13:30:34 <nickm> #startmeeting 13:30:34 <MeetBot> Meeting started Wed May 13 13:30:34 2015 UTC. The chair is nickm. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:30:34 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic. 13:30:37 <nickm> hi everybody! 13:30:42 <nickm> It's another beautiful wednesday. 13:30:53 <dgoulet> hi meeting! 13:30:58 <nickm> we've got at least two discussion topics pending, but let's start with quick status reports. 13:31:08 <nickm> Please start chatting while I pour myself a coffee? 13:31:24 <Yawning> uhm, did a bunch of stuff that's in 0.2.7.1 13:31:43 <Yawning> plans are, "do more stuff, some of it pt, some of it tech writing, and look at prop 188" 13:31:51 <Yawning> (yay, new release) 13:32:16 <Yawning> also probably look into some of the low hanging fruit stuff on the 0.2.7.x list 13:33:48 <dgoulet> ok I can go maybe? 13:34:05 <nickm> go for it 13:34:34 <Yawning> sure 13:34:48 <Yawning> (I also distracted dgoulet with torsocks stuff muhahahahaha) 13:35:00 <dgoulet> here are the tickets I did patch for: #4862, #8243, #15881 13:35:45 <Yawning> *looks at the missing desc thing* 13:35:58 <dgoulet> I'm waiting for a reply on #14917, this has been mostly my little-t tor work apart from torsocks and hs health 13:36:00 <nickm> Yawning: it's a bit tricksy 13:36:05 <dgoulet> it is... 13:36:34 <dgoulet> latest patch fixes the HSDir issue but I'm open to other ways (if any) to do it 13:37:01 <dgoulet> but the thing is without that patch, hs health tool is unable to work properly :S 13:37:05 * dgoulet done 13:37:20 <nickm> and for 14917, I don't see the question. I think we could do "warn loudly at startup for a while; later, don't start." 13:37:25 <nickm> does that seem plausible? 13:38:10 <nickm> I'll go now I guess 13:38:11 <Yawning> dgoulet: that doesn't add it to uploaded event right? 13:38:18 <dgoulet> Yawning: not yet 13:38:18 <Yawning> (oh, ok, questions can wait) 13:38:28 <dgoulet> (yeah discussion phase soon :) 13:38:57 <nickm> I got 0.2.7.1-alpha released, and added another libevent maintainer since libevent has been stalling hard. I also spent a lot of time in the weeds looking at Tor financial stuff and being in meetings. 13:39:30 <nickm> I merged the outline for the "how to write tests" doc I wanted to do for SponsorU, but it needs more work 13:39:43 <nickm> I started drafting ideas for how to kick old releases off the net 13:40:16 <nickm> at https://pad.riseup.net/p/deprecating-old-tors 13:40:33 <isabela_> nice 13:40:42 <nickm> coming up I am going to try to spend the next day or two hacking and writing as much as I can. 13:40:49 <nickm> though I should review and merge patches too I guess :) 13:40:58 <nickm> I hope I can churn out a few more proposals 13:41:01 <dgoulet> nickm: keep in mind that this pad will be deleted after 30 days of inactivity :) 13:41:10 * nickm saves 13:41:23 <isabela_> :) 13:41:35 <Yawning> riesup's pad takes a while to load for me 13:42:36 <nickm> who's next? 13:43:58 <nickm> ok. if anyone else steps in we can chat more then, but let's move on to discussion. 13:44:11 <nickm> isabela_ has a cool thing for us to look at. isabela_ , do you want to present that or should I? 13:44:30 <isabela_> sure 13:44:37 * dgoulet smells trac overhaul! 13:44:40 <isabela_> https://docs.google.com/document/d/1luDGhBRw2eJxJ4JKlPlF2CCIqMRCFP35NRuM2scBSh8/edit# 13:44:43 <isabela_> lol 13:45:06 <isabela_> so this is where I am documenting what I want to change 13:45:13 <nickm> dgoulet: smells like ... victory. 13:45:19 <dgoulet> WOOT! 13:45:25 <isabela_> hehehe 13:45:30 <Yawning> hm earthquake 13:45:32 <Yawning> sec 13:45:37 <isabela_> oh man :( 13:45:42 <Yawning> ok 13:45:43 <Yawning> just minor 13:45:48 <nickm> ! 13:45:51 <isabela_> good 13:45:52 <Yawning> hey 2nd one today 13:46:03 <Yawning> the one in the morning was 6.something 13:46:09 <Yawning> (at source) 13:46:22 <nickm> so, readikng through it, I like the basic idea and would like to suggest some refinements 13:46:30 <nickm> I think the TBB team is currently using points, for instance 13:46:55 <dgoulet> also we can actually add fields if we need to 13:47:02 <isabela_> yes 13:47:23 <nickm> I think we can do lots of discussion and comment on the document, and let isabela_ figure out what to fold in? 13:47:26 <isabela_> I wonder how is the best way to circulate this around before is implemented to trac, maybe send to tor-dev? 13:47:44 <nickm> let's do one round of in-team discussion, then send to tor-dev 13:47:46 <isabela_> btw the workflow is not something that needs to be part of the first implementation 13:47:51 <nickm> agreed 13:48:04 <isabela_> nickm: sounds good 13:48:05 <nickm> our current workflow is not the trac default workflow, btw 13:48:09 <nickm> we have needs_review and needs_revision 13:48:25 <isabela_> as keywords? 13:48:30 <nickm> no, as states 13:48:30 <isabela_> or states 13:48:35 <isabela_> oh nice 13:48:51 <Yawning> (is this where I say wtb gerrit) 13:48:55 <nickm> the idea is that needs_review can been reviewed and merged. if the review says that revisions are needed, it goes into needs_revision until it's ready. 13:48:57 <Yawning> (or does that come later?) 13:49:35 <nickm> Yawning: I am in favor of gerrit or any such tool so long as somebody admins it. The sysadmins will probably not mind providing the OS-level hosting so long as someone else is responsible for the app. 13:49:56 <Yawning> yeah just not sure we have the mystical someone 13:50:05 <Yawning> I'd offer to do it but I'm a horribad sysadmin 13:50:10 * nickm too 13:50:27 <isabela_> would more teams like gerrit? 13:50:32 <isabela_> twitter uses it 13:50:44 <dgoulet> I was a sysadmin in an other life some years ago :P not sure I want to go back though eheh 13:50:53 <isabela_> hehehe 13:51:16 <isabela_> gerrit would be cool, never installed/admin it just used it 13:51:31 <Yawning> (if this was an offtopic question, sorry) 13:51:42 <Yawning> (not trying to distract, but since we were talking about workflow stuff....) 13:51:56 <isabela_> Yawning: no is good 13:52:21 <isabela_> nickm: milestone will be 'january' instead of a project or a release 13:52:42 <Yawning> should january be qualified with a year? 13:52:53 <Yawning> like Jan2016 or w/e 13:52:58 <nickm> I'm adding lots of comments on the document. Is that a good way to do discussion? 13:53:36 <isabela_> nickm: yes it is 13:53:41 <isabela_> Yawning: agree 13:53:53 <isabela_> please add comments and I will improve it base on them 13:54:18 <nickm> also btw: when I say something like "We do it this way because..." 13:54:25 <nickm> or "we used to do it like that but it failed because..." 13:54:34 <nickm> I'm not saying that we can't change; only explaining the current rationale. 13:54:45 <isabela_> yes, history is good 13:57:49 <isabela_> i guess i missed on arrow on my workflow 13:58:15 <isabela_> from reviewable back to open.. witch is the revision state you have 13:59:01 <isabela_> I think the currect flow you have should be ok to keep / i will look into it more (how the current one is set up) 13:59:56 <isabela_> I am excited about this plugin because we can track dependencies 13:59:58 <isabela_> http://trac-hacks.org/wiki/ProjectPlanPlugin 14:00:28 <nickm> ok, I think I commmented on everything I saw. groovy 14:01:47 <dgoulet> hrm something I have in mind with the current situation 14:02:05 <dgoulet> when we backport a fix nickm flags it but doesn't close it yet even though it's merged upstream 14:02:10 <dgoulet> I wonder how that would work in this new workflow 14:02:25 <dgoulet> "merged" state ? but not "closed"? 14:02:56 <isabela_> oh sorry, this is bad habit 14:03:19 <isabela_> i would close after demos 14:03:28 <isabela_> but we can change tht 14:03:30 <isabela_> that 14:03:37 <dgoulet> " i would close after demos" ? 14:04:08 <nickm> I am fine with a new workflow for backports, but we _do_ need a workflow for backports. :) 14:04:18 <isabela_> at twitter the team would merge and at the end of sprint meeting we will have demos for me to see if all good then the tickt would be closed 14:04:18 <dgoulet> ^ 14:04:42 <isabela_> no need to do the same 14:04:51 <nickm> maybe merged == "somebody else confirms that it now works in master" ? 14:05:04 <dgoulet> I feel that this ^ will be chaotic :S 14:05:18 <dgoulet> means we have to ack twice the ticket? :) 14:05:21 <isabela_> nickm: yeah, is a small room for qa before going to the wild 14:05:46 <nickm> I'm not sure we have a "merged" / "in-the-wild" distinction. :) 14:05:55 <isabela_> cool ;) 14:06:01 <nickm> we have "merged" / "released", but people start testing our code as soon as it's merged. 14:07:24 <isabela_> yeah, i need to get the current flow correct before anything else 14:08:13 <nickm> I think the basic idea looks good to me though 14:08:14 <isabela_> so, I will work on your feedback and send it back to you before tor-dev 14:08:25 <nickm> do we have a sense of who can make these changes? 14:09:06 <isabela_> I think qbi might be able to help. I am applying then localy on a server I have here at home to test if it will work how I want it to 14:09:25 <Yawning> o.O 14:09:34 <Yawning> having a home trac server is p hardcore 14:09:48 <isabela_> heheh i dont use it for real 14:09:53 <nickm> always mount a scratch monkey :) 14:09:54 <isabela_> is just to play and test set ups 14:10:06 <dgoulet> "Ticket 7161: Buy coffee" 14:10:09 <dgoulet> epic 14:10:10 <isabela_> lol 14:11:11 <dgoulet> right in the workflow, I would propose a way to have a "revision" state or something that can go back and forth with reviewable 14:11:38 <dgoulet> we often have lots of round trip between an initial patch and mergeable one 14:11:50 <dgoulet> (Yea code review!) 14:11:54 <isabela_> hehehe 14:12:07 <isabela_> yes, i missed an arrow there coming back 14:12:19 <nickm> we use needs_revision for that 14:12:37 <dgoulet> yeah basically needs_review/needs_revision indeed 14:12:46 <isabela_> cool 14:12:57 <dgoulet> pretty neat stuff 14:13:32 <isabela_> cool - will update with feedback.. probably ask TBB folks to look at it before sending to dev 14:13:46 <isabela_> might start applying this with you first then move to other teams 14:13:53 <isabela_> sorry you my little rat lab :) 14:14:01 <nickm> no worries 14:14:05 <dgoulet> isabela_: hrm quick question 14:14:09 <nickm> just make sure the cheese is the expensive kind :) 14:14:28 <isabela_> dgoulet: ssup 14:14:38 <Yawning> is this oneof those experiments where I get cocaine if I press a button? 14:14:39 <dgoulet> isabela_: tor-hs, -relay, -client will be removed, is there a way to categorize ticket then? 14:14:41 <Yawning> :P 14:14:54 <Yawning> yeah I thing the varions tor-<subsystem> tags useful 14:15:03 <dgoulet> I use tor-hs 14:15:20 <isabela_> so I am adding projects 14:15:30 <dgoulet> ah that would be "project" field? 14:15:35 <isabela_> project core tor - components of this project can be 'relay' 14:15:54 <Yawning> ah ok 14:15:57 <dgoulet> a1 14:16:24 <isabela_> project hs can have a dependency on project core tor ticket (can be a way to do tor-hs things but it can also be a component) 14:16:53 <dgoulet> and what happened when the HS goes beyond core tor ? 14:17:00 <dgoulet> the hs ticket* 14:17:11 <Yawning> a bunch of stuff we do touches lots of subsystems 14:17:13 <isabela_> hs will be a project 14:17:29 <isabela_> tickets will have dependencies 14:17:38 <dgoulet> ok I see 14:17:41 <Yawning> ah 14:17:41 <isabela_> that is the plugin i was talking about and that is how projects work are connected 14:18:15 <isabela_> so you can open a dependeny against core tor or against bridges 14:18:33 <dgoulet> sounds good 14:20:48 <dgoulet> other topics? :D 14:21:09 <Yawning> wtb descriptor ID in UPLOADED HS_DESC events 14:21:12 <Yawning> ;_; 14:21:27 <nickm> dgoulet: maybe let's do a quick round of review-and-merge after the meeting? 14:21:38 <nickm> I was pretty distracted yesterday 14:21:46 <dgoulet> nickm: sure 14:21:49 <Yawning> I can review stuff, but I need to dinner 14:21:56 <Yawning> at some point 14:21:59 <nickm> ok 14:22:08 <Yawning> and kitty is giving me death glares 14:22:12 <nickm> i'll be around for hours and hours 14:22:17 <nickm> shall I endmeeting? 14:22:32 <isabela_> cool 14:22:33 <dgoulet> hmm 14:22:36 <dgoulet> nickm: wait 14:22:47 <nickm> ok 14:22:47 <dgoulet> ah well yeah sure 14:22:50 <dgoulet> sorry 14:22:51 <nickm> #endmeeting