13:38:05 <nickm> #startmeeting
13:38:05 <MeetBot> Meeting started Wed Dec  3 13:38:05 2014 UTC.  The chair is nickm. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:38:05 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
13:38:11 <nickm> hi, dgoulet; hi teor!
13:38:14 <dgoulet> hello all!
13:38:16 <nickm> who else is here?
13:38:23 <teor> Hi nickm, dgoulet, all
13:40:16 <nickm> So, this past week has not been a super-productive Tor time for me.  I've been doing managementy things with the board's executive committee, trying to find time to look at patches, and distracting myself with a fun little PRNG that's mostly done now, I think. :)
13:40:33 <nickm> Or rather, it's been a productive Tor time, but not a productive tor-the-program time.
13:40:52 <dgoulet> :)
13:41:04 <nickm> But we've got a bunch of stuff to deal with structurally to keep the company healthy and make it work better, so I'm probably not getting as much hacking done as I should.
13:41:36 <teor> There is only one of you
13:42:09 <teor> :-)
13:42:22 <nickm> we need more mes.
13:42:45 <nickm> dgoulet: did you have an opportunity to look at my writeup of the testing stuff, by chance?
13:42:46 <teor> nickm.clone() ?
13:43:24 <dgoulet> nickm: yeah I did read it, it was mostly fine be my tbh, do you need help on something?
13:43:28 <nickm> teor: if people could do that, the world would be very different.
13:43:41 <nickm> dgoulet: just a sanity check fwict.
13:44:10 <nickm> teor: Did you have a chance to look at the thing?  We've been trying to sketch out what the ecosystem around a broad set of integration tests would look like, and you've been paying a lot of attention to chutney recently
13:45:02 <teor> The post on tor-dev@lists ? Yes, it seemed sensible.
13:45:08 <nickm> keen
13:45:19 <teor> And I wanted it without all the effort getting there.
13:45:28 <nickm> hm?
13:45:41 <teor> It seemed so good, I wanted it now.
13:46:02 <nickm> well, I can always use more hands. :)
13:46:09 <dgoulet> +1 :)
13:46:42 <nickm> Step zero is probably sketching out v1 of the APIs, and getting chutney happier in general.
13:46:48 <karsten> nickm.attach(teor.removeHand());
13:46:50 <nickm> we shouldn't spend too much time on the APIs though...
13:47:15 <nickm> Designing APIs for a tool you are currently building is trying to generalize from a sample size of zero.
13:47:20 <teor> Happy to pitch in. And yes, chutney does seem to need some wrangling at times
13:47:35 <nickm> anything waiting for me to review and merge on that front?
13:47:49 <teor> nickm: ... while aiming at a moving target.
13:47:58 <dgoulet> is it #13718 the one about chutney bootstrap?
13:48:27 <nickm> also see #13823
13:48:29 <dgoulet> anything I can help because this one is powerfully annoying for me, takes ~45-60sec everytime I start chutney to reach a quiescent state of "usability"
13:48:39 <teor> Oh yeah. I'm working my way through that one. It has spawned a multitide of children.
13:49:09 <dgoulet> teor: yeah saw that, don't hesitate to ping me for test/review/code/coffee :), I would really like to see this one fix
13:49:23 <nickm> So, I've been assuming that your comments on those are mostly thinking aloud, plus hoping that somebody will tell you if you've made any obvious mistakes in your thinking. But sometimes I decide that about comments when people are actually hoping for advice and answers and suggestions.
13:49:46 <nickm> So pleae don't hesitate to ask explicitly
13:50:29 <teor> Oh, no, just me trying to keep track of the issues that have arisen
13:50:44 <nickm> ok, thought so
13:51:19 <teor> I tend to name names or use question marks when I want advice.
13:52:00 <dgoulet> so maybe I can just give you a quick update on my side
13:52:25 <nickm> teor: great protocol.  I wish everybody followed it. :)
13:52:25 <teor> Sounds good
13:52:32 <nickm> go ahead
13:52:49 <dgoulet> very simple: I've been doing some performance measurements for SponsorR non stop since last week, but I want to use today for little-t tor stuff (code/review)
13:52:56 <dgoulet> I have a couple of ticket in mind
13:53:05 <dgoulet> but please, you guys need any help, ping me ;)
13:53:43 <dgoulet> also maybe after this meeting, I would like your input on some "bizarre" behavior for introduction point I've observed with my measurements
13:54:03 <nickm> interesting, sure
13:54:18 <nickm> btw, I still think your lttng stuff is cool and useful.
13:54:33 <dgoulet> and could be good to start thinking about a plan to integrate that stuff ^ upstream
13:54:42 <nickm> Does anything else use the same message format?  Is it well supported across languages?
13:54:56 <nickm> (Looking for a reporting format for chutney, but can't have chutne be linux-only)
13:55:51 <dgoulet> nickm: so lttng creates data in the CTF format (Common Trace Format, http://www.efficios.com/ctf), it's a format that started to be used more and more
13:56:28 <nickm> What's the biggest/most interesting other project using it?  Can you generate it and parse it from most interesting languages?  What's it based on?  And would you recommend it for chutney reporting to test scripts?
13:56:44 <asn> hello friends
13:56:49 <dgoulet> nickm: should I maybe answer those on a ticket? :)
13:57:00 <nickm> the alternative is probably cobbling something together out of json, line-based text, or protocolbufs
13:57:07 <nickm> dgoulet: I guess I should open a ticket then? :)
13:57:08 <nickm> hi asn !
13:57:16 <asn> im about to go to lecture so I'll be very brief
13:57:25 <asn> during the past week, I've been hacking on the SponsorR statistics code.
13:57:29 <dgoulet> nickm: yeah and also know that there are some json2ctf parsers or pcap2ctf :)
13:57:40 <asn> (Based on the proposal 238 draft we sent to the mailing list last week)
13:57:48 <asn> (Since the deadline is in a month from now, we had to start implementation)
13:58:10 <asn> currently I'm running the stats on a relay. and I passed the code to Karsten. Now karsten is going to obfuscate the  statistics
13:58:22 <karsten> yep.
13:58:24 <asn> this week, I'll be fixing the guardfraction stuff.
13:58:32 <asn> and I'll be helping karsten with any code help he needs.
13:58:47 <asn> as far as little-t-tor is concerned.
13:58:52 <asn> and that's that from me :)
14:00:34 <asn> karsten: on this relay: https://globe.torproject.org/#/relay/B120B6310D626FA54B6667DCAFEBD17A1EF47890
14:00:37 <dgoulet> asn: cool thanks!
14:00:51 <asn> karsten: for the past 20 hours, it has received 3512 RP cells.
14:00:51 <nickm> dgoulet: #13887
14:01:07 <dgoulet> nickm: great, I'll add the info there
14:01:17 <asn> karsten: nowhere near 20k we were speculating for a signle connection. but it's a pretty weak relay so it doesn't get many RP requests.
14:01:52 <karsten> asn: ok. I'll see if I can test on a slightly faster relay.
14:02:05 <asn> i'll let you know when the 24 hours end.
14:02:08 <karsten> asn: but so far, I'm still writing code. fighting with laplace, but making progress.
14:02:15 <asn> also, in about 10 hours I should regain the HSDir flag.
14:02:33 <asn> and now I'm off to lecture. have a good day guys!
14:02:40 <karsten> enjoy!
14:02:50 <dgoulet> asn, karsten: that's mine, please let me know if you need a fast one :) https://globe.torproject.org/#/relay/A69221A7EC7498D2F88A0FB795261013FA36CAAE
14:03:40 <karsten> dgoulet: good to know, thanks.
14:07:04 <nickm> anything else to review at the meeting?
14:07:16 <nickm> I think Yawning's got the flu, and athena is decompressing, so it may just be us.
14:07:51 * dgoulet report done
14:07:57 <nickm> #endmeeting