:32] Matthew Dempsky joins the room [20:03:34] sean.s.shen joins the room [20:03:55] that's me! [20:04:02] thanks wes! [20:04:08] heh. [20:04:13] Stephen joins the room [20:04:15] hehehehe [20:04:20] we're now in: 0. Meeting comes to order, scribes appointed (5 minutes) [20:04:51] resnick joins the room [20:04:52] working on slide project functionality [20:05:10] and yes, I'm dating myself with the term "slide projector" [20:05:35] Hyong-Jong Paik joins the room [20:05:47] note well: anything you say here can and will be used against you [20:06:12] where are the slides? [20:06:14] now on to 2. Name equivalences (1 hour: earliest start 13:05 latest end 14:10) [20:06:24] https://datatracker.ietf.org/meeting/77/materials.html#wg-dnsext [20:06:52] stelager joins the room [20:06:55] tx [20:06:55] Also useful: http://trac.tools.ietf.org/wg/dnsext/trac/wiki [20:07:19] and for those looking for the agenda as a whole: http://www.ietf.org/proceedings/10mar/agenda/dnsext.txt [20:07:31] Paul Selkirk joins the room [20:07:37] Suz joins the room [20:07:47] jinmei joins the room [20:07:53] johani joins the room [20:07:59] Evan Hunt joins the room [20:08:22] Vogon! [20:08:24] ggm joins the room [20:08:32] I'd appreciate it during the discussion if anyone has something that should be said at the mic please preface your note with "RELAY: " [20:08:56] rdv@jabber.postel.org joins the room [20:09:07] discussing reqts for name equivalence [20:09:15] Andrew Sullivan talking [20:09:19] (is somebody doing this or not?) [20:09:31] Bernie joins the room [20:09:37] I am; but I'm assuming most people are listening to the audio. [20:09:47] cool. then I won't.! [20:09:50] is anyone not listening to the audio and needs more of a note-by-note capture? [20:10:00] now on slide "Quivelance": why [20:10:05] cary joins the room [20:10:09] There are some sitting in other rooms trying to track two WGs... [20:10:16] but I wouldn't say it's high priority... [20:10:21] Stephen leaves the room [20:10:28] kay; I'll summarize at least some then. [20:10:49] Stephen joins the room [20:10:53] mostly tracking agenda would be useful... [20:10:53] Cory von Wallenstein joins the room [20:11:00] sum: lots of reasons for doing name equiv derives from IDNA issues with name "variants" [20:11:13] raj joins the room [20:11:39] next slide: "equivalence": why? [20:11:49] and why CNAME/DNAME don't work as is [20:12:30] Kingy joins the room [20:12:34] jpc joins the room [20:12:43] keith joins the room [20:12:58] John: no one has been able to ask the question "what are all the *other* names with this node? [20:13:09] +1 [20:13:13] +1.5 [20:13:53] sounds impossible given some of the variants people have requested (no relay necessary) [20:13:58] Carl Wallace joins the room [20:14:01] Inverse problem! -- not what TLD registries are making like they need [20:14:09] a use for the reverse DNS? [20:14:12] RELAY: is there any other name lookup mechanisms that give you equivalent functionality? [20:14:15] skudou joins the room [20:14:46] patrik: you're in queue [20:15:07] if I wish to know if jorge is a legal synonym for me in Spanish, I need this [20:15:19] or Uri in russian [20:15:23] Dave Thaler joins the room [20:15:28] weiler joins the room [20:15:48] RussMundy joins the room [20:15:49] keith_nm joins the room [20:16:21] Kingy leaves the room [20:16:43] Chris Griffiths joins the room [20:17:42] bkuerbis joins the room [20:18:27] Kingy joins the room [20:18:39] benno joins the room [20:19:38] geoff leaves the room [20:19:58] JeremyHitchcock joins the room [20:20:16] next slide: "equivalance": out of scope [20:21:14] not solving specific language problems; not going to break the entire installed base [20:21:20] (ie, this is not DNSng) [20:21:22] Cory von Wallenstein leaves the room [20:21:51] wouter leaves the room [20:22:01] wouter joins the room [20:22:08] yuji joins the room [20:22:09] Jakob Schlyter joins the room [20:25:05] discussion surrounding if the solution means simultaneously synchronizing data and if we need immediate vs gradual updates as the data should be "identicalized" [20:26:13] naptee joins the room [20:27:02] ywang830 joins the room [20:28:43] Kingy leaves the room: Replaced by new connection [20:28:45] Kingy joins the room [20:29:11] tony joins the room [20:30:21] Michael Graff joins the room [20:30:27] Vincent joins the room [20:30:30] Need brief "Application Considerations" section in doc? [20:30:52] Before I get to the mic, can someone tell me why this isn't a tool issue? If you need 5 names, write out 5 names. If you need 5 zones, write out 5 zones. [20:30:53] daniel.brown joins the room [20:31:00] jr112 joins the room [20:31:33] geoff joins the room [20:31:37] i think that because this is dnsext, ppl only think of in-zone contents [20:31:39] 5 is a doable scale; 2**5 is something else [20:32:00] I have 10,000 spam zones. I want all www's to point to my godaddy A record until godaddy blocks me; then I want to move them all to the next sucker hosting service. [20:32:00] Do you expect 2**5? [20:32:37] Greek registry already has cases which need 2**3 for full coverage [20:32:49] i still haven't been convinced that it can't be solved (as far as we can solve it at all) by having a zone-clone *config option* instead of an in-zone solution [20:33:01] IDNA2008 will extend that to 2**5 [20:33:03] Are the 2**3 cases machine deterministically generated? [20:33:12] and no matter what you do, you'll get explosions once you start talking about multiple labels [20:33:15] Or do they require human intervention to know what those will be? [20:33:19] zone-clone would mean you would need to tell secondaries to do that too, or have them transfer all zones. [20:33:26] yep [20:33:38] and that would automatically solve the update issue as well :p [20:33:55] I don't know the .GR registry tool chain [20:33:59] CNNIC-AA5684158 joins the room [20:34:13] IMHO, the tools will have to change anyway. Change them to just handle it :) [20:34:23] Write out a DNAME + apex records. [20:34:30] :) [20:34:33] There ya go, BNAME with today's tools. [20:34:33] i wanted to send a mail about this after the interim meeting, but i completely forgot until just now [20:34:38] .GR current practice avoids filling all cases -- use 4 or 5 DNAME pseudo-deleg instead [20:35:06] the bigger problem is that this wg can't do a config-or-tool-solution (as andrew is just saying kinda) [20:35:07] .GR have declared BNAME adequate [20:35:17] (IIUC) [20:35:22] jaap joins the room [20:35:41] When you say apex records, are you including the dnssec rr's? [20:36:57] Rickard Bellgrim leaves the room [20:37:00] daniel.brown leaves the room [20:37:02] JeremyHitchcock leaves the room [20:37:09] sure [20:37:11] Rickard Bellgrim joins the room [20:37:19] JeremyHitchcock joins the room [20:37:26] Use the same crypto keys if you want. [20:37:32] Don't they have to be re-generated rather than transferred, given the current set? [20:38:00] daniel.brown joins the room [20:38:04] Michael Graff: yeah, but you still need to regen the rrsigs for the 10**5 zones. [20:38:10] if a tool writes zone-1.com and zone-2.com it can use the same DNSKEY records in both. Signatures would need to be special at the apex, but that's ok IMHO [20:38:44] you'll have to regen sigs for any solution afaict [20:38:53] or re-use keys [20:39:00] unless you want to special case this in validators and resolvers as well [20:39:06] It's ok to re-use the crypt keys, IMHO, it's the same security container. [20:39:18] that is, the keying data can still be identical. [20:39:21] benno leaves the room [20:39:23] jelte: actually, no... if one zone said "I'm that zone" there is only one sig that doesn't need to change when "that zone" is changed. [20:39:57] not if it has to appear as two zones to the outside [20:40:11] from [20:40:13] silde change recently to "Equivalance": interim (I missed the change; but believe it was recent) [20:40:19] remember that some RR include domain names in RDATA, and those needs new signatures as well [20:40:23] DNAME + zone apex would have to be re-signed only on apex changes. [20:40:26] (oops,wrong window) [20:40:48] why would they need new signatures? [20:41:14] weiler leaves the room [20:41:16] ah sorry any other solution that does not involve existing records [20:41:55] Unless the name inside the rdata changes, of course. I'd say "tools should use the non-aliased data" -- using CNAME synth means you would have to do this anyway I think, otherwise the synth cname as Paul just described would be invalid in a MX [20:42:19] benno joins the room [20:42:48] Cory von Wallenstein joins the room [20:44:15] cary leaves the room [20:46:27] mcharlesr joins the room [20:46:42] cary joins the room [20:46:46] it seems that we got "STARTTLS" wrong --- it should have specified the name that was being connected to. [20:47:00] now on "Equivalence": list and then quickly to "kinds" after that. [20:47:02] For STARTTLS, is there a reason that the client can't be responsible for knowing equivalent names too? I.e., if it connects to x.y.example.com and gets a cert for y.x.example.net, it can 'know' the same way the DNS 'knows' that they're the same? [20:47:13] rgonzalez joins the room [20:47:21] only if you have canonicalization [20:47:22] now TODO [20:47:23] Matthew Dempsky, if the client can see the CNAME/DNAME/BNAME chain, yes. [20:47:46] it seems that we got STARTTLS wrong though --- it should have listed which server the client thought it was talking to. [20:47:47] matthew: clients don't know now that they're the same. [20:48:11] The mechanism is "security by prayer". [20:48:13] They don't work under this scenario right now anyway, so they'll have to be upgraded regardless, right? [20:48:13] Andrew: "what are the requirements of desired behaviour" [20:48:21] yu kyung joins the room [20:48:32] desired behaviour must be done before design/results-choice [20:48:55] TTL 0 :p [20:49:01] They work when things work. :-) [20:49:04] Totally a tool issue. IMHO. [20:49:57] Cory von Wallenstein leaves the room [20:50:37] That is, if the cert happens to match the name used. But the world is getting more complicated (due to IDNA and other things) that clients need a problem solved. It may or may not be work for DNS, or Security folks, or clients. [20:51:16] certs can have more than one name [20:51:44] Current practice is for registrars to charge more for that though. [20:51:50] Michael Graff: interestingly enough, there was recently a message on openssl-users discussing that it breaks after you get to 500 names ;-) [20:52:06] openssl might break, but the cert doesn;t [20:52:10] Rickard Bellgrim leaves the room [20:52:21] the better solution would be to hand back an A record to the app along with a CERT fingerprint and the needed RRSIGs to validate it all. [20:52:23] hardaker ducks [20:52:25] Rickard Bellgrim joins the room [20:52:58] Did we mention that applications don't care what's below them? If openssl breaks or the cert breaks, it's all the same to the app. [20:53:06] hardaker, it's a good idea. [20:53:24] resnick: Yes, that's why openssl returns such lame error messages that leaves the user clueless! [20:53:39] :) [20:53:51] but, the 128 variants affects not just the zone, but also the certificate. [20:53:54] not a lot, but there are actually implementation that do not use openssl :) [20:53:58] hardaker slowly sticks his head above the fence and wonders if it was a trap [20:54:02] +s [20:54:40] why doesn't the mta present a cert for its hostname instead of certs for all the domains using the mta? [20:55:25] wow. we're designing x.400 [20:55:27] I think because Mark means that its hostname has 128 variants [20:55:28] question 1: is one name as the official and everything else is an alias for it good enough (not answering today)( [20:55:37] ggm: about time someone did [20:55:46] ah, the _hostname_, but that is not what we're discussing, is it? [20:56:06] slide is now "constraints", btw [20:56:08] wouter: But if those 128 variants are automatically generated, can't the MTA expand the 128 as well? [20:56:11] (to marka's comment) RDATA in MX needs to be absolute, not relative [20:56:22] palwal: And how exactly does the MTA know what it's hostname? [20:56:31] daniel.brown leaves the room [20:56:57] resnick: most of the time I tell it, or it uses the hostname of the box [20:57:05] daniel.brown joins the room [20:57:06] resnick: 1000 aliased domains with one MX, mx.blabla.example, pointing to an mta with a cert for mx.blabla.example? [20:57:20] niall: if you wanted me to relay that, please let me know (it's unclear) [20:58:13] You're assuming commonality of administrative control between the DNS (MX assigner) and the MTA. It doesn't happen. (And that's why we have this problem. And at least this one is not solvable.) [20:58:30] Suppose you wrote out 32 zones with identical data, other than the MX targets always pointed to the same host (where same == same DNS wire encoding here.) Why that is what hosting companies do now, and it works. Why change this? [20:58:36] JeremyHitchcock leaves the room [20:58:39] Jakob Schlyter leaves the room [20:58:39] Cory von Wallenstein joins the room [20:58:44] matthijs leaves the room [20:58:45] Rickard Bellgrim leaves the room [20:58:49] Jakob Schlyter joins the room [20:58:50] JeremyHitchcock joins the room [20:59:00] Rickard Bellgrim joins the room [20:59:04] raj leaves the room [20:59:44] i just lolled [20:59:45] matthijs joins the room [21:00:25] Protocol police: arrest that man! [21:00:40] it only works with "CNAME" if you never do TLS. [21:00:44] Mark's on the case. [21:02:01] ray joins the room [21:02:30] Andrew is now talking directly to the "Constraints" slide [21:04:04] Cory von Wallenstein leaves the room [21:04:50] wouter leaves the room [21:04:53] wouter joins the room [21:05:37] not even their resolvers, but their applications [21:06:13] now on the discussion slide (like THAT was needed) [21:06:27] (and now back to constraints) [21:06:57] hmz, audio lags way more than i had hoped [21:07:09] audio is back now to constraints now :) [21:07:13] reminder: please prefix relay requests with "RELAY: " [21:09:35] Cory von Wallenstein joins the room [21:10:32] mark just made my point, so not necessary :) [21:10:41] times up [21:10:46] for that topic [21:10:46] matthijs leaves the room [21:11:24] * [21:11:28] matthijs joins the room [21:11:34] ray leaves the room [21:11:35] daniel.brown leaves the room [21:11:44] next on agenda: 3.1 draft-ietf-dnsext-dnssec-bis-updates-10 (15 minutes: start [21:11:53] daniel.brown joins the room [21:11:58] sam weiler now speaking [21:13:06] (slides aren't up) [21:13:22] yu kyung leaves the room [21:13:36] yu kyung joins the room [21:13:46] hardaker: hm, I got the slides anyway [21:13:56] heh [21:14:25] niall leaves the room [21:14:59] now on "answering queries with CD" [21:15:10] sconte joins the room [21:15:56] cary leaves the room [21:16:28] now on "anything else?" [21:17:02] yuji leaves the room: Replaced by new connection [21:17:02] yuji joins the room [21:17:16] hmm, what about sha512? [21:17:43] Rickard Bellgrim: want me to relay that? [21:17:46] keith leaves the room [21:17:51] (though I bet I know the answer) [21:17:55] daniel.brown leaves the room [21:17:56] benno leaves the room [21:17:57] keith_nm leaves the room: Logged out [21:17:58] matthijs leaves the room [21:18:06] nah, just thinking why it its not mentioned in the document [21:18:16] kay. [21:18:22] daniel.brown joins the room [21:18:30] I'll mention it to Sam when he walks by again [21:19:01] matthijs joins the room [21:20:13] is the dnscurve slides available? [21:20:14] wouter leaves the room [21:20:20] wouter joins the room [21:21:02] it's really Paul V for anyone that doesn't recognize the voice [21:21:12] identity theft! [21:21:49] It's just a CNAME. [21:21:53] we're about 40 seconds behind on remote audio :p [21:22:20] should it be possible to express an empty set of clones for equivalence? [21:22:25] Antoin joins the room [21:22:47] weiler joins the room [21:22:57] 3.2 draft-dempsky-dnscurve-01 (15 minutes: start 14:25 end 14:40) [21:23:09] Matthew Dempsky [21:23:23] Cory von Wallenstein leaves the room [21:24:38] ogud has set the subject to: IETF-77 DNSEXT meeting in session [21:24:50] мαя¢σ joins the room [21:24:53] re: 512. we don't have an i-d on 512 do we? [21:25:07] woolf joins the room [21:25:11] Suz leaves the room [21:25:22] dempsky.org. 259200 IN NS uz5p4utwsxu5p3r9xrw0ygddw2hxh7bkhd0vdwtbt92lf058ny1p79.dempsky.org. [21:25:49] yes we do, it was for sha2 family [21:25:54] daniel.brown leaves the room [21:26:08] benno joins the room [21:26:30] daniel.brown joins the room [21:26:38] http://tools.ietf.org/html/rfc5702 [21:28:02] Stephen leaves the room [21:29:00] got it. we'll fix it. [21:29:17] question on the table: does the working group want to take this on? what track? [21:31:05] benno leaves the room [21:32:04] мαя¢σ leaves the room [21:33:51] keith joins the room [21:33:58] keith_nm joins the room [21:34:10] мαя¢σ joins the room [21:36:00] Rickard Bellgrim leaves the room [21:36:15] Rickard Bellgrim joins the room [21:36:31] FYI, middle boxes issues being discussed now [21:36:41] (and sorry for not summarizing enough) [21:37:39] Olaf wonders what we're solving [21:37:55] Cory von Wallenstein joins the room [21:38:01] daniel.brown leaves the room [21:38:28] daniel.brown joins the room [21:40:14] mcharlesr leaves the room [21:41:23] benno joins the room [21:41:33] hums: should we adopt it? should we not? how many unsure? [21:43:34] next item is 3.3 draft-vixie-dnsext-dns0x20-00 [expired draft] (3 minutes: Paul Vixie [21:43:35] i don't get a feeling of the humming on the mic, can scribe provide your feeling [21:43:45] sometimes known as Marc Andrews [21:43:59] andrew summarized afterward [21:44:30] but copying him: some low-level support for adopting, some low-level against and a large percentage unsure [21:45:00] thanks [21:45:02] np [21:45:17] many have read paul's draft [21:46:42] questions: how many people support adoption? [21:46:52] question: how many opposed, and not made up? [21:47:24] record: some hum for, one non-commital hum opposed and none not-made-up [21:47:31] can we extend 0x20 to idna variants? [21:47:34] KIM K joins the room [21:47:44] 3.4 Roll over and die (15 minutes: start 14:43 end 14:58) Roy Arends [21:47:46] jelte: :-) [21:47:51] jelte: yes [21:47:57] Jakob Schlyter leaves the room [21:48:41] Stephen joins the room [21:48:53] title slide [21:49:11] daniel.brown leaves the room [21:49:19] slide 2: we're under attack [21:49:36] slide 3: dnskey application attack [21:50:03] slide 4: dnskey response size [21:50:08] slide 5: who does this [21:50:11] jpc leaves the room: offline [21:50:20] daniel.brown joins the room [21:50:30] KIM K leaves the room [21:51:09] slide 6: what was special about the 16th? [21:51:37] slide 16.2: red circled [21:51:40] err... 6.2 [21:52:08] next slide: Hanlon's razor [21:52:13] next slide: why so many clients [21:52:20] (he can changes slides faster than I can type!) [21:52:47] doesn't matter wes, we're behind anyway :) [21:52:59] slide: "current load for in-addr.arpa" [21:53:03] 3 slides per second [21:53:29] slide: the load projection [21:53:50] pretty blue arrows and then red added [21:53:59] slide: "was this a one off event?" [21:54:13] then next, same title [21:54:17] Rickard Bellgrim leaves the room [21:54:27] +pretty blue arrows [21:54:37] slide: "why so many queries?" [21:54:42] Rickard Bellgrim joins the room [21:55:45] will do the upcoming break, then upstairs for a nap [21:55:45] second slide of same title [21:55:55] sorry, wrong window :-) [21:56:30] third slide of same plus pretty graphics and bouncing yellow smileys [21:57:33] next slide "ISC" [21:59:01] Carl Wallace leaves the room [21:59:27] slide: perfect storm [21:59:58] slides: skipping and skipping until "a series of unfortunate events" [22:00:22] did he just go even faster? :) [22:00:37] he broke the sound barrier :) [22:00:48] reaching ekr speed? [22:00:58] :) [22:01:01] they were slides: he broke the light barrier [22:01:02] only 1 minutes limitation by chair [22:01:03] yeah roy can be like that when he's trying to get a point across :) [22:01:20] CNNIC-AA5684158 leaves the room [22:01:27] slide "frequent rollover syndrom" [22:01:31] Michael Graff leaves the room [22:01:38] perhaps validators should not allow configuration of non-5011 anchors? (without complaining a lot) [22:01:45] bkuerbis leaves the room [22:01:46] Xiaodong joins the room [22:02:19] slide: solution [22:02:21] (last) [22:02:35] wouter: Or encode an expiration date into them. [22:02:43] wouter: that still would not prevent configuring stale keys [22:02:49] we need trust-history! [22:03:20] skudou leaves the room [22:03:22] slides: questions/remarks/observations [22:03:29] (though I'm not sure there is time) [22:03:43] could history not result in even more queries? [22:04:14] Rickard Bellgrim leaves the room [22:04:15] matthijs: give roy a beer from me :) [22:04:22] hardie leaves the room [22:04:26] Rickard Bellgrim joins the room [22:04:28] stelager leaves the room [22:04:33] JcK joins the room [22:04:41] Rickard Bellgrim leaves the room [22:04:44] koji leaves the room [22:04:48] jelte: shall i give the whole wg a round? :p [22:04:48] Chris Griffiths leaves the room [22:04:53] all done [22:04:55] Matthew Dempsky leaves the room [22:04:56] jaap leaves the room [22:04:57] Kingy leaves the room [22:04:58] good night LA! [22:04:59] jinmei leaves the room [22:05:01] weiler leaves the room [22:05:03] benno leaves the room [22:05:03] JeremyHitchcock leaves the room [22:05:06] ogud has set the subject to: IETF-77 DNSEXT meeting over [22:05:06] sconte leaves the room [22:05:09] ywang830 leaves the room [22:05:10] Paul Selkirk leaves the room [22:05:14] Patrik Wallström leaves the room [22:05:16] that might be a bit much :) [22:05:16] resnick leaves the room [22:05:25] JcK leaves the room [22:05:28] ;) [22:05:30] but you may also get one yourself [22:05:30] yuji leaves the room [22:05:33] will do [22:05:38] matthijs leaves the room [22:05:40] hsalgado leaves the room [22:05:41] Xiaodong leaves the room [22:05:44] geoff leaves the room [22:05:45] ogud leaves the room [22:05:45] RussMundy leaves the room [22:05:46] thank you chairs [22:05:50] Stephen leaves the room [22:05:51] danny leaves the room [22:05:52] Cory von Wallenstein leaves the room [22:05:57] Frederico Neves leaves the room [22:05:58] jelte, that is an important question, I'll note it. It does check all the upstream servers. [22:05:59] yone leaves the room [22:06:17] Bernie leaves the room [22:06:17] thanks [22:06:18] ggm leaves the room [22:06:39] woolf leaves the room [22:07:09] naptee leaves the room [22:07:09] ogud joins the room [22:07:20] keith_nm leaves the room [22:07:38] sebastian.castro leaves the room [22:07:48] yu kyung leaves the room [22:08:03] jr112 leaves the room [22:08:26] Kingy joins the room [22:08:50] rdv@jabber.postel.org leaves the room [22:09:20] daniel.brown leaves the room [22:10:02] daniel.brown joins the room [22:10:03] rgonzalez leaves the room [22:11:04] wouter leaves the room [22:12:31] Vincent leaves the room [22:12:48] Kingy leaves the room [22:14:45] ogud leaves the room [22:16:19] Evan Hunt leaves the room [22:16:43] hardie joins the room [22:19:49] Antoin leaves the room: Replaced by new connection [22:22:15] Dave Thaler leaves the room [22:22:19] hardaker leaves the room [22:22:23] hardie leaves the room [22:23:05] мαя¢σ leaves the room: Disconnected [22:23:30] sean.s.shen leaves the room [22:24:19] johani leaves the room [22:24:24] jinmei joins the room [22:25:20] Stephen joins the room [22:32:15] healthyao2000 leaves the room [22:33:56] Stephen leaves the room [22:33:56] Hyong-Jong Paik leaves the room [22:34:37] мαя¢σ joins the room [22:35:04] ggm joins the room [22:36:11] ggm leaves the room [22:38:01] johani joins the room [22:38:08] ggm joins the room [22:38:21] мαя¢σ leaves the room: Disconnected [22:38:33] johani leaves the room [22:39:44] jinmei leaves the room [22:40:01] ggm leaves the room [23:00:57] jelte leaves the room [23:28:24] tony leaves the room [23:53:13] keith leaves the room