19:28:39 #startmeeting 19:28:40 Meeting started Tue Jul 8 19:28:39 2014 UTC. The chair is TheSnide. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:28:40 Useful Commands: #action #agreed #help #info #idea #link #topic. 19:28:43 hi all 19:28:54 #topic Agenda 19:30:05 #topic 2.2 19:30:32 some progress has been done on that part, mostly by wens. 19:31:29 He pushed the new munin-cgi-html to be almost on par with the old one. It will be merged & released as 2.1.9. 19:31:46 :) 19:32:38 The storable2sql is almost completed either. It'll also be part of 2.1.9. 19:33:53 The objective of storable2sql is to be able to use parts of 2.2 on a 2.0 install. 19:34:21 which parts is still to be defined. JSON is sure. CGI-HTML should be possible. 19:35:03 We could then rollout parts of 2.2 directly alongside 2.0 for a smooth transition. 19:35:14 (avoid bigbangs) 19:35:50 #topic munin-c 19:36:34 I worked again on munin-c, specially to provide a posix_spawn port, instead of the usual fork/exec scheme. 19:37:21 I'm still searching a win32 specialist, that could offer native packages of it there. 19:37:55 ... and ideally porting the munin-node-win32 plugins inside munin-c would be awesome. 19:38:31 that'd be interesting. Easy monitoring windows boxes from munin would be a killer app :) 19:38:53 Actually, only 1 plugin needs to be ported : the one that bridges perfmon to munin :) 19:38:54 with a "msi" install. I talked to someone at work about that today, I'll ask again tomorrow 19:39:19 (me?) 19:39:21 I have some experience with NSIS, but no MSI. 19:39:33 Kvisle: didn't want to mention anyone by name :-D 19:40:28 ssm: but yes. monitoring OSX & win32 seemlessly would be a killer app IMHO. 19:40:56 ssm: put the JSON api on top of it for some sugar topping :) 19:41:13 .... coming to ... 19:41:16 #topic json 19:41:30 The JSON api is ready-to-use. 19:41:44 so, USE IT. ABUSE IT. 19:42:36 * ssm is packaging munin-api in 2.1.9 19:42:38 I'm still not very satified with the api itself, but as i lack skills to make an UI, I don't really know what's needed to make it much better. 19:43:01 so, the API might change until 2.2. 19:43:05 It'll make people happy, as long as the API itself is sane :) 19:43:28 ... yet, the core things won't. and if you lurk here you can even design it ! 19:43:57 the thing that bothers me the most is that the ID might change every munin-update run. 19:44:10 (since we don't have a persistant DB) 19:44:23 id = checksum of group/host/service/field? 19:44:30 s/checksum/hash 19:44:49 with a configurable secret to seed it, like mojolicious already have 19:44:51 nah, id = row sequence value :) 19:44:55 ah, that 19:45:10 i'd like to either : 19:45:34 #1 change the ID generation from seq() --> clever() 19:45:59 #2 have persistant IDs 19:46:18 #3 directly use URL paths instead of node_id. 19:46:57 * ssm needs to install and use it in production for a bit first :) 19:47:18 ... #3 would mean something like : /nodes/?node_path=localnet/localhost 19:47:40 ssm: ok, i'll be rushing storable2sql then 19:48:06 #topic misc 19:48:12 so, anything else ? 19:48:49 not much. Just summer and strawberries :) 19:53:04 ssm: ok. have a look @ https://github.com/munin-monitoring/munin/pull/235 19:53:09 #endmeeting