November 4, 2014 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | >
November 4, 2014 [00:02:00] *** marlinc has joined #gerrit[00:10:12] *** qchris has quit IRC[00:49:57] *** firemanxbr has quit IRC[00:59:31] *** jrnieder has joined #gerrit[00:59:51] *** jrnieder has left #gerrit[00:59:55] *** jrnieder has joined #gerrit[02:19:01] *** stixter has quit IRC[02:26:44] *** dougsland has quit IRC[02:40:10] *** wenlock has joined #gerrit[03:09:58] *** wenlock has quit IRC[03:41:33] *** Vampire0_ has joined #gerrit[03:44:23] *** Vampire0 has quit IRC[05:20:42] *** denizt has joined #gerrit[05:20:44] *** denizt_ has joined #gerrit[07:19:03] *** St_Marx has quit IRC[07:32:04] *** cmn has quit IRC[07:34:33] *** cmn has joined #gerrit[07:47:10] *** wenlock has joined #gerrit[07:51:43] *** wenlock has quit IRC[07:54:58] *** zxiiro is now known as zxiiro-away[08:09:51] *** wenlock has joined #gerrit[08:27:22] *** wenlock has quit IRC[08:32:51] *** srenatus has joined #gerrit[09:06:22] *** qchris has joined #gerrit[09:14:02] *** mountains__ has joined #gerrit[09:32:20] *** qchris is now known as qchris_away[09:49:46] <j416> if I push a branch to gerrit with commits A, B, C and reviewer says +2 on A and it gets merged by gerrit, then arrives at B and realises B should be squashed into A because it doesn't make sense for it to be in a separate commit -- how can that be handled? A is already merged so cannot easily be changed without manual fiddling, can it?[09:53:53] <tty1> j416: im not sure if there is a more elegant way but i can tell you how i handle it[09:54:01] <j416> please[09:54:30] <tty1> j416: I reject C and then comment on B telling the original author to squash C into it and then resubmit as a new version of B[09:54:50] <tty1> j416: in short, i handle it the same way i handle any commit which doesnt "pass" inspection[09:54:59] <j416> please read my question again :)[09:55:18] <tty1> j416: ohhh i misread[09:55:25] <tty1> you want to swuash it AFTER A is submited[09:55:36] <j416> but I agree with your answer, if it had been B and C[09:55:43] <j416> tty1: exactly[09:55:46] <tty1> right[09:55:52] <tty1> j416: ok in that case its a bit different[09:56:29] <j416> my spontaneous guess is that gerrit doesn't handle it and you'd have to work around it[09:56:42] <j416> but I'm not sure so that's why I'm asking[09:56:46] <tty1> j416: the way i setup things i dont allow rewriting of history, so in my case i just let A and B to remain seperate. Usually I wont let A pass unless it makes sense as a commit on its own anyway.. however i do allow the super user (me) to be the only one allowed to rewrite history[09:56:57] <j416> ok[09:57:05] <tty1> j416: so if it is important those commits get squashed then as the super user i would bypass gerrit and commit direct to master[09:57:14] <j416> but if you review the commits in order, there is a good chance that A makes sense on its own, but B doesn't[09:57:19] <j416> alright[09:57:21] <j416> thanks[09:57:28] <tty1> j416: generally i like to avoid rewriting of history because it is a HUGE pain for allt he contributors who have outstanding commits of their own[09:57:49] <j416> one thought I had was to perhaps make gerrit avoid merging until the entire patch set was +2:ed[09:57:57] <j416> that could solve it[09:58:00] <tty1> j416: well i agree soemtimes you might see A and B as being a single feature, and B just slightly expands on A[09:58:18] <tty1> but i just choose to accept A as an additional, albeit minor, feature enhancement[09:58:31] <j416> and it doesn't even have to be A and B, it is often A and .. H[09:58:35] <j416> alright.[09:58:39] <j416> ty[09:58:44] <tty1> j416: well B might not be pushed/commit until days later[09:58:52] <tty1> so im not sure how gerrit would know that B and C were coming[09:58:52] <j416> true, true.[09:59:06] <j416> commit-based, right[09:59:14] <j416> I'm still in branch-based world :)[09:59:15] <tty1> j416: even before gerrit i wouldnt allow history rewrites on the official server[09:59:38] *** qchris_away is now known as qchris[09:59:40] <j416> we don't do that either[09:59:42] <tty1> j416: i usually encourage my devs to do all their squashing and rebasing as best they can befor ethey commit because after a commit makes it into production i personally never allow it to be rewritten[09:59:48] <j416> but we wouldn't have gerrit merge directly to master anyway so[10:00:09] <tty1> j416: oh well in that case you can always just squash the commits yourself before manually merging into master[10:00:18] <tty1> j416: i do setup my gerrit so it merges right to master[10:05:22] <j416> tty1: we wanted to make it part of the review process, but I see your point[10:05:34] <j416> thanks, I got the answers I wanted.[10:06:28] <tty1> j416: it isnt a problem unique to gerrit. it is really a general git question.. namely.. when is it ok to rewrite history[10:06:47] <j416> it depends on workflow[10:07:10] <tty1> j416: yea its a question that is to some degree unique to every team and workflow[10:07:18] <j416> yes[10:07:30] <tty1> for me ive always just said, if its a shared GIT server, no history rewrites unless everyone who is using the server knows about it[10:07:48] <tty1> j416: i leave ssquashing and rebasing largely to the individual coder to do[10:08:54] <tty1> j416: even on a topic branch i dont usually like to squash commits unless its part of the final rebase into a development branch (in which case gerrit handles that well)[10:09:23] <tty1> j416: i also treat rebasing from topic branches into a development branch a sort os "minirelease". so care is taken to make sure it is "complete"[10:09:27] <tty1> and wont need rewriting[10:14:20] <j416> I'm talking about review, not git per-se[10:14:49] <j416> I got your point. Thanks.[10:19:57] <tty1> j416: np[10:23:19] <jkt> j416: there's AFAIK no submit type which would force-rewind a branch (i.e. losing commits)[10:24:34] <jkt> j416: so about the only way is force-pushing (asuming you have proper ACL in place for a force push)[10:30:50] *** timothy has joined #gerrit[10:34:52] *** St_Marx has joined #gerrit[10:35:25] *** St_Marx has quit IRC[10:36:21] *** srenatus has quit IRC[10:36:21] *** srenatus has joined #gerrit[10:36:47] *** timothy has quit IRC[10:38:07] *** [1]test has joined #gerrit[10:40:53] *** timothy has joined #gerrit[10:48:40] *** franred has joined #gerrit[11:36:25] *** teran has joined #gerrit[11:58:45] *** teran has quit IRC[12:05:51] *** teran has joined #gerrit[12:06:48] *** [1]test has quit IRC[12:07:45] *** teran has quit IRC[12:18:52] *** teran has joined #gerrit[12:41:51] *** plinio has joined #gerrit[12:54:57] *** dougsland has joined #gerrit[13:02:12] *** dougsland has quit IRC[13:02:36] *** dougsland has joined #gerrit[13:03:12] *** teran has quit IRC[13:15:57] *** dougsland has quit IRC[13:20:21] *** dougsland has joined #gerrit[14:12:51] *** teran has joined #gerrit[14:30:04] *** hugares has joined #gerrit[14:36:43] *** aswadr has joined #gerrit[14:50:43] *** qchris is now known as qchris_away[15:08:11] *** dougsland has quit IRC[15:08:36] *** dougsland has joined #gerrit[15:11:06] *** teran has quit IRC[15:18:54] *** zxiiro-away is now known as zxiiro[15:26:08] *** qchris_away is now known as qchris[15:38:50] *** aricg has left #gerrit[15:45:05] *** teran has joined #gerrit[15:50:15] *** teran_ has joined #gerrit[15:51:05] *** teran__ has joined #gerrit[15:52:23] *** teran has quit IRC[15:54:48] *** teran_ has quit IRC[15:59:30] *** qchris is now known as qchris_meeting[16:19:52] *** bluework has joined #gerrit[16:19:54] <bluework> Hi here.[16:21:49] <bluework> I'm having few troubles using the "verified" flag in gerrit. I've added it to all project, and added the permissions for about everyone to verify it. I still can't see it in my dashboard, and using gerrit cli doesn't help :/[16:22:39] <bluework> I mean, it does prevent me from submitting a change, and I have a "Verified" entry in label block, but I can't find a way to set it[16:27:12] *** teran__ has quit IRC[16:44:35] *** wenlock has joined #gerrit[16:46:35] <bluework> Ok, fixed.[16:46:44] <bluework> The permissions should be in refs/heads, not in refs/for :-[17:19:25] *** teran has joined #gerrit[17:31:57] *** teran_ has joined #gerrit[17:32:51] *** wenlock has quit IRC[17:34:52] *** teran has quit IRC[17:38:16] *** teran_ has quit IRC[17:44:08] <micah> i have a repository on gerrit that has a commit that needs to be dropped[17:44:35] <micah> how do I remove that commit? it doesn't need to be purged from git entirely, its just a stray commit that shouldn't have gone in[17:44:37] <micah> I can't force push[17:48:32] <micah> i dont have a revert change button[17:56:30] <jkt> micah: if you're OK with reverting, you can `git revert` and push the result for review[17:56:52] <jkt> micah: if you would like to completley undo it, i.e. rewrite the history, you got to have a force push permission, there's no way around it[17:57:16] <micah> jkt: i'd like to completely undo it - but I am not finding force push as a permission[17:57:57] <jkt> micah: are you using a GUI, or are you doing that directly with git on refs/meta/config?[17:58:07] <micah> jkt: i was using the web interface[17:58:14] <micah> and looking at the Add permission... drop down[17:58:31] <micah> but I'll do it no matter how will work :)[18:00:36] <jkt> micah: there's a checkbox which appears once you add "push"[18:00:45] <jkt> if you check it, you have a force push permission[18:00:46] <jkt> iirc[18:01:49] <micah> aha![18:03:15] <micah> i missed that checkbox, thanks![18:11:07] *** qchris_meeting is now known as qchris_away[18:11:35] *** docwhat is now known as zz_docwhat[18:18:42] *** mountains__ has quit IRC[18:19:27] *** zz_docwhat is now known as docwhat[18:26:00] *** teran has joined #gerrit[18:27:18] *** cmn has quit IRC[18:31:05] *** cmn has joined #gerrit[18:35:42] *** cmn has quit IRC[18:50:48] *** aswadr has quit IRC[18:55:42] *** teran has quit IRC[18:58:40] *** qchris_away is now known as qchris_meeting[19:01:42] *** hugares has quit IRC[19:03:47] *** hugares has joined #gerrit[19:11:14] *** timothy has quit IRC[19:41:51] *** qchris_meeting is now known as qchris_away[19:43:08] *** srenatus has quit IRC[19:46:00] *** kad has quit IRC[19:47:53] *** kad has joined #gerrit[19:50:54] *** marlinc has quit IRC[20:48:45] *** plinio has quit IRC[20:55:49] *** marlinc has joined #gerrit[20:57:03] *** qchris_away is now known as qchris[21:17:36] *** marlinc has quit IRC[21:52:53] *** marlinc has joined #gerrit[22:01:59] *** hugares has quit IRC[22:15:51] *** talios has joined #gerrit[22:46:00] *** marlinc has quit IRC[22:51:30] *** marlinc has joined #gerrit[23:13:57] *** russt has joined #gerrit[23:36:33] *** qchris is now known as qchris_away[23:41:43] *** qchris_away is now known as qchris