[13:12:41] erosen, around? [13:12:45] average, around? [13:12:51] milimetric: moooorning! [13:26:57] mooorning milimetric [13:27:04] hello doctor [13:28:25] dee [13:31:28] yoooo [13:55:06] hey drdee, this is awesome: http://micheles.googlecode.com/hg/decorator/documentation.html#decorator-is-a-decorator [13:55:20] it's a library in python that makes decorators less annoying to write [13:56:39] cool! [14:00:23] oh did you see my opinion about the meetings next week drdee? [14:00:24] what do you think? [14:01:32] no [15:26:34] drdee: let me know if you want to chat about making the zero rollups or dashboards [15:26:47] erosen join the hangout! [15:26:51] will do [16:19:53] milimetric i pmed you [16:42:35] drdee: kul is ready [16:42:38] can you meet in hangout [16:42:52] k coming [17:02:15] drdee: hangout [17:02:27] soryr [17:12:47] hey milimetric - got my notes? [17:13:00] yes DarTar [17:13:07] gr8 [17:13:13] thank you [17:13:20] I haven't had time to look at them [17:13:26] but I will today [17:13:54] np, I'm going to port a few more graphs and maybe add new ones in preparation for the VE test next week [17:14:33] the main issue at the moment is fine-tuning/setting defaults for all dashboards [17:15:09] it'd be practical for one to have a handle to toggle off globally datasource labels from legends [17:15:52] anyway, no rush - ping me if you wish/have time for a short chat later [17:21:02] hi drdee , milimetric , rounce123 , ottomata [17:21:20] yo [17:21:27] Hey we are on the standup can you join? [17:22:12] noisy place here, can't hear or speak because of it [17:22:20] I can write though [17:23:25] ok, please give a quick update [17:24:29] quick update: I added .shlibs and .links to the package [17:24:37] it is ready for another review from Andrew [17:24:52] I've followed all the needed guidelines from Andrew so far [17:25:31] I believe it's ready for a review from Faidon (unless Andrew thinks there are other things to change/add) [17:26:52] what's the link to gerrit? [17:27:57] drdee: Andrew told me there needs to be a patchset with only the debian specific changes [17:28:04] drdee: I will make a gerrit patchset now [17:28:07] ok [17:28:11] drdee: the branch is called debian [17:28:50] drdee: https://git.wikimedia.org/log/analytics%2Fdclass.git/refs%2Fheads%2Fdebian [17:32:02] can a gerrit patchset be made for a different branch than master ( what I mean is, can the gerrit patchset, be merged into a different branch than master) ? [17:42:36] 19:38 < AzaToth> average_drifter: git review [17:42:40] ok looks like it's possible [17:42:44] I'm gonna use it [17:49:39] I'd like more permissions please on the dclass repo [17:50:08] why? just use git review [17:50:29] drdee: well, I've already just pushed to that repo [17:50:38] so now I need to delete the branch and submit it as a gerrit patchset [17:51:09] a remote branch needs to be deleted? [17:51:13] yes [17:51:16] what's the name? [17:51:17] the branch's name is debian [17:51:22] ok i will delete it [17:51:25] you are sure? [17:51:25] thank you [17:51:29] I am 100% [17:51:52] done [17:51:56] thank you [17:56:07] drdee: https://gerrit.wikimedia.org/r/#/c/68703/ [17:56:10] drdee: please merge this patchset [17:56:16] drdee: after this will follow the debian patchset [17:56:42] done [17:56:45] thanks [17:56:49] will make the debian patchset now [18:16:05] New patchset: Stefan.petrea; "Debianization" [analytics/dclass] (debian) - https://gerrit.wikimedia.org/r/68711 [18:16:18] getting there.. [18:16:23] have to add the changes now [18:16:26] gerrit is tricky [18:17:33] New patchset: Stefan.petrea; "Debianization to conform with Ops requirements" [analytics/dclass] (debian) - https://gerrit.wikimedia.org/r/68711 [18:19:12] drdee: added you, Andrew and Faidon to the review above ^^ [18:19:39] thx [18:20:45] I am looking for new tasks at this moment [18:21:00] if you have any on the wall which you think I could tackle please let me know [18:33:28] drdee: are those the results of the latest run, or did you run into an issue? [18:33:46] i ran into issues; same data as yesterday's analysis [18:33:51] cool [18:34:05] stupid pig [18:34:21] also regarding the digi-malysia point, if you are only counting x-cs headers, they should only be set for opera UAs, because varnish checks for that IIRC [18:34:42] right, but i haven't verified that [18:34:52] so actually that number could be correct [18:35:48] although the varnish regex for Opera is very liberal; it might incorrectly identify user agent strings :) [18:36:08] kul asked for caveats [18:36:15] so i threw in a bunch [18:37:47] very true [18:38:00] totally reasonable, I just figured i would mention it [18:41:49] thx! [18:52:27] can I be of help ? [18:52:36] in some form to someone ? :) [19:02:39] yes [19:03:11] please pick up card 738 [19:03:14] ok [19:04:52] oh, for 738 csv is being generated [19:05:00] so it's done [19:05:18] please email me and erik and update us on the status and what the next stepsare [19:17:11] drdee: update sent on 738 [19:19:04] ty [19:21:30] np [19:21:49] can't you fix the URL issue by making it a variable in the config file [19:21:49] ? [19:22:41] drdee: the config files of Wikistats ? [19:22:47] yes [19:29:10] drdee: it would be very hard, since it would involve WikiReports.pl which from what I know .. has no config file [19:29:32] can you verify that with Erik? [19:29:32] hm, but for the squid reports it would be possible [19:29:38] drdee: yes [19:29:45] or ask him for advice [19:29:50] alright [19:43:18] average_drifter: why are you hardcoding the paths in the report? why not use relative urls? then you don't have this problem [19:50:25] https://github.com/wikimedia/analytics-wikistats/blob/master/squids/perl/SquidReportArchive.pl#L151 [19:50:25] https://github.com/wikimedia/analytics-wikistats/blob/master/squids/perl/SquidReportArchive.pl#L633 [19:50:28] https://github.com/wikimedia/analytics-wikistats/blob/master/squids/perl/SquidReportArchive.pl#L2601 [19:50:31] https://github.com/wikimedia/analytics-wikistats/blob/master/squids/perl/SquidReportArchive.pl#L3335 [19:50:34] https://github.com/wikimedia/analytics-wikistats/blob/master/squids/perl/SquidReportArchive.pl#L3511 [19:50:37] https://github.com/wikimedia/analytics-wikistats/blob/master/squids/perl/SquidReportArchive.pl#L3990 [19:50:40] https://github.com/wikimedia/analytics-wikistats/blob/master/squids/perl/SquidReportArchive.pl#L4606 [19:50:43] https://github.com/wikimedia/analytics-wikistats/blob/master/squids/perl/SquidReportArchive.pl#L4785 [19:50:49] these links confirm that the paths are not hardcoded