16:00:47 <asn> #startmeeting SponsorR
16:00:47 <MeetBot> Meeting started Tue Feb 10 16:00:47 2015 UTC.  The chair is asn. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:47 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:00:54 <asn> hello friends
16:01:08 <asn> let's get going with the usual drill
16:01:14 <asn> let me start with a status report, and you go on with yours
16:01:29 <asn> over the past week, I worked on the blog post about the HS statistics
16:01:42 <asn> karen is looking at it today, so it should be ready to post RSN
16:01:53 <asn> i also did various SponsorR bureaucracy
16:02:10 <asn> and I started a [tor-dev] thread about more statistics that we could collect
16:02:22 <asn> i skimmed over aaron's reply and it had some smarte comments.
16:02:25 <asn> *Smart
16:02:32 <asn> and that's that I think from the SponsorR world.
16:02:36 <asn> who next?
16:02:39 <ohmygodel> ok i can go
16:02:42 <asn> ohmygodel: please
16:02:51 <ohmygodel> i helped work out the number for the HS stast blog post
16:03:10 <ohmygodel> i made a wiki page summarizing the terminology discussion (linked to from the sponsor r wiki page)
16:03:15 <ohmygodel> thats it
16:03:31 <asn> thanks that's useful
16:03:34 * karsten can go next
16:03:35 <asn> karsten: please
16:03:50 <karsten> I looked at ohmygodel's math to calculate fraction of hidden-service traffic,
16:04:05 <karsten> and found a factor-two error introduced in the extrapolation report.
16:04:34 <karsten> I posted the calculation to tor-dev@. that's ohmygodel's, asn's, and my joint work.
16:04:55 <karsten> I also pushed sources for report graphs and report sources to git.
16:05:09 <karsten> and updated the published report to take out that factor-two bug.
16:05:12 <karsten> that's all.
16:05:20 <asn> that's good. thanks!
16:05:26 <asn> yeah good we found that factor-two bug.
16:05:31 <karsten> yes!
16:05:46 <ohmygodel> is the tech report updated ?
16:05:47 <asn> dgoulet: syverson ?
16:05:49 <karsten> it is.
16:05:50 <ohmygodel> https://research.torproject.org/techreports/extrapolating-hidserv-stats-2015-01-31.pdf
16:05:57 * dgoulet can go
16:06:10 <karsten> ohmygodel: see the footnote on page 3.
16:06:19 <ohmygodel> ok cool, the 1/31 date threw me off
16:06:50 <asn> dgoulet: go!
16:07:36 <dgoulet> I'm no more sick and 0.2.6 is almost out so now more time on sponsorR, mostly right now working on hs health measurer (#13209)
16:08:04 <dgoulet> last week was ETOOMANY little-t tor so this week I want to focus solely on SponsorR
16:08:05 * dgoulet done
16:08:14 * syverson next
16:08:20 <asn> syverson: yes
16:08:36 <syverson> I resolved terminology discussions with ohmygodel.
16:08:59 <syverson> Began diving into problems with Eigenspeed for BW stats.
16:09:04 * syverson done
16:09:11 <asn> thanks
16:09:16 <asn> OK, that's good.
16:09:20 <asn> Quite fast.
16:09:27 <asn> Should we move to discussion phase?
16:09:30 <asn> What do we want to discuss?
16:09:57 <ohmygodel> terminology thoughts ?
16:10:07 <asn> Sure.
16:10:17 <asn> Maybe we could also discuss the next stats?
16:10:22 <asn> and of course what we will bve doing next week...
16:10:23 <ohmygodel> ok
16:10:28 <asn> anything else?
16:10:59 <asn> sounds good to me.
16:11:00 <asn> ok
16:11:04 <asn> let's go with terminology thoughts.
16:11:10 <ohmygodel> so as i said in my email
16:11:11 <asn> unfortunately, i haven't had time to read the thread or wiki entry.
16:11:14 <asn> how can I help here?
16:11:20 <ohmygodel> im going to start preferring the term “onion service” to “hidden service”
16:11:42 <ohmygodel> you can help by considering if “onion service” is right for you :-)
16:11:50 <asn> i think onion service is right for me.
16:11:55 <ohmygodel> also i want some opinions about broadening the discussion
16:11:58 <asn> i'm not sure if we should start switching now.
16:12:10 <asn> like, i don't think we should say "onion service" on this blog post.
16:12:23 <syverson> I already have been using this terminology in papers and discussions, e.g., with DARPA and DoJ.
16:12:30 <asn> mainly because "hidden service" is super established brand name, and switching from it will require some marketing skills.
16:12:57 <syverson> It's branded with a coward's shame.
16:12:58 <ohmygodel> i think you could make the change with a footnote explaining that “onion service” = “hidden service” and why you prefer that term
16:13:01 <asn> i've been using "onion space" instead of "darknet", because darknet is terrible and I don't think it's official terminology (unliuke hidden service)
16:13:09 <syverson> Sorry, cultural reference that dates me.
16:13:24 <ohmygodel> yeah syverson *whoosh*
16:13:39 <ohmygodel> hm, asn, that is not how i had understood “onion space”
16:13:42 <ohmygodel> but actually i like it better
16:13:49 <ohmygodel> i had understood it to be the set of onion addresses
16:14:02 <ohmygodel> but maybe it it better used as the set of onion services
16:14:09 <asn> what is the set of onion services in your lingo?
16:14:10 <ohmygodel> i like it
16:14:15 <ohmygodel> nothing haha
16:14:19 <ohmygodel> im changing the wiki !
16:14:22 <asn> there is no replacement for darknet?
16:14:35 <ohmygodel> no, there really should be one
16:14:37 <asn> i think that's the term we really need to replace. it's so shit.
16:14:42 <ohmygodel> totally agree
16:14:51 <syverson> Not sure there should be one honestly.
16:15:11 <asn> many people seem to want a word for this cloud.
16:15:26 <asn> for this network of computers.
16:15:37 <ohmygodel> new def’n: '''onionspace''' should be used to refer to the set of onion services that have been available "recently" (context-dependent).
16:16:39 <asn> i don't think we should change the blog post fwiw.
16:16:40 <ohmygodel> fine if you dont want to change the blog post. it will be a gradual change in any case.
16:16:46 <asn> ok.
16:16:59 <syverson> Perhaps we can split hairs: onionspace= space of onion services.   .onion space= space of .onion names.
16:17:04 <syverson> Is that too sublte?
16:17:16 <syverson> subtle
16:17:17 <ohmygodel> i would prefer “onion address space” for the latter
16:17:48 <ohmygodel> obvious analog to, e.g., “IPv4 address space”
16:17:49 <syverson> or onion domain space?
16:18:10 <karsten> what about protocol details? hidden-service directory -> onion directory? hidden-service client -> onion client? hidden server -> onion server?
16:18:23 <asn> karsten: oh wow, you just opened the pandoras box.
16:18:29 <syverson> Maybe you and I aren't actually done, ohmygodel. ;>)
16:18:41 <asn> change the consensus flags to OnionDir.
16:18:48 <ohmygodel> karsten: we havent discussed new terms for those
16:18:49 <karsten> asn!
16:19:03 <syverson> Ermm, I thought we had a bit.
16:19:12 <ohmygodel> i think theyre less necessary because fewer people need to refer to those things
16:19:27 <ohmygodel> at least HSDir
16:19:30 <karsten> true. I'm just thinking how to write the next tech report.
16:19:31 <asn> ohmygodel: what's your replacement for "encrypted services", which is also terrible name but fortunately far from established right now.
16:19:40 <ohmygodel> hidden-service client -> onion service client seems fine
16:19:55 <ohmygodel> similarly, hidden server (new term to me actually) -> onion server seems to work
16:19:58 <syverson> We've been quibbling over "encrypted service"
16:20:01 <dgoulet> karsten: oh wow massive rebranding! :)
16:20:13 <asn> yeah this is serious rebranding stuff.
16:20:23 <syverson> it depends whether these are aonother kind of onion service or a different animai.
16:20:26 <dgoulet> sounds like 224 + rebranding would be perfect
16:20:31 <asn> I think Tor has a PR team these days. they should definitely be informed of this.
16:20:32 <ohmygodel> asn: given that the details of such services arent finished, the name doesn’t need to/can’t be made right now
16:20:38 <syverson> I was pushing for "Tor-required service"
16:20:47 <ohmygodel> best suggestions are: “direct onion service” and “Tor-require service"
16:20:49 <asn> dgoulet: yes exactly. that's also what I think. Do official rebranding when the onion address changes to 56 characters.
16:20:52 <syverson> What he said.
16:21:06 <asn> interesting names.
16:21:14 <asn> i don't realy *like* them though.
16:21:17 <asn> :P
16:21:30 <asn> more useful than "encrypted services" for sure though.
16:21:44 <ohmygodel> fair enough, im not in love either :-)
16:22:09 <syverson> I think we need to change how we refer to things amongs ourselves and with the public sooner rather than later, but not needed for tor-required services before they're designed.
16:22:49 <asn> right
16:22:57 <asn> OK. I think we can move on to next subject?
16:22:58 <syverson> Switching to "onion" for existing HS stuff should be doable and recommended now.
16:23:04 <syverson> Yes please.
16:23:05 <ohmygodel> asn: one question
16:23:16 <asn> I think a [tor-talk] mail with our intentions and ideas might be a good step at this time.
16:23:19 <ohmygodel> advice for asking for more opnions ?
16:23:45 <syverson> How about [tor-dev] rather than [tor-talk]?
16:23:48 <asn> sure
16:23:49 <ohmygodel> ok i dont subscribe to tor-talk. what is it used for? who listens to it?
16:23:54 <ohmygodel> ok tor-dev instead
16:24:04 <asn> more community (people who setup HSes), less developers.
16:24:08 <asn> but [tor-dev] is fine too.
16:24:21 <asn> can you send this mail ohmygodel ?
16:24:26 <ohmygodel> yes
16:24:29 <asn> ok that's great.
16:24:30 <asn> thanks!
16:24:51 <asn> next subject? or question left?
16:25:19 <ohmygodel> next subject
16:25:20 <asn> ok
16:25:27 <asn> Should we discuss more stats?
16:25:37 <ohmygodel> sure
16:25:46 <asn> not sure if that's the best use of our time though.
16:26:10 <syverson> asn, what do you suggest?
16:26:21 <asn> not sure how to make this a worthwhile discussion.
16:26:29 <asn> we could walk over the 4 proposed stats and discuss them.
16:26:36 <asn> but it's going to take a while, and maybe mailing list is better.
16:26:44 <asn> what do you think?
16:26:50 <asn> do we have more discussion topics?
16:26:52 <asn> if not, we can do it.
16:26:54 <ohmygodel> yeah i think thats better over the mailing list
16:27:03 <asn> you mean you prefer mailing list?
16:27:12 <ohmygodel> i prefer the mailing list
16:27:13 <ohmygodel> i do wonder if someone could remind me what our plans are for stats this quarter
16:27:26 <asn> yes that's a good topic.
16:27:41 <asn> we are suppose to advance both the "more stats" thing, and the "stats aggregation" thing9~.
16:27:52 <asn> at this point, I think advancing the stats aggregation thing might be more worthwhile.
16:27:52 <syverson> ohmygodel, asn, you mean the memex sri list or the hs-stats list?
16:28:20 <asn> because if we have the stats aggregation thing, we can do more stats, and we should also port the existing stats there too.
16:28:33 <ohmygodel> syveson: we have a thread on tor-dev
16:28:45 <asn> i meant the [tor-dev] thread that was tstarted yuesterday in this case.
16:29:03 <syverson> OK sounds good. (Too many lists.)
16:29:12 <ohmygodel> asn: i agree
16:29:27 <asn> ohmygodel: ok.
16:29:30 <ohmygodel> i had thought that we could knock out some easy additional stats for a quick win
16:29:53 <karsten> ohmygodel: agreed.
16:30:01 <karsten> we should think about what's realistic to present in april.
16:30:19 <ohmygodel> and maybe we still can (the # of RPs established seems useful and easy)
16:30:21 <karsten> if we only aim for better stats aggregation, maybe we'll have a design, but no results.
16:30:52 <asn> i can very
 easily see the "easy additional stats" draining all the remaining time though.
16:30:56 <karsten> I'm saying that without really following what stats we might want to add.
16:31:02 <asn> that is proposal, implementation, testing, deployment, analysis.
16:31:07 <karsten> asn: yeah..
16:31:11 <asn> this is easy 2 months.
16:31:28 <asn> allowing enough time on the meanwhile for public review etc.
16:31:42 <asn> i'm not sure. the tradeoffs are interesting here.
16:31:48 <karsten> we can focus on one thing this week and re-evaluate next week whether that'll work out until april.
16:32:14 <asn> is nickm interested in helping out with the stats aggregation thing?
16:32:29 <ohmygodel> asn: yes
16:32:35 <asn> ok.
16:32:50 <asn> i can read the AnonStats3 and other emails this week.
16:32:52 <ohmygodel> in particular, he was interested in helping implement the required crypto
16:33:03 <ohmygodel> he suggested that it might be a fun Rust project
16:33:13 <asn> hah oh wow
16:33:13 <asn> ok
16:33:37 <ohmygodel> (i sent him the abe and okamoto paper on partially-blind signatures)
16:33:49 <asn> i just remembered that AnonStats has crypto.
16:33:56 <asn> and a dedicated StatsAuth.
16:34:03 <ohmygodel> yup :-D
16:34:26 <asn> OK, I really don't think it can be finished by mid-April.
16:34:31 <asn> so we are looking at something like this:
16:34:49 <asn> - Either we get 1-2 extra stats for April but not much progress on the AnonStats front
16:35:04 <asn> - Or we do some progress (proposal, start of implementation) on the AnonStats front, but no extra stats
16:35:38 <asn> I think that the latter is more worthwhile.
16:35:41 <Yawning> ohmygodel: hmmm, japanese cryptographers, a bit of an oddity
16:35:42 <Yawning> :P
16:35:48 <asn> not at all!
16:35:58 <Yawning> (lemmie guess, they're with NTT or something)
16:36:21 <ohmygodel> there are some great japanese cryptographers
16:36:22 <asn> However, I don't want to start implementing AnonStats so soon.
16:36:27 <ohmygodel> haha yeah NTT
16:36:32 <asn> Which pushes me towards the former...
16:36:32 <Yawning> hah, knew it
16:36:41 <Yawning> the NTT folks did elligator^2 which was spiffy
16:36:48 <ohmygodel> their english-as-a-second-language makes their papers often a challenge to read, though…
16:36:58 <Yawning> not for me
16:37:09 <CGIIIIIIIIIIII1111> dictatorship!
16:37:14 <asn> Since both projecst are currently in the discussion stage though.
16:37:22 <asn> Maybe we can advance the discussion during the upcoming week?
16:37:26 <asn> And see where it takes us next week?
16:37:27 <Yawning> (I just immediatly think of stuff like MISTY/Camellia when it's japanese crypto)
16:37:32 <asn> I will focus on the ANonStats discussion personally.
16:37:43 <CGIIIIIIIIIIII1111> freemode for CGI!
16:37:46 <karsten> I should re-read that thread, too.
16:37:47 <dgoulet> well did secure bw measurement is needed for AnonStats anyway?
16:38:04 <asn> oh
16:38:11 <ohmygodel> dgoulet: no
16:38:26 <Sebastian> dgoulet: not a good time atm
16:38:27 <syverson> How about the people who invented re-encryption mixes, Yawning? There is lots of Japanese crypto.
16:38:41 <dgoulet> Sebastian: no worries
16:38:44 <Sebastian> dgoulet: what's up? I'll answer asap
16:38:57 <Yawning> syverson: yeah, guess I just need to read more papers
16:38:59 <ohmygodel> i mean, the bandwidth-weighted stats vote will be more secure if you can trust the bandwidth weights, but it will be no less secure than tor itself...
16:39:37 <CGIIIIIIIIIIII1111> why you need gun?
16:40:12 <asn> OK. As you can see, it's not clear to me what we should do here.
16:40:31 <ohmygodel> asn: i think exploring anon stats collection to see how that might work is a fine idea
16:40:32 <asn> Maybe I can read AnonStats3, and see if it's something that can be written into an official proposal.
16:40:35 <karsten> also remember that we want to do other stats-related things this quarter.
16:40:44 <karsten> finish that other tech report, implement stats on Metrics, etc.
16:40:54 <asn> karsten: can you handle the metrics side?
16:41:07 <asn> i'm not that concerned with the other tech report being finished.
16:41:07 <karsten> asn: yes. but then I cannot handle other things.
16:41:13 <asn> karsten: what other things would you handle this week?
16:41:32 <karsten> asn: nothing yet. whatever is most urgent/important.
16:41:46 <asn> OK. Maybe you can take the metrics incorporation.
16:41:57 <karsten> sure.
16:41:58 <asn> I can take the AnonStats thing, and maybe talk some more about extra stats.
16:42:25 <karsten> sounds good.
16:42:29 <asn> And I'll try to have some sort of plan wrt stats results presentable by April by next week.
16:43:00 <asn> sounds plausible?
16:43:05 <dgoulet> if we can decide on which stats to collect also, I can get started on the code/proposal if you need to be offloaded
16:43:33 <asn> well, if we decide that AnonStats is doable soon-ish, maybe we shouldn't implement the other stats just yet and just do them in AnonStats when it's ready.
16:43:56 <dgoulet> probably better
16:44:14 <CGIIIIIIIIIIII1111> any dictators online in here?
16:44:15 <karsten> will there be a proposal for anonstats?
16:44:17 <asn> like *maybe* we can have some AnonStats stats by the *next* quarterly meeting on July.
16:44:20 <asn> karsten: yes.
16:44:34 <karsten> asn: great.
16:44:35 <asn> it's something that relays will need to start doing, so it really needs proposal.
16:44:51 <asn> it's going to be quite a big project. and I don't really volunteer to implement it.
16:44:57 <karsten> heh
16:45:06 <karsten> and maintain, of course.
16:45:09 <asn> ...
16:45:15 <asn> but the ground work needs to happen anyway.
16:45:42 <ohmygodel> yeah it would be great to make it as modular as possible
16:45:42 <karsten> yes. just saying that a quick-and-dirty implementation is not what we want.
16:45:42 <asn> so OK...
16:45:56 <asn> karsten: yeah
16:46:07 <asn> karsten: some sort of prototype might be useful.
16:46:12 <karsten> oh sure.
16:46:19 <asn> but what the Tor relays actually use needs to be *proper*.
16:46:38 <asn> OK. Next topic?
16:46:52 <asn> I think we are done with topics.
16:47:09 <ohmygodel> great
16:47:14 <asn> Do we have things to do next week?
16:47:21 <asn> I think I just solved this for me.
16:47:27 <ohmygodel> i do
16:47:30 <asn> Karsten, you do work on metrics integration?
16:47:31 * karsten starts adding graphs to Metrics.
16:47:33 <asn> dgoulet, you do?
16:48:00 <dgoulet> HS health! + Testing
16:48:01 <asn> karsten: btw, I wanted to tell you, that even the smaller figures of the extrapolation tech report would be very interesteing in metrics
16:48:22 <asn> karsten: like, if metrics could display the probs of becoming an HSDir over time for all relays, like Figure 6
16:48:27 <asn> that would be very nice. since we could monitor it in real time
16:48:33 <asn> and spot any anomalies.
16:48:41 <asn> dgoulet: OK, I assume you know what you will be doing there.
16:48:42 <dgoulet> yes that falls into hs health also quite well ^ :)
16:48:43 <asn> dgoulet: do you need help?
16:48:53 <dgoulet> asn: yes I know, and yes I'll need help on some reviewing
16:49:08 <asn> ok
16:49:21 <ohmygodel> did we pick a time for next week? i should mention that the only sponsor r thing im planning to spend major time on for the next two weeks is peerflow, and so its not absolutely crucial that i am able to attend.
16:49:32 <dgoulet> (and help on not falling too much into little-t tor fixes :P)
16:49:33 <asn> ohmygodel: ah good!
16:49:48 <asn> so we are talking about next week on wednesday?
16:49:55 <asn> but karsten wont attend?
16:50:06 <dgoulet> ohmygodel: oh btw, I might have missed it but did you manage to convince chris that peerflow is the new black? :)
16:50:06 <karsten> asn: will think about what graphs to add.
16:50:15 <karsten> asn: I'll try to attent.
16:50:16 <asn> if ohmygodel is spending time on peerflow, maybe we can restore the meeting to the normal time, and have karsten but not ohmygodel?
16:50:19 * karsten gotta run; phone
16:50:33 <asn> ah karsten cannot do tuesdays either...
16:50:33 <sjmurdoch> Have folks seen http://www.wired.com/2015/02/darpa-memex-dark-web/
16:50:41 <asn> sjmurdoch: not yet.
16:51:06 <asn> sjmurdoch: oh wow just read the title..
16:51:10 <ohmygodel> ok wed. 2/18 at 1600UTC
16:51:21 <asn> ohmygodel: seems so.
16:51:45 <asn> OK, so we are done here guys?
16:51:52 <ohmygodel> oh god. onionspace!
16:51:57 <ohmygodel> done asn
16:52:12 <ohmygodel> dgoulet: im interested in your health service
16:52:20 <dgoulet> ohmygodel: sure
16:52:21 <ohmygodel> could you summarize it ?
16:52:39 <sjmurdoch> ohmygodel: If you haven't seen, I've used some comments from you to inspire this: https://theconversation.com/tor-the-last-bastion-of-online-anonymity-but-is-it-still-secure-after-silk-road-35395?utm_medium=email&utm_campaign=The+Weekend+Conversation+-+2419&utm_content=The+Weekend+Conversation+-+2419+CID_f257b679c0ae38a0bf3823cee90d82f4&utm_source=campaign_monitor_uk&utm_term=Tor%20the%20last%20bastion%20of%20online%20anonymity%20b
16:52:40 <sjmurdoch> A nicer URL: https://theconversation.com/tor-the-last-bastion-of-online-anonymity-but-is-it-still-secure-after-silk-road-35395
16:53:46 <dgoulet> ohmygodel: last point at the bottom: https://lists.torproject.org/pipermail/tor-dev/2015-February/008230.html
16:54:17 <dgoulet> ohmygodel: there are multiple things we can do with it, mostly right now i'm interested in knowing the impact of relay churn and reachability
16:54:24 <asn> #endmeeting