13:31:21 #startmeeting 13:31:21 Meeting started Wed Oct 8 13:31:21 2014 UTC. The chair is nickm. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:31:21 Useful Commands: #action #agreed #help #info #idea #link #topic. 13:31:26 don't have much to report though 13:31:34 i'm now squashing my guardfraction branch 13:31:36 hi meeting 13:31:37 so that it's reviewable 13:31:42 keen 13:31:50 it has 4 XXXs in it, that are questions to the reviewer. 13:31:53 I think we've been building up things that need reviewing 13:31:56 other than that, it should be fine I think. 13:31:58 yes. 13:32:00 there's a global scheduler branch with configurable thresholds now 13:32:08 cool. thanks, athena! 13:32:10 i tried to review some of your branches... but I couldn't get confidence... 13:32:11 and i've got a review of #11243 on the way 13:32:34 i reviewed the zlib one with the !finished. the idea looked sane, but I didn't look at the zlib API to understand what's really going on with avail_in... 13:32:36 nice. I did a huge pile of tests for make sure #11243 was right 13:33:08 also, i looked at #12862 and tentatively decided we should leave it as-is to preserve symmetry between the pre- and post-switch tests 13:33:19 ok 13:33:24 any other opinions before i close it? 13:33:48 seems ok to me 13:33:54 ok 13:36:20 athena: did you see the questions Yawning and I had about the branch when we started reviewing it? I was supposed to write those up after Yawning and I had some time to go through it more... but I think I should just dig up my notes and send them unedited since I apparently haven't been on the ball wrt that for the last couple of days :/ 13:36:34 s/the branch/the scheduler branch/ 13:37:05 it's fine - no, i didn't see, but you can send me some notes 13:37:39 ok 13:38:24 So as for me, I've been poking #11243, writing a huge pile of tests for that, thinking about how to get our router parsing code to be less scary, and implementing more prop#220 stuff 13:38:37 as of this morning, it works with chutney, and the authorities do the right thing with microdescs 13:40:27 I should take a break some time (even though I am really enjoying myself) to review more code and answer some emails. 13:40:35 anybody want to review code with me later today? 13:41:01 sure 13:41:11 but I can't do this too late. 13:41:17 i can do it in 1-2 hours, if you want. 13:41:45 by then maybe I will have my guardfraction branch ready, so we can trade reviews :) 13:42:58 okay. I should be around a lot today. 13:43:23 ok 13:43:42 We should also be working on chutney stuff to try to get coverage working with chutney, and more sponsorS prelims done 13:43:48 but that's an email from roger I need to re-read first 13:44:14 anything else we should talk about today? Anybody else need something to do? 13:44:18 what needs to be done with coverage and chutney? 13:44:25 (sorry to jump in) 13:45:41 So, when coverage is enabled, we generate a tor binary called "tor-cov" 13:45:50 that's supposed to be tor build with gcov support 13:45:52 *built 13:46:08 (no worries; you're welcome here) 13:46:10 yeah, I've had coverage working locally 13:46:31 I tried making it go with chutney a month or two ago, and it didn't work for me. I haven't had time to investigate why yet. 13:46:44 and chutney working locally 13:46:59 I always wondered how it would work together with multiple processes 13:50:01 I thought gcov knows how to do locking as appropriate? But tbh I don't actually know 13:52:41 anyway, it's something I should look at 13:52:52 along with all your chutney patches, and mvdan's, and the tor-related patches 13:52:56 err 13:53:00 the chutney-related tor patches 13:54:10 yes, and #13163, which is only related to chutney insofar as I discovered it using TestNetwork 14:00:51 ok. sounds like end-of-the-meeting to me. Thanks, everyone! 14:00:54 #endmeeting