IRC Logs for #circuits Tuesday, 2015-06-30

prologiccan you elaborate?00:03
*** koobs has joined #circuits05:38
y0noprologic, yep. The normal usage of systemd library to get journald log is done with that: http://www.freedesktop.org/software/systemd/python-systemd/journal.html#example-polling-for-journal-events05:50
prologicyeah kk05:51
prologicgone are the days of using tail/grep huh05:51
prologicawesome stuff!05:51
y0nowith circuits I have done something like that: http://codepad.org/frWN2Xo505:52
prologico'really?05:52
y0noI have take inspiration from File Class that use poller.05:52
prologicplease please contribute this!05:52
prologic:P05:52
y0nobut that don't work, I don't know why XD05:53
prologicahh05:53
prologicI can help you fix that probably05:53
prologichow would I test it?05:53
y0noJust need to have a virtual machine with Systemd and python :p05:54
prologicI can see one issue already05:54
prologic    @handler('ready')05:54
prologic    def __on_ready(self, comonent):05:54
prologic        self._poller.addReader(self, self.journal.get_events())05:54
prologicthe poller interface expects a file desciprtor05:54
prologicis that what jounal.get_events() is returning?05:54
y0nohttp://www.freedesktop.org/software/systemd/python-systemd/journal.html#systemd.journal.Reader.get_events05:55
prologicalso to make this JounalPoller component more effective05:55
prologicyou should try to find an existing poller in the running system and use that05:55
prologicyou cannot have two pollers05:55
prologiclook at the sockets components for further inspiration :)05:55
prologicfileno() → int05:56
prologicGet a file descriptor to poll for changes in the journal. This method invokes sd_journal_get_fd(). See man:sd_journal_get_fd(3).05:56
prologicjuse use this instead I thinik05:56
prologicand add .fileno() to the underlying poller05:56
y0noOkay =)05:57
prologicdo it now :)05:58
prologiclemme know if it works05:58
y0noYep =)05:58
y0noprologic, the first thing is that when using fileno() instead of get_events(), the script works better :D06:07
y0noprologic, http://codepad.org/6a4kcIVb seems to work well. Thanks for your help07:30
*** tverrbjelke has joined #circuits07:34
*** nizox has joined #circuits07:44
prologicy0no, so it all works now? :)08:48
*** travis-ci has joined #circuits08:58
travis-cicircuits/circuits.github.io#35 (build - 7ed3423 : James Mills): The build passed.08:58
travis-ciChange view : https://github.com/circuits/circuits.github.io/compare/40ab1ee0c7fa...7ed342324f0c08:58
travis-ciBuild details : https://travis-ci.org/circuits/circuits.github.io/builds/6894040508:58
*** travis-ci has left #circuits ()08:58
y0noprologic, yep08:59
prologicbeautiful :)09:02
prologicnow let's find a way for you to get that out there for others to use09:02
prologicand further promote how nice circuits is :)09:02
prologicblog post, tweet, a new circuits-systemd package?09:02
prologicmaybe try to make use of the namespaces support we're trying to add09:02
y0nothe namespace support ? What is it ?09:04
prologicahh09:07
prologicthat's where another package can inject a new subpackage/module into another09:07
prologicso we've declared circuits as a namespace09:07
prologicthanks to @spaceone09:07
prologicso if you created circuits-systemd09:07
prologicyou'd be able to possibly do something like09:07
prologicfrom circuits.systemd import JournalPoller09:08
prologicor similar09:08
prologic@spaceone, wouldn't we also have to declare circuits.core, circuits.net, circuits.protocols, circuits.web and circuits.tools as namespaces too?09:08
prologic@spaceone, I'd still like you to comment on https://github.com/circuits/circuits/pull/101#issuecomment-112981813 too09:09
prologicAs the way I'm thinking is; if this is purely to support users without setuptools I'm not sure we should. if you're not using setuptools you're doing it wrong? at least the whole packaging ecosystem has moved to setuptools/wheel/pip09:10
prologicdistutils is dead really at least the old distutils from Python 2.x09:10
*** travis-ci has joined #circuits09:22
travis-cicircuits/circuits.github.io#36 (build - 4d51e1f : James Mills): The build passed.09:22
travis-ciChange view : https://github.com/circuits/circuits.github.io/compare/7ed342324f0c...4d51e1fb361d09:22
travis-ciBuild details : https://travis-ci.org/circuits/circuits.github.io/builds/6894373509:22
*** travis-ci has left #circuits ()09:22
y0noprologic, something like that ? https://github.com/y0no/circuits.journald10:16
*** travis-ci has joined #circuits10:20
travis-cicircuits/circuitsweb.github.io#1 (build - 6653251 : James Mills): The build passed.10:20
travis-ciChange view : https://github.com/circuits/circuitsweb.github.io/compare/11a80e19f4d1...6653251f2b8e10:20
travis-ciBuild details : https://travis-ci.org/circuits/circuitsweb.github.io/builds/6895193710:20
*** travis-ci has left #circuits ()10:20
*** koobs has quit IRC10:24
*** koobs has joined #circuits10:25
*** travis-ci has joined #circuits10:28
travis-cicircuits/circuitsweb.github.io#1 (build - 6653251 : James Mills): The build passed.10:28
travis-ciChange view : https://github.com/circuits/circuitsweb.github.io/compare/11a80e19f4d1...6653251f2b8e10:28
travis-ciBuild details : https://travis-ci.org/circuits/circuitsweb.github.io/builds/6895193710:28
*** travis-ci has left #circuits ()10:28
*** tverrbjelke has left #circuits ("Once you know what it is you want to be true, instinct is a very useful device for enabling you to know that it is")10:51
prologicy0no, good news13:29
prologicI think I solved the memory leak13:29
y0noprologic, oh really ? :D13:29
prologicwell almost13:47
prologic:)13:47
y0noIf you need someone to test it in production tell me :)13:57
prologicy0no, https://github.com/circuits/circuits/commit/3bed08820f333239f008ba858422a37ce9918b7e13:58
prologicif you could test that :)13:59
prologicyeah13:59
*** travis-ci has joined #circuits15:10
travis-cicircuits/circuits#252 (bug-fix##54 - 1842759 : James Mills): The build is still failing.15:10
travis-ciChange view : https://github.com/circuits/circuits/compare/0762dc912eb6...18427597381f15:10
travis-ciBuild details : https://travis-ci.org/circuits/circuits/builds/6898546215:10
*** travis-ci has left #circuits ()15:10
*** Coldblackice has quit IRC15:31
*** bleep has quit IRC20:14
*** FSX has joined #circuits21:02
spaceoneprologic: namespacing of circuits.*  → i thought aboout it, too. but not sure. circuits.protocols would proably be nice21:17
spaceonebut well, i created circuits.http instead of circuits.protocols.http21:19
spaceonee.g.21:19
spaceoneso dunno if one wants to have such a long path21:19
spaceoneprologic: i currently need the following functionality:21:20
spaceonei have a event A and 2 handlers for the event21:21
spaceonehandler1 is a coroutine which yields self.wait(B())21:21
spaceonei want that handler2 is not executed immediately after this, i want that it is executed after the handler1 coroutine is finished21:22
*** Coldblackice has joined #circuits21:25
*** travis-ci has joined #circuits21:45
travis-cicircuits/circuits.github.io#37 (build - b0afe9c : James Mills): The build passed.21:45
travis-ciChange view : https://github.com/circuits/circuits.github.io/compare/4d51e1fb361d...b0afe9cbb7c521:45
travis-ciBuild details : https://travis-ci.org/circuits/circuits.github.io/builds/6904904421:45
*** travis-ci has left #circuits ()21:45
*** travis-ci has joined #circuits21:47
travis-cicircuits/circuitsweb.github.io#2 (build - 78f589d : James Mills): The build passed.21:47
travis-ciChange view : https://github.com/circuits/circuitsweb.github.io/compare/6653251f2b8e...78f589dab1ae21:47
travis-ciBuild details : https://travis-ci.org/circuits/circuitsweb.github.io/builds/6904914021:47
*** travis-ci has left #circuits ()21:47
y0noprologic, the first test about memory leak fix seems to be good22:50
y0noI put the modification on production tomorrow if I have time to see if it's really good for us22:50
prologicy0no, sweet23:24
prologicyeah not sure how to solve the other two atm23:25
prologicstill investigating23:25

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!