Wednesday, 2014-01-08

*** ruhe is now known as _ruhe00:21
*** SergeyLukjanov is now known as _SergeyLukjanov00:37
*** asalkeld has joined #openstack-ceilometer00:40
openstackgerritA change was merged to openstack/ceilometer: Sync config generator workaround from oslo  https://review.openstack.org/6329100:57
*** ozialien has joined #openstack-ceilometer01:09
*** herndon has quit IRC01:11
*** yfujioka has joined #openstack-ceilometer01:21
*** ozialien has quit IRC01:57
openstackgerritJohn Herndon proposed a change to openstack/python-ceilometerclient: Support the Event API  https://review.openstack.org/5458202:01
openstackgerritliusheng proposed a change to openstack/ceilometer: Fix use the fact that empty sequences are false  https://review.openstack.org/6476802:03
openstackgerritA change was merged to openstack/ceilometer: Handle the metrics sent by nova notifier  https://review.openstack.org/4283802:14
*** tongli has quit IRC02:18
*** tianst20 has quit IRC02:29
*** dhellmann has quit IRC02:55
*** Alexei_987 has quit IRC03:22
*** ozialien has joined #openstack-ceilometer03:44
*** flwang has quit IRC04:56
*** flwang has joined #openstack-ceilometer04:56
*** flwang has quit IRC04:58
*** dhellmann has joined #openstack-ceilometer05:04
openstackgerritA change was merged to openstack/ceilometer: Avoid false negatives on message signature comparison  https://review.openstack.org/6486305:13
*** urulama has joined #openstack-ceilometer05:19
*** urulama has quit IRC05:23
*** dhellmann has quit IRC05:31
*** dhellmann has joined #openstack-ceilometer05:42
*** flwang has joined #openstack-ceilometer05:44
*** ozialien has quit IRC05:47
openstackgerritJenkins proposed a change to openstack/ceilometer: Imported Translations from Transifex  https://review.openstack.org/6280806:14
openstackgerritA change was merged to openstack/ceilometer: AlarmChange definition added to doc/source/webapi/v2.rst  https://review.openstack.org/6511206:16
openstackgerritA change was merged to openstack/ceilometer: 1st & last sample timestamps in Resource representation  https://review.openstack.org/6511806:17
openstackgerritA change was merged to openstack/ceilometer: Utilizes assertIsNone and assertIsNotNone  https://review.openstack.org/5110006:17
*** ildikov has quit IRC06:29
openstackgerritYuuichi Fujioka proposed a change to openstack/ceilometer: Implements monitoring-network-from-opendaylight  https://review.openstack.org/6389006:33
*** SergeyLukjanov has joined #openstack-ceilometer07:05
*** ildikov has joined #openstack-ceilometer07:28
*** urulama has joined #openstack-ceilometer07:33
*** Alexei_987 has joined #openstack-ceilometer07:48
*** nadya_ has joined #openstack-ceilometer08:10
*** yfujioka has quit IRC08:19
*** nadya_ has quit IRC08:25
*** coolsvap has joined #openstack-ceilometer08:30
gibiboris-42: ping08:32
boris-42gibi png08:32
gibiboris-42: hi! is there a reason in the mongo query syntax to have $and and $or written in prefix way and $gt and other in infix way?08:33
boris-42gibi not so sure=)08:34
gibiboris-42: We just thinking about making every operator prefix for consistency08:34
gibiboris-42: somthing like {"and": [{'=': [{"field", 13}, {'field': 14}]}08:35
gibiboris-42: what do you think about this syntax?08:40
*** SergeyLukjanov is now known as _SergeyLukjanov08:44
*** _SergeyLukjanov has quit IRC08:44
boris-42gibi probably we should use '$' for operators?08:46
gibiboris-42: we think that "<=" "!=" is more readable than $lte $ne Moreover "=" does not have an operator in mongo syntax which is also strange inconsistency for me08:51
boris-42gibi probably some legacy code is the reason of mongo syntax=)08:51
gibiboris-42 we don't want to follow mongo syntax blindly due to these inconsitency08:54
boris-42gibi okay but the major idea of mongo syntax is ok08:56
boris-42gibi because it could be validate by jsonschema08:56
boris-42gibi so we can take from it best=)08:57
gibiboris-42 I dont have jsonschema experience but I guess the above prefix example is close enough to the mongo to be able to validate08:57
boris-42gibi I think yes08:58
boris-42but why "=" works as "and" ?)08:59
s2r2hi.  Could someone please enlighten me wrt ceilometer's api (stable/havana branch)?  I get an array of things like this one https://gist.github.com/srenatus/8313800 when I GET /v2/meters.  What makes me wonder is the "\n" inside the meter_id; plus I cannot find any way to actually get a meter via GET /v2/meters/$meter_id (only [])08:59
*** SergeyLukjanov has joined #openstack-ceilometer08:59
gibiboris-42 I don't get your last question. Could you elaborate a bit?08:59
boris-42{"and": [{"=": {"field", 13}}, {"=": {'field': 14}} ]}09:00
boris-42gibi ^ why not such syntax?09:01
*** nadya_ has joined #openstack-ceilometer09:03
gibiboris-42 two things 1) consistent prefix syntax for operators 2) have human readable "=" "<=" etc09:03
boris-421) but I saved "="09:03
boris-422) just repeat it 2 times instead of 1 so "and" becomes more attractive09:03
gibiahh you mean, you want something like {"=": [{field: 13}, {field: 14}]} ?09:05
boris-42gibi I mean:  {"and": [{"=": {"field", 13}}, {"=": {'field': 14}} ]}09:06
boris-42gibi is better09:06
gibiboris-42 I see, you are right that is better09:07
boris-42gibi yep because you can use > or < >= or <= in the same list of operations09:07
gibiboris-42 ok, then your last example is somthing you can be happy with?09:10
boris-42gibi it seems good09:10
gibiboris-42 great, then we will have think about the implementation of this syntax with ildikov today afternoon09:13
boris-42gibi nice09:15
ildikovgibi, boris-42: it seems to be good to me too, at the afternoon we will start the design based on this to see if it will be good from every point of view or not09:17
*** nadya_ has quit IRC09:19
*** jhurlbert has quit IRC09:20
*** nadya_ has joined #openstack-ceilometer09:21
openstackgerritMehdi Abaakouk proposed a change to openstack/ceilometer: [dont review] mysql gate tests  https://review.openstack.org/6543409:21
openstackgerritMehdi Abaakouk proposed a change to openstack/ceilometer: Run unit tests against MySQL  https://review.openstack.org/5948909:21
openstackgerritMehdi Abaakouk proposed a change to openstack/ceilometer: [dont review] mysql gate tests  https://review.openstack.org/6543409:24
*** jhurlbert has joined #openstack-ceilometer09:30
*** eglynn has joined #openstack-ceilometer09:31
*** nadya_ has quit IRC09:33
s2r2ok, rtfm'ing helped.09:36
silehtjd__, around ?09:49
jd__sileht: yes09:52
jd__(not sure if i was connected)09:52
jd__sileht: missed you yesterday, want to hang out?09:52
*** flwang has quit IRC09:52
silehtjd__, yup09:53
*** nadya_ has joined #openstack-ceilometer10:01
*** SergeyLukjanov has quit IRC10:17
*** nadya_ has quit IRC10:18
*** SergeyLukjanov has joined #openstack-ceilometer10:22
*** SergeyLukjanov has quit IRC10:23
*** nadya_ has joined #openstack-ceilometer10:28
*** nadya_ has quit IRC10:34
*** nadya_ has joined #openstack-ceilometer10:35
*** nadya_ has quit IRC10:45
*** SergeyLukjanov has joined #openstack-ceilometer10:58
*** jhurlbert has quit IRC11:15
*** jhurlbert has joined #openstack-ceilometer11:19
openstackgerritEugeniya Kudryashova proposed a change to openstack/python-ceilometerclient: Use Resource() class from common Oslo code  https://review.openstack.org/6288311:42
*** urulama has quit IRC11:58
*** yassine_ has joined #openstack-ceilometer12:33
*** thomasem has joined #openstack-ceilometer12:38
*** coolsvap has quit IRC12:47
*** yassine_ has quit IRC13:15
*** yassine_ has joined #openstack-ceilometer13:16
*** yassine_ has quit IRC13:16
*** yassine_ has joined #openstack-ceilometer13:17
*** yassine_ has quit IRC13:17
*** yassine_ has joined #openstack-ceilometer13:19
*** yassine_ has quit IRC13:19
*** yassine has joined #openstack-ceilometer13:21
*** sandywalsh_ has joined #openstack-ceilometer13:27
*** sandywalsh_ has quit IRC13:31
*** yassine has quit IRC13:31
*** yassine has joined #openstack-ceilometer13:33
*** sandywalsh_ has joined #openstack-ceilometer13:36
*** eglynn is now known as eglynn-lunch13:50
*** eglynn-lunch has quit IRC13:55
thomasemHey everyone!13:59
openstackgerritJulien Danjou proposed a change to openstack/ceilometer: requirements: drop netaddr  https://review.openstack.org/6546913:59
*** yassine has quit IRC14:01
*** jdob has joined #openstack-ceilometer14:01
*** yassine has joined #openstack-ceilometer14:01
*** sandywalsh_ has quit IRC14:04
*** nadya_ has joined #openstack-ceilometer14:05
*** prad_ has joined #openstack-ceilometer14:08
*** sandywalsh_ has joined #openstack-ceilometer14:14
openstackgerritJulien Danjou proposed a change to openstack/ceilometer: tests: pass /dev/null as config for mongod  https://review.openstack.org/6547314:20
*** nadya_ has quit IRC14:34
*** ildikov has quit IRC14:37
*** ildikov has joined #openstack-ceilometer14:38
*** prad_ has quit IRC14:40
*** flwang has joined #openstack-ceilometer14:41
*** eglynn-lunch has joined #openstack-ceilometer14:45
*** herndon_ has joined #openstack-ceilometer14:46
*** eglynn-lunch is now known as eglynn14:58
eglynnanyone remember the purpose of the meter attribute on the Resource storage model?14:58
eglynnin the mongo case, this ends up as an extremely long and repetitive list of meter (name,type,unit) tuples14:59
eglynn... one for every single sample!14:59
eglynn(as the $push operator in the mongo aggregation isn't dupe-avoiding)14:59
*** gordc has joined #openstack-ceilometer14:59
eglynnthen IIUC it seems we just ahead and drop the meters on the floor14:59
eglynn... as the link hrefs in the API Resource representation are recreated from a *separate* query on the storage driver15:00
eglynnwould anyone object if that logic was blown away from get_resouces on all storage drivers?15:00
thomasemeglynn, So, the storage/models.py Resource model?15:01
eglynnthomasem: yeap15:01
thomasemI was of the understanding that that was intended to be a list to be turned into links in an API response15:01
thomasemI think I asked that question a while back and that was the response (looking for e-mail).15:02
eglynnthomasem: here's where it's set in the mongo case https://github.com/openstack/ceilometer/blob/master/ceilometer/storage/impl_mongodb.py#L66415:02
eglynnthomasem: the result is that len(resource.meter) == number of samples for that resource15:03
eglynnthomasem: AFAICS its dropped on the floor because the API resource representation doesn't have an analogous attribute15:04
eglynn... https://github.com/openstack/ceilometer/blob/master/ceilometer/api/controllers/v2.py#L97715:04
thomasemhttp://docs.openstack.org/developer/ceilometer/webapi/v2.html#Resource15:05
thomasemoh wait15:05
thomasemnumber of samples? yikes15:05
eglynnyikes indeed :)15:06
eglynnthomasem: re. that link to the API docco, were you thinking of the "links" field in Resource?15:06
thomasemYeah, I thought that the meters attribute on a resource model returned from the DBAPI would translate to links.15:07
thomasemSince those links were linking to the various counters (meters) for the resource in the sample output.15:07
eglynnthomasem: yeah that would be more logicial15:07
eglynnthomasem: ... but the links are actually populated separately via a subsequent call back into the storage driver get_meters15:07
thomasemOh? well then15:08
eglynnthomasem: yeah that's my read of this line https://github.com/openstack/ceilometer/blob/master/ceilometer/api/controllers/v2.py#L103515:09
thomasemOh, jesus? it actually does a full get_meters call for every resource request15:09
eglynnyeap ... well constrained by resource ID15:10
thomasemSure15:10
eglynnso actually mutliple get_meters calls, one per resource15:10
eglynnyikesers!15:11
*** herndon_ has quit IRC15:11
thomasemYep15:11
eglynnso two possible approaches ...15:11
eglynn1. just stop populating the Resource.meter in the storage model15:12
eglynn2. ensure the Resource.meter list is dupe-free and stop the multiple calls into get_meters for each resource ID15:12
* eglynn suspects the out-of-memory condition on GET /v2/resources was mostly caused by the len of the Resource.meter lists15:13
*** terriyu has joined #openstack-ceilometer15:13
eglynn(as opposed to the usage of the mongo aggregation framework)15:13
eglynn(... WRT to this bug https://bugs.launchpad.net/ceilometer/+bug/1262571 )15:14
thomasemI would prefer 1, trying to think of a way to tidy up the meter calls in order to build links too.15:15
thomasemBecause the meters list appears to be a DBAPI trying to be a web API.15:16
thomasemAnd that seems to blur some responsibilities.15:16
eglynnyeap15:16
thomasemI need coffee? been writing this silly random query generator for storage testing.15:20
thomasembrb15:20
eglynnin any case, I'll file a new bug to capture this badness, since it's specific to the mongodb driver15:20
thomasemeglynn, I think we could drop it in sqlalchemy as well: https://github.com/openstack/ceilometer/blob/master/ceilometer/storage/impl_sqlalchemy.py#L471-L47615:23
thomasemI figure that's an additional join that need not happen if the API is doing another call anyway15:23
eglynnthomasem: yeap, sorry I meant to say it's *not* specific to the mongodb driver15:23
eglynnthomasem: (whereas the existing bug https://bugs.launchpad.net/ceilometer/+bug/1262571 is mongo-specific)15:24
thomasemOhhh, haha. No worries at all!15:24
thomasemeglynn, Ireland gave me a cold.15:25
thomasem:P15:25
*** prad_ has joined #openstack-ceilometer15:26
*** flwang has quit IRC15:33
*** flwang has joined #openstack-ceilometer15:36
*** flwang has quit IRC15:43
eglynnthomasem: be thankful you got away with only a cold ;)15:43
eglynnthomasem: when did you visit?15:43
thomasemeglynn, Past two weeks, over the holidays.15:43
eglynnthomasem: you got to experience some dramatic weather so!15:44
eglynnthomasem: business or pleasure or both?15:44
*** herndon_ has joined #openstack-ceilometer15:45
thomasemeglynn, pleasure, was there with my roommate (he has family over there and as such I have many many friends now) :P15:45
eglynncool!15:45
thomasemYeah, we stayed at Clontarf Castle, which I had no idea had so many weddings going on all of the time.15:46
thomasemI'd stayed in city centre the last two times I was there.15:46
eglynnthomasem: yeah that's a hotspot for folks getting hitched15:46
*** coolsvap has joined #openstack-ceilometer15:47
eglynnthomasem: ... also a lot of people plan weddings for the holiday season when family from abroad are more likely to be home to visit15:47
thomasemApparently. Made it rather difficult to get a quiet drink and a fire most of the time, which was kind of why I wanted to stay in Clontarf as opposed to city centre.15:47
thomasemyeah15:47
thomasemBut, it was still fun, despite getting a bit sick.15:47
thomasemWhich area do you live in?15:48
eglynnyeah, the Irish wedding clich├ęs (drunk father-of-the groom, bridesmaids on the tiles, cover band belting our 60s numbers etc.) can get a bit old15:49
thomasemLol15:49
eglynn... I'm in Dublin too, other side of the bay from Clontarf15:49
thomasemOhhhh, neat15:50
thomasemDon't think I've been down on that side15:51
eglynnpretty anywhere along the coast in Dublin is nice15:51
eglynn(... as long as you don't mind being flooded every so often)15:52
thomasemHaha, keep the sandbags handy15:53
*** jd__ has quit IRC15:54
*** jd__ has joined #openstack-ceilometer15:55
*** flwang has joined #openstack-ceilometer15:56
openstackgerritA change was merged to stackforge/pycadf: Python 3: the request body should be bytes in test_api.py  https://review.openstack.org/6527915:58
*** ildikov has quit IRC15:59
openstackgerritA change was merged to stackforge/pycadf: Python3: use six.moves.urllib.parse instead of urlparse  https://review.openstack.org/6528016:02
*** ildikov has joined #openstack-ceilometer16:02
eglynnthomasem: FYI I've captured that earlier discussion here https://bugs.launchpad.net/ceilometer/+bug/126716216:12
thomasemeglynn, Gotcha. Great, thanks for doing that! :)16:14
thomasemeglynn, Also, regarding the ceilometer client update for events, this is the nested formatting I wrote for the traits when an event is returned: http://paste.openstack.org/show/60808/16:16
thomasemI kind of wanted to get input on whether that was a bad idea or not. Not sure of the standards around the openstack client output.16:16
thomasemThe few folks I've asked have said it looks fine. herndon_ was thinking it looked a bit cramped and I agree, but I feel there's an absence of other options that really delineate the blocks appropriately.16:17
eglynnthomasem: hmmm, interesting16:17
eglynnthomasem: so I guess it's kinda similar to what I need to do for alarm history in the CLI for Hanava16:18
eglynnthomasem: i.e. multiple change events per alarm, each with a type and desc16:18
thomasemYep16:19
eglynnthomasem: I ended with something simpler, without the nested boxing16:19
* eglynn digs for an example16:19
thomasemAh, neat. Okey dokey16:19
eglynnthomasem: ... search for alarm-history in http://openstack.redhat.com/CeilometerQuickStart16:20
eglynnnah I described that badly16:20
*** nadya_ has joined #openstack-ceilometer16:20
eglynnnot "multiple change events per alarm" more like "multiple fields for an individual change event"16:21
thomasemSo, that looks like the single resource 'get' property/value table. Is that what you're referring to?16:21
thomasemOh here we go16:22
thomasemI see? it's the same thing without the borders16:22
eglynnthomasem: in the Detail section for the creation event for example, there are multiple fields16:23
*** SergeyLukjanov has quit IRC16:23
eglynnthomasem: ... yep, same kind of thing without nested borders16:23
*** Alexei_987 has quit IRC16:23
thomasemI guess it's a UX concern.16:23
thomasemIn this example it's easy to see what belongs where, but that's because it's only 3 alarms? what if it were 15?16:24
thomasemThings would start to blur together. That's why on a web page it's usually a good idea to zebra-stripe16:24
thomasemThe purpose for the nested traits was to create a lot of distinction, of course at the cost of it being a little offset.16:25
thomasemsorry, nested borders*16:25
eglynnalarm-history always only applies to a single alarm16:25
eglynn(that's just how I implemented the CLI)16:26
thomasemI suppose in the case of events the traits define the event.16:26
eglynn... but that alarm can have many change events16:26
thomasemYeah16:26
eglynnyeah, so the cases are not really that similar16:26
thomasemI'd be happy with doing either, I just wasn't sure if there was some standard about it (I think there ought to be).16:26
thomasemGoing from client to client can be a jarring experience without it.16:27
thomasemBut, with configurable events, we're going to be having 20 traits per event16:28
thomasemand things like that...16:28
thomasemSo, idk what we ought to do there.16:28
eglynndhellmann might have an idea re. the state of the art on existing/emerging standards for such things16:29
eglynnthomasem: ^^^ (if this issue was encountered in cliff ... https://github.com/dhellmann/cliff )16:29
*** herndon_ has quit IRC16:29
thomasemCool, thanks. I'll chat with him about it. I can't remember - that was months ago when I was dusting off these branches.16:30
*** sandywalsh_ has quit IRC16:31
*** sayali has joined #openstack-ceilometer16:45
*** ildikov has quit IRC16:49
*** nadya_ has quit IRC16:51
*** nadya_ has joined #openstack-ceilometer16:53
*** ildikov has joined #openstack-ceilometer16:53
*** terriyu has quit IRC16:54
*** tongli has joined #openstack-ceilometer16:56
*** SergeyLukjanov has joined #openstack-ceilometer17:00
* dhellmann perks up his ears and reads the backlog17:15
dhellmannthomasem: the unfiied client expects the command to return the data and allow the framework to format it, so the user can chose different output formats17:17
dhellmannthomasem: I'm not sure how well that nesting will work under that model, so we'll have to experiment a bit17:18
dhellmannthomasem: s/unified client/unified command line interface/17:18
*** herndon has joined #openstack-ceilometer17:28
*** nadya_ has quit IRC17:28
*** herndon has quit IRC17:33
openstackgerritJulien Danjou proposed a change to openstack/ceilometer: tests: kill all started processes on exit  https://review.openstack.org/6550917:34
openstackgerritJulien Danjou proposed a change to openstack/ceilometer: tests: allow to skip if no database URL  https://review.openstack.org/6508117:34
openstackgerritJulien Danjou proposed a change to openstack/ceilometer: storage: bases of a Cassandra driver  https://review.openstack.org/6277917:34
*** nadya_ has joined #openstack-ceilometer17:34
*** herndon_ has joined #openstack-ceilometer17:41
*** ozialien has joined #openstack-ceilometer17:43
thomasemdhellmann, Thanks for the reply. Yeah, It's not terribly difficult to change if it turns out not to work in the future. So, I think I understand what you mean, I saw that the implementor could supply a formatter function for the output - is that what you're referring to?17:45
*** _ruhe is now known as ruhe17:53
*** ozialien has quit IRC17:54
*** herndon_ has quit IRC17:59
*** jmckind has joined #openstack-ceilometer18:01
*** thomasem has quit IRC18:01
openstackgerritA change was merged to openstack/ceilometer: Fixed spelling mistake  https://review.openstack.org/6530718:03
*** thomasem has joined #openstack-ceilometer18:09
*** sayali has quit IRC18:16
dhellmannthomasem: cliff uses a set of plugins to format the output, so the command plugin returns python objects with a standard schema (columns, values, etc.) and the formatter processes them18:16
dhellmannthomasem: if a command plugin is calling print or otherwise formatting data, then it's doing something wrong18:17
*** ildikov has quit IRC18:17
dhellmannthomasem: e.g., see http://git.openstack.org/cgit/openstack/python-openstackclient/tree/openstackclient/image/v2/image.py for the commands to list and show images18:18
dhellmannthomasem: this case with traits, where we might not know all the names up front, may need some additional work in the framework18:19
*** ozialien has joined #openstack-ceilometer18:23
*** sandywalsh has quit IRC18:54
*** dmsimard has joined #openstack-ceilometer19:00
dmsimardIs this channel more about development or can I ask a question about an issue i'm having ?19:00
*** ildikov has joined #openstack-ceilometer19:02
*** sayali has joined #openstack-ceilometer19:06
*** sandywalsh has joined #openstack-ceilometer19:07
*** herndon_ has joined #openstack-ceilometer19:12
*** ruhe is now known as _ruhe19:21
*** dmsimard has left #openstack-ceilometer19:33
*** nadya_ has quit IRC19:34
*** _ruhe is now known as ruhe20:14
*** sayali has quit IRC20:36
*** eglynn has quit IRC20:39
*** jdob has quit IRC20:39
*** jdob has joined #openstack-ceilometer20:42
*** ozialien has quit IRC21:05
*** prad has joined #openstack-ceilometer21:24
*** prad_ has quit IRC21:26
*** SergeyLukjanov_ has joined #openstack-ceilometer21:34
*** SergeyLukjanov has quit IRC21:35
*** jdob has quit IRC21:42
*** jdob has joined #openstack-ceilometer21:43
*** kwhitney has left #openstack-ceilometer21:54
*** jdob has quit IRC22:08
*** SergeyLukjanov_ has quit IRC22:08
*** eglynn has joined #openstack-ceilometer22:09
*** SergeyLukjanov has joined #openstack-ceilometer22:10
*** thomasem has quit IRC22:16
*** SergeyLukjanov_ has joined #openstack-ceilometer22:47
*** SergeyLukjanov_ has quit IRC22:47
*** SergeyLukjanov has quit IRC22:48
*** SergeyLukjanov_ has joined #openstack-ceilometer22:48
*** eglynn has quit IRC22:48
*** SergeyLukjanov_ has quit IRC22:48
*** ozialien has joined #openstack-ceilometer22:56
*** ozialien has left #openstack-ceilometer23:03
*** ozialien has quit IRC23:04
*** ozialien has joined #openstack-ceilometer23:04
*** ozialien has quit IRC23:04
*** jmckind has quit IRC23:20
*** yassine has quit IRC23:21
*** asalkeld_ has joined #openstack-ceilometer23:29
*** asalkeld_ has quit IRC23:30
*** asalkeld_ has joined #openstack-ceilometer23:31
*** asalkeld has quit IRC23:32
*** asalkeld_ has quit IRC23:36
*** ruhe is now known as _ruhe23:41
*** prad has quit IRC23:41
*** gordc has quit IRC23:43
*** herndon_ has quit IRC23:46

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!