Displaying #traffic-server/2015-11-16.log:

Mon Nov 16 00:50:13 2015  psp:Joined the channel
Mon Nov 16 01:29:35 2015  danielxu:Joined the channel
Mon Nov 16 01:36:57 2015  ihabunek:Joined the channel
Mon Nov 16 01:38:02 2015  briang:Joined the channel
Mon Nov 16 01:38:02 2015  briang:Joined the channel
Mon Nov 16 01:38:07 2015  phibs:Joined the channel
Mon Nov 16 01:38:34 2015  dcarlin:Joined the channel
Mon Nov 16 01:38:51 2015  maskit:Joined the channel
Mon Nov 16 01:39:03 2015  kichan:Joined the channel
Mon Nov 16 01:39:48 2015  ggherdov:Joined the channel
Mon Nov 16 02:48:41 2015  leprechau:Joined the channel
Mon Nov 16 03:41:00 2015  psp1:Joined the channel
Mon Nov 16 03:55:15 2015  psp:Joined the channel
Mon Nov 16 03:58:06 2015  ibrezac:Joined the channel
Mon Nov 16 04:02:13 2015  PSUdaemon:jpeach: so amc and I are looking at the Hash stuff more now. we think you actually want to separate the function and the state, not the function and the hash.
Mon Nov 16 04:02:20 2015  PSUdaemon:well, maybe separate all three
Mon Nov 16 07:11:30 2015  adedomme1in:Joined the channel
Mon Nov 16 07:19:40 2015  ggherdov:Joined the channel
Mon Nov 16 09:11:38 2015  Lethalman:Joined the channel
Mon Nov 16 09:52:11 2015  tomr_frnd:Joined the channel
Mon Nov 16 10:47:13 2015  Randeep:Joined the channel
Mon Nov 16 10:50:21 2015  Randeep:hi, I have created a Parent and Child ats server and streaming videos from the system conencted to child. Its working. But on the child its not caching. its giving parent hit but tc miss
Mon Nov 16 10:50:22 2015  Randeep:1447670883.766 232 172.16.0.17 TCP_MISS/200 390210 GET http://jitp.lukup.com/dash/mimarathi/3_289-0-I-1217322.m4s?_=1447671001880 - PARENT_HIT/172.16.0.40 video/mp4
Mon Nov 16 10:50:35 2015  Randeep:what can be the issue?
Mon Nov 16 10:51:24 2015  Randeep:on the child I have configured parents.config as below
Mon Nov 16 10:51:24 2015  Randeep:dest_domain=. parent="172.16.0.40:80" go_direct=false
Mon Nov 16 11:05:24 2015  niq:Joined the channel
Mon Nov 16 12:01:25 2015  JSeymour:Joined the channel
Mon Nov 16 12:47:46 2015  niq:Joined the channel
Mon Nov 16 12:47:46 2015  niq:Joined the channel
Mon Nov 16 12:49:19 2015  Randeeppr:Joined the channel
Mon Nov 16 12:57:08 2015  ibrezac:Joined the channel
Mon Nov 16 13:21:46 2015  gancho:Joined the channel
Mon Nov 16 14:12:36 2015  ibrezac:Joined the channel
Mon Nov 16 14:55:31 2015  esproul:Joined the channel
Mon Nov 16 15:01:28 2015  shinrich1:Joined the channel
Mon Nov 16 15:15:58 2015  gancho:Joined the channel
Mon Nov 16 15:28:53 2015  reveller:Joined the channel
Mon Nov 16 15:32:55 2015  gancho:Joined the channel
Mon Nov 16 15:37:47 2015  shinrich1:Joined the channel
Mon Nov 16 15:38:42 2015  es:Joined the channel
Mon Nov 16 15:40:07 2015  dxu_:Joined the channel
Mon Nov 16 15:54:02 2015  jrickman:Joined the channel
Mon Nov 16 16:13:26 2015  amc:Hey hey, summit time! With Leif!
Mon Nov 16 16:14:13 2015  Humbedooh:oh boy!
Mon Nov 16 16:16:27 2015  amc:I realize that a Leif is a poor substitute for a Humbedooh but we do what we can.
Mon Nov 16 16:19:20 2015  zwoop:Well, I'm here, I don't that many danes in here
Mon Nov 16 16:22:32 2015  Humbedooh:your favorite dane isn't there
Mon Nov 16 16:22:36 2015  Humbedooh:and neither am I :(
Mon Nov 16 16:39:13 2015  niq:Joined the channel
Mon Nov 16 16:45:37 2015  ibrezac:Joined the channel
Mon Nov 16 16:51:28 2015  davet_:Joined the channel
Mon Nov 16 16:54:53 2015  zwoop:My favorite dane is mr Carlsberg
Mon Nov 16 16:56:26 2015  psp:Joined the channel
Mon Nov 16 16:57:32 2015  reveller:Thanks for setting up the Google Hangout... I will sit in when I can throughout the day
Mon Nov 16 16:57:45 2015  amc:Can I wave at you?
Mon Nov 16 16:57:56 2015  reveller:waves back
Mon Nov 16 16:58:25 2015  reveller:no audio yet
Mon Nov 16 16:58:38 2015  amc:We've decided to have the bug scrub in your house.
Mon Nov 16 16:58:45 2015  zwoop:reveller which island are you on again ?
Mon Nov 16 16:58:57 2015  amc:reveller - can you hear Dr. H?
Mon Nov 16 16:59:07 2015  reveller:St Croix, US Virgin Islands
Mon Nov 16 16:59:51 2015  reveller:I can but barely... I can't unmute my mic to respond tho
Mon Nov 16 17:00:14 2015  reveller:I can!
Mon Nov 16 17:00:21 2015  reveller:YEAH!
Mon Nov 16 17:04:10 2015  es:Joined the channel
Mon Nov 16 17:05:35 2015  PSUdaemon:Changed the topic to: Apache Traffic Server | https://docs.trafficserver.apache.org | Current release: 6.0.0 | Channel logs at http://wilderness.apache.org/channels/?c=traffic-server | ATS Summit: https://plus.google.com/hangouts/_/yahoo-inc.com/ats
Mon Nov 16 17:06:43 2015  PSUdaemon:sounds like leif is saying grace
Mon Nov 16 17:07:08 2015  PSUdaemon:"we thank the yahoo for this great food and this roof over our heads"
Mon Nov 16 17:07:25 2015  reveller::D
Mon Nov 16 17:15:39 2015  PSUdaemon:yeah, lets do a bug scrub there
Mon Nov 16 17:15:41 2015  PSUdaemon:great idea amc!
Mon Nov 16 17:38:40 2015  jpeach:PSUdaemon: seems like the EVP API has a reasonable design pattern for this
Mon Nov 16 17:39:31 2015  jpeach:you have a set of constant algorithms, then bind that to a context (which can be a generic struct), then on finalize you extract a hash value
Mon Nov 16 17:39:33 2015  meeramn:Joined the channel
Mon Nov 16 17:42:04 2015  zwoop:jpeach does that imply that it has to keep (dupe?) any data that you feed into it ?
Mon Nov 16 17:42:12 2015  jpeach:no
Mon Nov 16 17:42:25 2015  amc:It does mean you need to know the maximum size of any context.
Mon Nov 16 17:42:26 2015  zwoop:it has some higher level HMAC algorithm ?
Mon Nov 16 17:42:30 2015  jpeach:the context is just scratch space for the algorithms
Mon Nov 16 17:42:32 2015  jpeach:amc: yes
Mon Nov 16 17:42:57 2015  jpeach:amc: but you can enforce the right combinations at compile time I think
Mon Nov 16 17:43:03 2015  amc:At that point, though, you can create a union of all hash functions with a specific output size, get the size of that,
Mon Nov 16 17:43:16 2015  amc:have that as a chunk of memory and placement new the hash function class in it.
Mon Nov 16 17:44:25 2015  jpeach:if you did that then you might as well just use a hash object directly
Mon Nov 16 17:44:50 2015  jpeach:unless you want to do something more complicated and type-erase the thing you placement-new?
Mon Nov 16 17:45:16 2015  amc:Yes, but the point would be to make it run time selectable without allocation.
Mon Nov 16 17:46:29 2015  es:Joined the channel
Mon Nov 16 17:47:54 2015  jpeach:which requires a concrete context type, a concrete hash value type, and a variable algorithm selection
Mon Nov 16 17:48:28 2015  PSUdaemon:yeah
Mon Nov 16 17:48:42 2015  PSUdaemon:but the concrete context type can be cast from a memory blob from the caller
Mon Nov 16 17:49:47 2015  PSUdaemon:that is amc's idea
Mon Nov 16 17:50:20 2015  jpeach:PSUdaemon: you just need the max size and the right alignment
Mon Nov 16 17:50:33 2015  PSUdaemon:yeah
Mon Nov 16 17:51:30 2015  jpeach:and it ends up looking very similar to the EVP pattern
Mon Nov 16 17:52:09 2015  PSUdaemon:yup
Mon Nov 16 17:53:51 2015  PSUdaemon:can we do like a class scoped struct that you can pull from the ATSHashFunc?
Mon Nov 16 17:55:15 2015  jpeach:not without knowing the concrete type of the context
Mon Nov 16 18:00:54 2015  reveller:amc: PSUdaemon: bug scrub or not, "come on down!"
Mon Nov 16 18:04:14 2015  PSUdaemon:already checking flight prices :)
Mon Nov 16 18:04:18 2015  PSUdaemon:~$700
Mon Nov 16 18:04:25 2015  PSUdaemon:from DEN
Mon Nov 16 18:05:44 2015  PSUdaemon:jpeach: if you are just interested in the subject, Comcast has done some multi-cast stuff like you are describing
Mon Nov 16 18:06:18 2015  jpeach:PSUdaemon: I would like to hear about that
Mon Nov 16 18:06:34 2015  jpeach:zwoop: can you pls run regressions on OS X and let me know?
Mon Nov 16 18:06:43 2015  reveller:PSUdaemon: bring your dive gear! Water temps 79-82 year around with great visibility
Mon Nov 16 18:09:17 2015  zwoop:TS-4025
Mon Nov 16 18:09:20 2015  zwoop:PSUdaemon ^
Mon Nov 16 18:18:20 2015  jpeach:hey y'all ... regressions seem to be broken on OS X
Mon Nov 16 18:19:10 2015  jpeach:http://apaste.info/I8U
Mon Nov 16 18:30:25 2015  zwoop:woohoo
Mon Nov 16 18:30:44 2015  amc:This might be TS-3885
Mon Nov 16 18:31:11 2015  zwoop:TS-3885
Mon Nov 16 18:31:22 2015  jpeach:zwoop: yeh but you reverted that?
Mon Nov 16 18:31:41 2015  amc:I have seen you back trace as a result of that.
Mon Nov 16 18:31:51 2015  amc:But there's a PR with a different and working fix.
Mon Nov 16 18:32:12 2015  zwoop:yeah, looks like I reverted it
Mon Nov 16 18:32:19 2015  amc:I've been meaning to merge it but haven't gotten there.
Mon Nov 16 18:34:12 2015  jpeach:wel what broke the tests on master not?
Mon Nov 16 18:34:52 2015  kang1:Joined the channel
Mon Nov 16 18:35:41 2015  zwoop:jpeach you sure it's not a conflicting port? I've seen the synserver failing if something else has that port ?
Mon Nov 16 18:35:50 2015  zwoop:making a build now to test
Mon Nov 16 18:36:40 2015  jpeach:nothing that I can see on port 3300
Mon Nov 16 18:39:19 2015  amc:IT's a stall on the ACCEPT. It depends on the OS.
Mon Nov 16 18:39:29 2015  amc:For instance, it's not on FC but it is on Y! Linux.
Mon Nov 16 18:40:33 2015  amc:For the OS with the problem, the OS doesn't sent the accept notification to user space until data arrives on the socket.
Mon Nov 16 18:41:10 2015  zwoop:amc Weird. Is Yahoo Linux different than CentOS / RHEL linux kernel ?
Mon Nov 16 18:41:29 2015  blattj:Joined the channel
Mon Nov 16 18:41:54 2015  amc:Yes.
Mon Nov 16 18:42:14 2015  amc:This particular patch was a big one, beacuse it has a lot of performance impact.
Mon Nov 16 18:44:27 2015  zwoop:that explains why no one else sees this :)
Mon Nov 16 18:44:44 2015  zwoop:amc you made improvements to defer accept I assume
Mon Nov 16 18:44:52 2015  amc:Yes.
Mon Nov 16 18:44:54 2015  zwoop:(deferred accept in Linux is a bloody joke)
Mon Nov 16 18:44:58 2015  amc:Well, not me. Quentin I think.
Mon Nov 16 18:45:02 2015  zwoop:roger
Mon Nov 16 18:45:07 2015  jpeach:this is not the point
Mon Nov 16 18:45:14 2015  jpeach:the point is it crashes on OS X
Mon Nov 16 18:45:26 2015  amc:Yes. Does it still crash with that fix?
Mon Nov 16 18:45:36 2015  jpeach:bin/traffic_server -R 1 -r SDK_API_TSHttpConnectServerIntercept
Mon Nov 16 18:45:47 2015  amc:We saw crashed and hangs on Y!Linux until we made thiscahnge in about the same spot.
Mon Nov 16 18:46:17 2015  jpeach:it crashes on master and earlier ... have not gone back far enough to find a working version
Mon Nov 16 18:47:00 2015  zwoop:jpeach did you confirm it? (Im running it on my laptop now)
Mon Nov 16 18:47:17 2015  zwoop:never mind, mine just crashed :)
Mon Nov 16 18:47:35 2015  zwoop:this must be a fairly recent problem
Mon Nov 16 18:47:35 2015  amc:Try that fix, it's a trivial one, and see if that helps.
Mon Nov 16 18:48:08 2015  amc:Yes, that's odd, because that code is old. It might be because some DEFER_ACCEPT thing is being done elsewhere
Mon Nov 16 18:48:14 2015  jpeach:that fix must be irelevant
Mon Nov 16 18:48:33 2015  amc:Perhaps. But we should try it and verify if Leif sees it locally on his laptop.
Mon Nov 16 18:48:34 2015  jpeach:because this test always passed before that fix, and the fix was reverted
Mon Nov 16 18:48:52 2015  zwoop:amc I did
Mon Nov 16 18:48:59 2015  amc:Didn't help?
Mon Nov 16 18:49:16 2015  zwoop:well, current master still crashes, sorry. What's the patch to try ?
Mon Nov 16 18:49:31 2015  amc:https://github.com/apache/trafficserver/pull/313
Mon Nov 16 18:51:25 2015  ibrezac:Joined the channel
Mon Nov 16 18:51:33 2015  jpeach:it crashes on 5edd58cd860d2c74339b5a090b7c4c27e4102938
Mon Nov 16 18:51:46 2015  jpeach:let's not patch stuff ... let's find the problem
Mon Nov 16 18:54:48 2015  amc:For TS-3885 (which may be this bug) the problem was it woudl time out waiting for the ACCEPT. It would do that because of how DEFER_ACCEPT is handled by the OS.
Mon Nov 16 18:55:07 2015  amc:The solution is to write data to the socket for force the ACCEPT to show up in user space.
Mon Nov 16 18:55:41 2015  amc:Sometimes this would generate the "could not bind" assert.
Mon Nov 16 18:56:50 2015  jpeach:ok, so how does that manifest on OS X?
Mon Nov 16 18:56:50 2015  Randeep:Joined the channel
Mon Nov 16 19:00:40 2015  blattj:Joined the channel
Mon Nov 16 19:00:44 2015  blattj: https://plus.google.com/hangouts/_/yahoo-inc.com/ats
Mon Nov 16 19:01:16 2015  biilmann:Joined the channel
Mon Nov 16 19:01:43 2015  msekimura:Joined the channel
Mon Nov 16 19:09:15 2015  jpeach:zwoop, amc: network guys tell me that there is no deferred accept behaviour on OS X
Mon Nov 16 19:11:10 2015  reveller:we've lost audio in the Summit Hangout
Mon Nov 16 19:12:05 2015  PSUdaemon:can you hear masaori?
Mon Nov 16 19:12:08 2015  PSUdaemon:no one else is using a mic
Mon Nov 16 19:12:30 2015  reveller:only audio we hear is the passcode prompt for the conference call
Mon Nov 16 19:12:38 2015  amc:Josh is working on it.
Mon Nov 16 19:13:29 2015  shinrich1:Joined the channel
Mon Nov 16 19:16:22 2015  amc:Do you have a core for that? At thsi point I expect it's geting an unanticipated event.
Mon Nov 16 19:18:05 2015  Randeep:For live mpeg dash streams with small chunks, which ram cache algorithm is better clfus Or lru?
Mon Nov 16 19:18:55 2015  PSUdaemon:for live, probably LRU
Mon Nov 16 19:19:07 2015  PSUdaemon:or even FIFO :)
Mon Nov 16 19:19:10 2015  PSUdaemon:but that is not implemented
Mon Nov 16 19:19:26 2015  amc:Is sound still not available ?
Mon Nov 16 19:19:30 2015  sbeards:has a good use case for server push that he will be presenting today
Mon Nov 16 19:19:38 2015  reveller:just came back
Mon Nov 16 19:19:45 2015  meeramn:audio came back
Mon Nov 16 19:19:58 2015  meeramn:thanks
Mon Nov 16 19:20:04 2015  Randeep:I see very less mem hit in logs
Mon Nov 16 19:20:51 2015  meeramn:oh no! It stopped again
Mon Nov 16 19:21:08 2015  PSUdaemon:Randeep: with LRU?
Mon Nov 16 19:21:22 2015  Randeep:With clfus
Mon Nov 16 19:21:58 2015  PSUdaemon:linear has very special properties that most cache alg's don't take into acct properly
Mon Nov 16 19:22:06 2015  reveller:amc: we lost audio again
Mon Nov 16 19:22:19 2015  amc:Susan is working on it.
Mon Nov 16 19:22:55 2015  shinrich1:Audio now?
Mon Nov 16 19:23:00 2015  reveller:nope
Mon Nov 16 19:23:05 2015  Randeep:psudaemon: I'll try with lru
Mon Nov 16 19:23:40 2015  PSUdaemon:also possible that the very short lifespan means by the time it gets into cache it's not requested anymore
Mon Nov 16 19:23:56 2015  PSUdaemon:we use ram drives to pretty much just force this
Mon Nov 16 19:23:59 2015  PSUdaemon:but it's kind of a hack
Mon Nov 16 19:24:17 2015  reveller:there it is... we hear you loud and clear (for now)
Mon Nov 16 19:24:27 2015  shinrich1:Cool!
Mon Nov 16 19:24:32 2015  reveller:shinrich1: Thanks!
Mon Nov 16 19:24:33 2015  shinrich1:Keep us posted :-)
Mon Nov 16 19:24:39 2015  jpeach:zwoop: fix the build!
Mon Nov 16 19:24:53 2015  zwoop:jpeach make it so!
Mon Nov 16 19:24:57 2015  Randeep:Psudaemon: ramdrives? Ram or ramfs
Mon Nov 16 19:25:53 2015  ibrezac:is there any way to restrict access to the stats_over_http.so url?
Mon Nov 16 19:27:02 2015  zwoop:ibrezac not other than making an obfuscated URL
Mon Nov 16 19:27:23 2015  ibrezac:yep, that's what we are doing
Mon Nov 16 19:27:30 2015  zwoop:we'll take patches ;)
Mon Nov 16 19:27:36 2015  ibrezac:I thought @internal might work :)
Mon Nov 16 19:27:42 2015  ibrezac:working on other patches :)
Mon Nov 16 19:27:55 2015  Randeep:How do we benchmark ats. I tried to benchmark ats with a particular processor and mother board. It's taking 18% memory and 8%cpu while playing 20 streams
Mon Nov 16 19:28:28 2015  ibrezac:set-redirect 301 %<cquuh>
Mon Nov 16 19:28:37 2015  ibrezac:but master seems to be broken
Mon Nov 16 19:29:54 2015  jpeach:ibrezac: @internal is for requests that originate from inside traffic_server
Mon Nov 16 19:31:04 2015  ibrezac:Variable expansion in header_rewrite does not work and certain global header_rewrite configurations crash traffic_server, any ideas? Or do I need dig?
Mon Nov 16 19:32:10 2015  ibrezac:@jpeach, hmm, is this a map that points back to traffic_server?
Mon Nov 16 19:33:02 2015  jpeach:is what a map?
Mon Nov 16 19:39:23 2015  PSUdaemon:Randeep: ram drives. fixed size block devices backed by ram. we config them as volumes in ATS and direct linear directory services to that volume.
Mon Nov 16 19:44:02 2015  dxu_:Joined the channel
Mon Nov 16 19:49:17 2015  _klk_:Joined the channel
Mon Nov 16 20:02:05 2015  ibrezac:@jpeach, map from remap.config
Mon Nov 16 20:02:43 2015  es:Joined the channel
Mon Nov 16 20:31:47 2015  jpeach:no, @internal looks at internal transaction state, not remap.config
Mon Nov 16 20:33:56 2015  kang1:Joined the channel
Mon Nov 16 20:38:31 2015  _klk_:Joined the channel
Mon Nov 16 20:53:58 2015  Humbedooh:zwoop: tut tut
Mon Nov 16 20:56:54 2015  biilmann:Joined the channel
Mon Nov 16 20:58:22 2015  psp:Joined the channel
Mon Nov 16 21:05:36 2015  PSUdaemon:who uses the geoip plugin?
Mon Nov 16 21:27:49 2015  zwoop:PSUdaemon no one I think
Mon Nov 16 21:27:56 2015  zwoop:I use it on the "test" sites
Mon Nov 16 21:28:56 2015  psp:Joined the channel
Mon Nov 16 21:46:38 2015  blattj:Joined the channel
Mon Nov 16 21:47:07 2015  biilmann:PSUdaemon: I used it as inspiration for our own GeoIP functionality, but not using the plugin as is
Mon Nov 16 21:50:20 2015  PSUdaemon:biilmann: are you going to contribute your changes?
Mon Nov 16 21:50:31 2015  PSUdaemon:we are looking at this sort of functionality too
Mon Nov 16 21:55:23 2015  jpeach:zwoop, amc: http://fpaste.org/291222/44771091/
Mon Nov 16 22:00:00 2015  jpeach:zwoop: I figured it out
Mon Nov 16 22:00:53 2015  biilmann:PSUdaemon: in the end they're pretty specific to how we use GeoIP with our Origin servers, so probably too specific to our setup to really be browadly useful
Mon Nov 16 22:02:22 2015  biilmann:jpeach: btw - at the ATS summit today - if you want to talk HTTPS with loads of certificates next break...
Mon Nov 16 22:02:36 2015  jpeach:cool, come grab me
Mon Nov 16 22:05:23 2015  reveller:biilmann: we still need to get together at some point to share experiences with large scale SSL.
Mon Nov 16 22:05:49 2015  biilmann:reveller: yeah
Mon Nov 16 22:05:58 2015  biilmann:reveller: are you at the summit?
Mon Nov 16 22:06:15 2015  reveller:biilmann: only virtually on the Hangout
Mon Nov 16 22:08:04 2015  _klk_:Joined the channel
Mon Nov 16 22:08:45 2015  jpeach:biilmann: DrH is here too ... we should all chat
Mon Nov 16 22:09:14 2015  PSUdaemon:biilmann: you might be surprised...
Mon Nov 16 22:11:40 2015  biilmann:PSUdaemon: might try to open-source something more generalized based on this
Mon Nov 16 22:12:42 2015  biilmann:we basically set an X-Country header in the requests to the origin servers that either has the country detected by a maxminddb GeoIP lookup or a country based on a Cookie value
Mon Nov 16 22:14:32 2015  zwoop:biilmann I made some changes to header_rewrite to support that. It's not OpenSourced yet
Mon Nov 16 22:22:07 2015  biilmann:zwoop: neat - in our case we then do some really specific things to our usecase around sometimes taking country/language into account when picking alternates
Mon Nov 16 22:22:14 2015  jpeach:zwoop: does that make geoip and header_rewrite cooperate?
Mon Nov 16 22:23:07 2015  zwoop:jpeach I have another patch for header_rewrite too (still pending same nonsense reviews....) which lets you do everything in header_rewrite if you'd want to. We also could change geoip participate in this I guess, but I haven't touched that part yet.
Mon Nov 16 22:23:29 2015  zwoop:jpeach what's missing in header_rewrite is a way to "reject" the request.
Mon Nov 16 22:23:32 2015  zwoop:(easy fix)
Mon Nov 16 22:23:37 2015  jpeach:so header_rewrite does the geoip lookup?
Mon Nov 16 22:23:43 2015  zwoop:in this patch, yes
Mon Nov 16 22:34:40 2015  jpeach:zwoop: https://github.com/apache/trafficserver/pull/190
Mon Nov 16 22:35:07 2015  zwoop:is that still needed ?
Mon Nov 16 22:39:36 2015  jpeach:checking a build
Mon Nov 16 22:45:15 2015  jpeach:Humbedooh: can you pls close pull #190?
Mon Nov 16 22:52:56 2015  shinrich1:Did we just lose sound on the hangout again?
Mon Nov 16 22:52:59 2015  biilmann_:Joined the channel
Mon Nov 16 22:53:33 2015  shinrich1:Beyond the extended unmic'ed audience discussion.
Mon Nov 16 22:54:27 2015  dxu_:I think so
Mon Nov 16 22:54:27 2015  reveller:shinrich1: yes
Mon Nov 16 22:54:49 2015  dxu_:It's back!
Mon Nov 16 22:55:21 2015  shinrich1:Ok. Will fix for real at next break. The dial in user went away. I just turned on the hangout mic
Mon Nov 16 22:55:52 2015  PSUdaemon:Joined the channel
Mon Nov 16 22:56:07 2015  ke4qqq:Joined the channel
Mon Nov 16 22:58:34 2015  shinya_:Joined the channel
Mon Nov 16 22:58:49 2015  Humbedooh:Joined the channel
Mon Nov 16 22:59:23 2015  Humbedooh:Joined the channel
Mon Nov 16 23:01:29 2015  PSUdaemon:Joined the channel
Mon Nov 16 23:08:06 2015  biilmann_:jpeach: so how do I spot you during the coffee break?
Mon Nov 16 23:08:27 2015  higgins:Joined the channel
Mon Nov 16 23:08:36 2015  jpeach:grey shirt, brown trousers
Mon Nov 16 23:13:28 2015  rhand:Joined the channel
Mon Nov 16 23:13:47 2015  biilmann_:heh, realize I'm wearing the same sweater as on my about page photo :P http://mathias-biilmann.net/about
Mon Nov 16 23:21:52 2015  kang1:Joined the channel
Mon Nov 16 23:33:33 2015  biilmann:Joined the channel
Mon Nov 16 23:46:42 2015  Uri:Joined the channel
Mon Nov 16 23:47:27 2015  Uri:Any chance of improving the audio on the hangout session? The mike keeps turning on and off....?
Mon Nov 16 23:48:14 2015  shinrich1:That better?
Mon Nov 16 23:48:33 2015  Uri:Sounds like it - - thanks :-)

Comments