| [06:19] <TheMue> goooooood morning | |
| [06:36] <davecheney> helllo | |
| === TheMue_ is now known as TheMue | |
| [07:58] <rogpeppe> TheMue, davecheney: mornin' | |
| [07:59] <TheMue> rogpeppe: heya | |
| [08:00] <fwereade> yay, I has internets again, I hope | |
| [08:00] <fwereade> mornings :) | |
| [08:09] <TheMue> fwereade: internet makes stupid (topic of a German TV show yesterday) | |
| [08:27] <rogpeppe> just rebooting | |
| [09:22] <Aram> yo. | |
| [09:34] <wrtp> Aram: hiya | |
| [09:41] <TheMue> Aram: hi | |
| [09:42] <TheMue> Aram: jfi, i'm working on mstate/unit | |
| [10:24] <TheMue> lunchtime | |
| [12:14] <wrtp> niemeyer: yo | |
| [12:14] <wrtp> ! | |
| [12:16] <niemeyer> wrtp: Buenas! | |
| [12:16] <wrtp> niemeyer: do you know of any way of doing an "unmerge"? | |
| [12:17] <wrtp> niemeyer: i.e. i've got a branch that merged another one some time ago, and i want to back out from all the changes that that merge introduced | |
| [12:18] <niemeyer> wrtp: The only reasonable path for that is to apply a reversed diff | |
| [12:18] <niemeyer> wrtp: It's not free of side effects, unfortunately | |
| [12:19] <wrtp> niemeyer: thanks. | |
| [12:19] <niemeyer> wrtp: The old branch won't merge properly anymore, so it'll need to be recreated by branching and applying the forward diff again | |
| [12:19] <wrtp> niemeyer: i'll probably just build the branch from scratch again. | |
| [12:32] <wrtp> lunch | |
| [12:37] <niemeyer> wrtp: Enjoy | |
| [12:45] <wrtp> pwd | |
| [12:57] <niemeyer> :) | |
| [13:18] <TheMue> Aram: after unit, as said, i'm now starting with relation units (first step w/o presence and watcher) | |
| [13:45] <niemeyer> I have a scheduled visit to the doctor.. back soon | |
| [14:20] <wrtp> TheMue, Aram: a fairly small cleanup of the juju.Conn interface: https://codereview.appspot.com/6488077 | |
| [14:20] <wrtp> niemeyer: ^ | |
| [14:28] <TheMue> wrtp: you've got a review | |
| [14:28] <wrtp> TheMue: thanks | |
| [14:28] <TheMue> wrtp: yw | |
| [14:29] <wrtp> TheMue: about that comment - you're probably right, but i think i'll leave it be for the time being, as it's an unrelated change. | |
| [14:30] <TheMue> ok | |
| [14:54] * niemeyer waves | |
| [15:16] <wrtp> niemeyer: hiya | |
| [16:07] <niemeyer> wrtp: Review sent | |
| [16:07] <wrtp> niemeyer: ooh you're a lovely man | |
| [16:07] <niemeyer> Lunch time, biab | |
| [16:07] <niemeyer> wrtp: :-) | |
| [16:36] <TheMue> so, sports is calling, cu tomorrow | |
| [16:57] * niemeyer is back | |
| [16:57] <niemeyer> and heading to a meeting | |
| [16:57] <niemeyer> I suppose | |
| [17:30] <wrtp> niemeyer: is that CL LGTY with the suggested change? | |
| [17:31] <niemeyer> wrtp: Yeah, it does look reasonable | |
| [17:31] <wrtp> niemeyer: thanks | |
| [17:32] <wrtp> niemeyer: done. will submit. | |
| [17:33] <niemeyer> wrtp: Thanks! | |
| [17:33] <wrtp> niemeyer: good to get one more dependency out of the way! | |
| [17:33] <niemeyer> wrtp: Yeah, feels good doesn't it :) | |
| [17:33] <niemeyer> wrtp: Progress! | |
| [17:33] <wrtp> niemeyer: yeah. | |
| [17:33] <wrtp> niemeyer: upgrade-juju is done; just writing the tests. | |
| [17:34] <wrtp> niemeyer: i have an idea for the provisioner agent stuff - i'll see if it works out. | |
| [17:34] <niemeyer> wrtp: Woohay | |
| [17:34] <wrtp> niemeyer: if i'm right (and you think it's ok) it could save lots of time and code. we'll see. | |
| [17:35] <niemeyer> wrtp: Sweet | |
| [17:35] <niemeyer> wrtp: I'm just finishing the presence stuff with docs and whatnot | |
| [17:37] <wrtp> niemeyer: right, that's me for the day. good to finish with a submit... | |
| [17:37] <niemeyer> wrtp: +1 | |
| [17:37] <wrtp> niemeyer: have a good rest-of-day! | |
| [17:37] <niemeyer> wrtp: Have a good evening | |
| [17:38] <niemeyer> wrtp: thanks! | |
| [17:38] <wrtp> niemeyer: will do | |
| [23:32] <davecheney> heads up, while i'm waiting on a few things stabalise I'm doing juju ssh | |
| [23:33] <davecheney> i'll do scp at the same time as it is very similar | |
| [23:34] <niemeyer> davecheney: Cheers | |
| [23:36] <davecheney> niemeyer: could you have a read of the email I sent to you and Aram yesterday about mstate.Open/Dial | |
| [23:37] <davecheney> when you have a chance | |
| [23:37] <niemeyer> davecheney: It does support TLS | |
| [23:38] <niemeyer> davecheney: But I think we can chew one thing at a time | |
| [23:38] <niemeyer> davecheney: We'll have to define a new protocol to establish the connection | |
| [23:45] <davecheney> niemeyer: this is mainly for juju.Conn, what will replace the current UseSSH flag on state.Info ? | |
| [23:45] <davecheney> I guess I can do UseTLS and the define mstate.Dial/DialTLS | |
| [23:45] <niemeyer> davecheney: I think we should make mstate work entirely before we shift that over | |
| [23:45] <niemeyer> davecheney: Because otherwise we'll have new issues | |
| [23:45] <davecheney> of course | |
| [23:46] <davecheney> i'm not trying to put the cart before the horse | |
| [23:46] <niemeyer> davecheney: How to validate the certificate, how to provide the initial password, etc | |
| [23:46] <davecheney> yeeesh | |
| [23:46] <davecheney> all that :( | |
| [23:46] <davecheney> so, do you want mongo tunneled over ssh ? | |
| [23:47] <niemeyer> davecheney: UseSSH should at least so we can take the blocker out of everybody's plate | |
| [23:47] <davecheney> understood | |
| [23:47] <niemeyer> davecheney: Yeah, same thing we're doing now with ZooKeeper | |
| [23:47] <niemeyer> davecheney: After we get everything else working, we can fix that | |
| [23:47] <davecheney> sounds like a good plan to me | |
| [23:48] <davecheney> i'm all booked for lisbo, gonna get there on time this time :) | |
| [23:48] <davecheney> lisbon | |
| [23:49] <niemeyer> davecheney: That's great to hear. We'll have a good time there | |
| [23:50] <niemeyer> I'll get some dinner and bbiab | |
| [23:59] <davecheney> niemeyer: we have unit.AssignedMachineId() | |
| [23:59] <davecheney> would I be able to add | |
| [23:59] <davecheney> (or replace it with) | |
| [23:59] <davecheney> unit.AssignedMachine | |
| [23:59] <davecheney> as most of the other machine related methods on unit take a *Machine, not an id | |