[11:29:32] hey qchris [11:29:35] hey average [11:29:41] Morning drdee [11:29:42] qchris wanna chat to sync up [11:29:43] ? [11:29:48] i am in amsterdam now :) [11:30:03] drdee: Good afternoon then :-) [11:30:11] at least the same timezone and i can run the scrum standup [11:30:14] Gimme time to boot the Google machine. [11:30:17] tomorrow will fly to toronto [11:30:18] sure [12:26:18] ping average [13:15:34] average, if you need qchris or me to close the dclass patch set, just give us a shout [13:16:17] ok [13:28:47] average, can you reply to faidon's email as well? [13:31:42] yes [13:47:45] hey ottomata, around? [14:04:27] yoooo ottomata!! [14:04:42] whaaazzzz uuuppp? [14:05:14] YOOOO [14:05:15] drdee! [14:05:20] you are alive?!!!! [14:05:28] or are you a reincarnated bot? [14:05:36] hard to tell over this IRC thing [14:05:45] i am alive [14:05:56] just landed this morning in amsterdam from hongkong [14:06:07] wanna do a hangout? [14:06:36] shoooroorurururu [14:07:09] batvcave [14:19:22] hey qchris, can you make those two mingle cards [14:19:22] ? [14:19:32] :-) [14:19:38] average: when you will push the next patch set for dclass? [14:19:39] I came back this minute :-) [14:19:47] drdee: Sure. [14:19:52] ty [14:20:01] (I know :) ) [14:48:01] drdee, libdclass now on gallium [14:48:11] awesome! [14:49:28] has puppet been applied? [14:49:54] qchris, got a git review question [14:51:16] i would like to learn, for once and for all, how to get rid off "You have more than one commit that you are about to submit." when running git review [14:53:05] drdee: You know git. So why use git review? [14:53:24] isn't that how we are supposed to use gerrit? [14:53:24] 'git review' is to make git behave a bit more like svn :-/ [14:53:32] WHATT??????? [14:53:34] I never used git review. [14:53:40] It's a pain. [14:53:49] i don't hate many things in life, in fact almost nothing [14:53:56] but git review is the one thing i truly despise [14:54:02] Get rid of it then :-) [14:54:03] it *never* works for me [14:54:06] git review is just a useful shortcut [14:54:13] it runs straight git commands [14:54:21] but it never works :) [14:54:28] i think i've figured out how to make it work [14:54:35] how? [14:54:35] : [14:54:36] :D [14:54:44] don't do nonstandard things? [14:54:47] always use topic branches [14:54:51] :-)) [14:54:58] how to get rid off "You have more than one commit that you are about to submit." when running git review [14:54:59] rebase your topic branches often [14:55:06] that's a rebase [14:55:07] usually [14:55:35] ok so how do i do that? [14:55:52] but, you only get yourself in that position if you don't use topic branches, or if you commit more than once in a topic branch before your changes are merged in gerrit [14:56:09] so, to stick to one change per topic branch [14:56:13] whenever you need to make a change [14:56:14] git pull [14:56:17] well [14:56:19] ok [14:56:21] git checkout master [14:56:22] git pull [14:56:28] git checkout -b my_new_branch_whatever origin/master [14:56:37] (that creates a topic branch tracking origin/master) [14:56:41] then do stuff [14:56:46] git commit -a [14:56:47] git review [14:56:48] fine. [14:56:50] sure [14:56:52] make some a new patchset: [14:56:54] git commit -a --amend [14:56:55] but i already did stuff [14:56:55] git review [14:56:56] fine [14:57:01] but! [14:57:16] if you need to bring in new changes from the remote [14:57:21] you should rebase at this point, not pull [14:57:26] as long as your commit hasn't been merged [14:57:34] rebase instead of pull(merge) [14:57:36] i've got some local commits [14:57:36] so [14:57:48] as long as none of them have been merged [14:57:49] you can rebase [14:57:54] git fetch —all [14:57:54] how? [14:57:57] git rebase origin/master [14:57:58] i did that [14:58:08] (or whatevr branch you want to rebase on) [14:58:25] now, if you have multiple commits in your topic branch that have not been merged [14:58:27] $ git rebase origin/master [14:58:28] Current branch master is up to date. [14:58:33] then yes, every time you do git review [14:58:38] it will submit them both [14:58:50] it won't create a new patchset if they haven't changed though [14:59:31] so i am still stuck? [14:59:37] no, you shoudl be able to submit them [14:59:41] rebase did not change anything [14:59:41] can't you just say 'yes' [14:59:41] ? [14:59:47] what's the error? [14:59:51] drdee: To get rid of the message, use "--yes" [15:00:06] sure but then it will be a very fucked up patchset [15:00:13] with 4 different commits [15:00:20] that have nothing to do with each other [15:00:29] 45f8857 (HEAD, master) Add Maven Checkstyle [15:00:29] cf9d6ed (origin/master, origin/HEAD) Comment out Kraken ETL to make Jenkins work [15:00:30] 3235514 Merge branch 'master' of github.com:wikimedia/kraken [15:00:31] a0ca935 (dev/master) Send email on Oozie failure using xInclude [15:00:43] i only want to submit 45f8857 [15:00:50] Can you push them somewhere for us to see? [15:00:57] sure to github [15:00:58] drdee, have any of those been merged? [15:01:07] on github the first 3 [15:01:12] oh github?? [15:01:20] are you just trying to submit your most recent commit to gerrit? [15:01:25] yes [15:01:28] ok do this [15:01:36] git reset HEAD^ [15:01:38] git stash [15:01:45] git checkout -b new_topic_branch origin/master [15:01:48] git stash pop [15:01:51] git commit -a [15:01:53] git review [15:01:58] that will make a new commit on a new topic branch [15:02:06] and hopefully get you away from your other commits [15:03:48] nope [15:03:55] it still wants to submit 4 commits [15:04:26] are you pushing to github and to gerrit? [15:04:31] i mean [15:04:34] have you pushed to both places? [15:04:54] github is supposed to be a mirror, right? [15:05:23] sure, but that change only happened recently [15:05:33] and i have not yet fully incorporated that change into my habits [15:06:50] drdee: Rereading the git commands above, the 'git checkout' line is based on the wrong ref, origin/master is your github remote. [15:07:25] drdee: Can you try: [15:07:31] git checkout 8e92ba60b0698876e11d75700a327ef6f00d230b [15:07:39] git cherry-pick 45f8857 [15:07:50] git push gerrit HEAD:ref/for/master [15:07:58] (if gerrit is your gerrit remote) [15:08:28] Oh ... typo ... git push gerrit HEAD:refs/for/master [15:10:10] try this in a new branch? [15:10:48] The above commands will take you into detached mode. But you can use a separate branch: [15:10:53] ok [15:10:57] git checkout 8e92ba60b0698876e11d75700a327ef6f00d230b [15:11:02] git checkout -b my_new_branch [15:11:02] i am in detached state [15:11:09] git cherry-pick 45f8857 [15:11:27] git push gerrit my_new_branch:refs/for/master [15:12:36] (PS1) Diederik: Add Maven Checkstyle [analytics/kraken] - https://gerrit.wikimedia.org/r/78960 [15:12:50] drdee: Looks like it worked :-) [15:12:56] ty qchris, and you have the honors for the review ;) [15:13:16] hehe :-) [15:13:29] i think we will have to tweak the check style specifics in the coming weeks [15:13:32] but it's a start [15:15:03] average: you need help with wrapping up the dclass patchset? [15:19:13] heya qchris: can you recommend a java json package that will be easy to use and compile? :) [15:19:58] drdee, i think if the mirror has been setup [15:20:02] you should never push to git hub again [15:20:09] you should just make a clean clone from gerrit [15:20:14] and work with that from now on [15:20:31] right i know but that's why i said "i have not yet fully incorporated that change into my habits" [15:40:06] ottomata: Sorry, I missed your ping. Gerrit for example uses Gson. [15:51:05] hmm, Gson, ok [15:51:40] ok qchris, can you help me figure out how to add gson to my maven dependencies? [15:52:24] oo, might have found it [15:52:27] https://sites.google.com/site/gson/gson-user-guide/using-gson-with-maven2 [15:52:42] ottomata: Oh. I've been too slow :-) [15:53:11] that is exaclty what I found [15:53:12] :) [15:53:18] thank you google and qchris! [15:53:37] what's compile? [15:54:13] Thereby you can restrict when dependencies are used [15:54:27] hmm [15:54:45] https://maven.apache.org/guides/introduction/introduction-to-dependency-mechanism.html#Dependency_Scope [15:54:48] do I want compile? i guess so, because then it will be compiled into my jars? [15:55:15] Yes, compile should be correct. [15:55:22] It's the default anymays Iirc. [15:55:26] Let me check ... [15:55:40] Yes, it's the default. [15:56:29] "i guess so, because then it will be compiled into my jars?" I am not sure what exactly you mean by "into". To get sommething int, you'll need to use shaded jars. [15:57:03] camus does [15:57:14] Ok. Than that should just work out of the box. [15:57:29] why gson over something else? i just want something real easy right now. i just want to parse json so I can read the timestamp field [15:58:21] Gson gets the job done :-) [15:58:28] It's not like all other libraries are bad. [15:58:42] i was just expecting to see a really simple usage example, hehh, haven't found one yet [15:59:08] to parse json? [15:59:38] reading though [16:01:03] Sorry. I do not understand. [16:02:21] i meant, i am reading docs... [16:02:23] i think i found it [16:02:33] Ok. [16:02:38] its easy, the user guide just was complicated, the api was easy to figure out [16:02:49] it looked like gson was better for serializing java objects [16:02:50] but i found it [16:19:26] (CR) QChris: [C: -1] "(5 comments)" [analytics/kraken] - https://gerrit.wikimedia.org/r/78960 (owner: Diederik) [16:53:26] ok qchris, i think I need some more help :) [16:53:35] I've gotten somewhere [16:53:36] K. [16:53:42] now I'm getting that JobContext error again :/ [16:53:51] Oh :-( [16:53:53] i was hoping if I just redid what I wanted to do that I would avoid it [16:53:56] but nope! [16:54:04] but! i do understand the class hierarchy much better now [16:54:10] i mostly understand what should be done [16:54:10] That's great! [16:54:23] hop in batcave? [16:54:26] 5 mins before standup? [16:54:37] I'll hop in the batcave. [16:54:45] Lot's see if we can get it done before... [16:54:49] hehe [16:54:52] probably not :) [16:55:20] it compiles so slowy in labs [16:55:26] i'm hoping that if I compile on that machine it will go away too :0 [16:59:09] average, scrum? [17:19:47] actually [17:20:17] ... [17:21:00] ottomata: this is one of the kafka 0.8 release blockers: https://issues.apache.org/jira/browse/KAFKA-990 [18:22:30] (PS2) Diederik: Add Maven Checkstyle [analytics/kraken] - https://gerrit.wikimedia.org/r/78960 [18:25:01] (CR) Diederik: "(4 comments)" [analytics/kraken] - https://gerrit.wikimedia.org/r/78960 (owner: Diederik) [18:31:22] in kraken we use jackson [18:31:35] and that's a really good one [18:35:08] (PS3) Diederik: Add Maven Checkstyle with default settings and disable Storm ETL module [analytics/kraken] - https://gerrit.wikimedia.org/r/78960 [18:36:03] (CR) Diederik: "(1 comment)" [analytics/kraken] - https://gerrit.wikimedia.org/r/78960 (owner: Diederik) [18:37:58] ottomata [20:56:47] (CR) QChris: "(1 comment)" [analytics/kraken] - https://gerrit.wikimedia.org/r/78960 (owner: Diederik) [21:01:58] qchris, you still around to look at a java thang? [21:02:05] sure. [21:02:38] music is llouuud in here! [21:02:41] in batcave [21:02:46] Booting into batcave... [21:02:58] Yeah loud music! Party! \o/ [21:03:47] notorius B.I.G. [21:03:56] i htink the cafe is closing [21:03:58] Oh ... wrong music :-) [21:04:00] so they turni t up real loud [21:04:01] haha [21:04:07] naw, i don't mind, although not so good for working [21:50:58] (CR) QChris: [C: -1] "(4 comments)" [analytics/kraken] - https://gerrit.wikimedia.org/r/78960 (owner: Diederik) [22:51:28] (PS1) Milimetric: default to sqlite for testing [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/79024 [22:52:11] (CR) Milimetric: [C: 2 V: 2] default to sqlite for testing [analytics/wikimetrics] - https://gerrit.wikimedia.org/r/79024 (owner: Milimetric)