17:01:08 <nickm> #startmeeting weekly network team meeting
17:01:08 <MeetBot> Meeting started Mon Oct 24 17:01:08 2016 UTC.  The chair is nickm. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:01:08 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
17:01:10 <nickm> hello all!
17:01:14 <asn> hello meeting
17:01:17 <dgoulet> o/
17:01:22 <nickm> greetings to asn and dgoulet and isabela !
17:01:28 <nickm> good to see everybody
17:01:51 <nickm> no yawning right now
17:01:59 <isabela> he emailed the list his update
17:02:03 <nickm> ah, cool
17:02:07 <nickm> still behind in today's email.
17:02:13 <nickm> isis, athena: ping?
17:02:36 <nickm> Would anybody like to do the 1st status update?
17:03:02 <isabela> i could go
17:03:07 <nickm> great!
17:03:24 <isabela> last week i was offline must of the week because of otf summit - which was pretty good btw
17:03:44 <isabela> this week i aim to pick up the testing effort again and follow up with people on that
17:03:54 <isabela> some people were even reporting bug on twitter :) !
17:04:19 <isabela> i will also be working on drl q3 report and on 3 different presentations i have to do next week in chile
17:04:28 <isabela> so that also means next week i will be mosstly offline again
17:04:30 <isabela> done!
17:04:36 <isabela> ah
17:05:11 <asn> i can go next
17:05:11 <isabela> btw had a call about the modularization proposal that got rejected, received feedback on that and i think we should try again in feb (next deadline for proposals)
17:05:17 <isabela> done for realz
17:05:38 <asn> i go next
17:05:39 <asn> ====
17:05:42 <asn> Hello. During past week I did some reviews and worked further on revising the
17:05:42 <asn> prop224 client auth logic. I also attended the OTF summit in Baltimore; I had
17:05:42 <asn> interesting discussions there about network team hiring. I also talked with
17:05:42 <asn> some users of client auth about HS client auth (result: people would like
17:05:42 <asn> stealth auth in prop224), and I also talked about the Tails server project with
17:05:44 <asn> the Tails people.
17:05:47 <asn> This week, I have two main tasks, and infinite smaller ones. One big one is to
17:05:49 <asn> finalize the client auth torspec patch, and the other is to revise #19043 based
17:05:52 <asn> on dgoulet's review. Unclear if I will manage to do both of these in a week. I
17:05:54 <asn> should also revise the name system API proposal, but I doubt I'll have time for
17:05:57 <asn> this anytime soon.
17:05:59 <asn> ====
17:06:12 <asn> EOF
17:06:26 <nickm> whoa. Thanks, asn!
17:07:49 <nickm> asn: i know we talked about whether I should look at the auth proposals. Let me know when you're ready for me to do that.  I'm excited to look, but I have so many other things that I can probably chill till you're ready :)
17:07:58 <nickm> So, here's my last-week:
17:08:05 <asn> nickm: yep. still not ready for your eyes!
17:08:07 <nickm> I did releases for the security issues TROVE-2016-10-001.
17:08:31 <nickm> Relatedly, I decided that it was non-crazy to have a Tor CVE thing, and I grabbed TROVE as the acronym
17:08:37 <nickm> since why not.
17:08:58 <nickm> Also, I spent a while chasing bugs, reviewing code.
17:09:10 <nickm> Also, I banged on #15056, frustrating myself with how many tiny little details there are
17:09:25 <nickm> Also, I made master into 0.3.0.x and merged some pending patches
17:09:58 <nickm> And I kept writing module-level doxygen.  I have about a dozen modules in src/or/*.c left to write summaries for
17:10:12 <nickm> Pffew. Kind of overwhelming. :/
17:10:54 <nickm> this week, huh.  I want to get #15056 and the doc stuff done by end-of-month.  But we'll also be accumulating bugs in 0.2.9.4-alpha, which should get fixed.
17:11:11 <nickm> if we release another 029 before the next TBB alpha comes out, we can get better feedback there.
17:11:32 <nickm> I also want to think about end-of-life timing for 024, 025, 026, and 027, and what our longterm support plan is.
17:11:55 <nickm> maybe let's talk about how to figure that out later.  Right now we're in a hard situation.
17:12:10 <nickm> that's it for me
17:12:32 * dgoulet can go
17:12:49 <dgoulet> My last week had some fighting with Schleuder (fail) and doing some infrastructure work on the test network services which are now on TPO's infra :).
17:13:01 <dgoulet> I've reviewed some tickets and work on minor ones. Some tor-dev email on the thread about prop224 client authorization. I did a torsocks 2.2.0 stable release as well.
17:13:09 <dgoulet> I've emptied the email queue last week also so now in terms of code, I plan to start #20029.
17:13:17 <isabela> nickm: are we using a keyword or anything special to track 0.2.9.4-alpha bugs on trac?
17:13:31 <nickm> isabela: let's talk about that at discussion-time?
17:13:36 <dgoulet> --
17:13:44 <dgoulet> (I think that wraps it up)
17:13:47 <isabela> nickm: ok
17:14:30 <nickm> any more folks with status updates?
17:14:33 <nickm> For discussion topics I have seen "How do we handle bug-tracking for 0294alpha" and "Let's end-of-life old tors" and... anything else?
17:14:42 <nickm> let's add topics while we wait a minute to see if there's another update
17:15:27 <nickm> (any other topics?)
17:15:38 <isabela> (not from me)
17:15:39 <tjr> I have a question
17:15:41 <nickm> hi tjr !
17:16:05 <tjr> Not asking about timeline or fitting it into schedule, but I'm wondering what the current opinion/appetite for Consensus Diffs is. :)
17:16:17 <nickm> yawning was interested in finishing it up.
17:16:47 <nickm> we've been seeking funding for making the directory protocol generally have lower overhead, so I hope that's on track that way too.
17:17:01 <tjr> okay, great - thanks!
17:17:17 <nickm> Personally I'm wondering if the existing code is clean enough we can resolve it, or whether the protocol as specified is just too hard for mortals :(
17:17:20 <nickm> we'll see!
17:17:27 <nickm> ok, bug-tracking for 029alpha.
17:17:41 <nickm> isabela: what do you think about using the 0.2.9.x-final milestone for that?
17:18:15 <isabela> sounds good / i wonder how is it looking right now, let me check
17:18:48 <isabela> 16 tickets there
17:19:08 <nickm> I pulled everything out of it that I was sure was not "must fix before 029"
17:19:36 <nickm> some of the remaining things might also be "okay to wait".  But I'd prefer not to decide those solo; having another dev looking too would be better IMO
17:19:43 <nickm> sounds ok?
17:19:53 <isabela> yes
17:20:10 <isabela> I just want to have the right instructions for the testers
17:20:32 <nickm> sounds good.
17:20:36 <nickm> there's also the "version" field.
17:20:41 <nickm> but that gets used weirdly.
17:20:51 <isabela> yeah, that's true
17:21:06 <nickm> (If anybody wants to look at the 029 milestone with me after the meeting, just poke me here or in #tor-project)
17:21:17 <isabela> about not deciding solo - should i drive a call for folks to review and triage things there / maybe do it on the list?
17:21:29 <isabela> ah
17:21:33 <isabela> that works too :)
17:21:36 <nickm> let's see how "informally" goes before we start begging.
17:21:42 <nickm> (it's pretty fun IMO)
17:21:44 <isabela> hehehe
17:21:48 <isabela> sounds good
17:23:06 <isabela> btw we should organize a triage for 0.3.0 / and this release will be 3months as we discussed in seattle?
17:23:29 <nickm> Maybe in November?  What do you think, dgoulet and asn?
17:23:55 <dgoulet> yup sounds good to me in November
17:23:57 <isabela> nickm: if is the week of nov 7 would be better for me
17:24:07 <asn> sounds good
17:24:07 <isabela> or after nov 7
17:24:23 <nickm> No strong preference from me
17:24:34 <dgoulet> I'll be at R meeting on the last of Nov. so before would be great
17:24:40 <isabela> ok
17:24:41 <dgoulet> last week*
17:24:53 <isabela> i will prep for it around the week of nov 7
17:25:12 <dgoulet> cool
17:25:34 <asn> #item there is no bug triager for this week
17:25:39 <dgoulet> ^!
17:25:42 <nickm> So, wrt end-of-life on old tors.  With the fix for TROVE-2016-10-001 a.k.a #20384, I was kinda banging my head against the fact that we're kind of sort of supporting 6 series.
17:26:11 <nickm> (024 through 029)
17:26:48 <isabela> hmm
17:27:02 <dgoulet> yeah that's quite a large amount of version to support...
17:27:02 <nickm> I think we want to get in a position where some of our series are "extended support", and some aren't.
17:27:14 <isabela> yep
17:27:37 <nickm> But I think that supporting all of 024...028 for as long as any OS has them is not going to work out.
17:27:39 <isabela> i think we should think of that considering current capacity as well
17:28:23 <nickm> I'm okay with being a bit lenient on 024..027, since we haven't previously announced any support schedule for them, and people may have gotten to expect us to support things forever...
17:28:35 <isabela> we should decide how many series back we will always support / then start a sunset plan for the ones we are now cutting out / then make sure that moving forward we will only support x series back
17:28:39 <nickm> ...but I think that for 028 and 029 and 030 we should probably announce something ahead of time and stick to it.
17:28:42 <nickm> yup.
17:29:08 <nickm> I started gathering data from weasel about how debian's situation looks, but it appears that there are also ubuntu and *bsd and etc to think about
17:29:37 <nickm> We could look for a volunteer LTS maintainer for every old series... :)
17:29:49 <isabela> if we give people time to upgrade it should not be a terrible thing
17:29:51 <nickm> Maybe I should start a wiki page and we can start adding stuff to it?
17:30:27 <dgoulet> nickm: so I agree with the above that is from 024 to 027, we haven't have any "policy" in place but 028+ we should be more strict
17:30:46 <dgoulet> nickm: maybe a pad as we gather information and then we take a decision on our support life cycle?
17:30:58 <dgoulet> and then add the decisions on the Wiki
17:31:15 <isabela> +1
17:31:29 <isabela> and then
17:31:40 <isabela> we should also have a blog psot and a lot of promotion of such decision
17:31:41 <nickm> okay, https://pad.riseup.net/p/tor-lts .
17:31:43 <dgoulet> I have this feeling also this ties to what's being discussed since yesterday on #20431
17:31:53 <isabela> and a date for people / a deadline
17:32:59 <nickm> dgoulet: I think that ticket is kind of presuming an answer to this question, and we might want to answer it first.
17:34:26 <nickm> or rather, if we answer our EOL/LTS plans, we have a much better answer for that ticket.
17:34:29 <nickm> y/n?
17:34:34 <dgoulet> yes exactly
17:36:51 <nickm> asn: two options for triage this week: nick does it; we try to find somebody else who can.
17:37:03 <asn> i can do it if no one else steps up
17:37:12 <isabela> wow wheezy will support 0.2.4 till 2018
17:37:26 <nickm> let's look among the folks who haven't done it yet first.  Would you like to ask around or shall I?
17:37:29 <asn> yeah that's crazy. we should not have to support 0.2.4 since then just because wheezy is ancient.
17:37:36 <nickm> isabela: LTS is not the same as actual support...
17:37:38 <asn> nickm: sure i can ask around
17:37:45 <dgoulet> asn, nickm: it's not that crazy, I did last week and I actually enjoyed it as I could learn all new things :)
17:37:58 <nickm> asn: very much appreciated.  I'll take this turn if no steps in.
17:38:00 <asn> athena: you wanna be bug triager of this week?
17:38:04 <isabela> nickm: what is the difference here?
17:38:29 <asn> i dont think there is anyone else in line apart from athena actually. teor and isis come to mind.
17:38:30 <nickm> isabela: I think that LTS is like "older than oldstable that still gets some security fixes" and that it's a separate support team?
17:38:56 <isabela> asn: maybe yawning too
17:39:18 <asn> yawning is not even in channel
17:39:46 <asn> ok i can become bug triager this week again np
17:39:59 <asn> i will add my name to the wiki
17:40:28 <isabela> ok
17:40:30 <asn> done
17:40:34 <isabela> i will send a note via email as well
17:40:50 <asn> a note on what? :)
17:41:06 <isabela> asking others to step up for the upcoming weeks
17:41:17 <asn> right
17:41:32 <isabela> specially who is not a volunteer
17:42:02 <nickm> let's see how it goes :)
17:42:13 <nickm> anything else for this week? or shall I endmeeting and shift attention to the pad?
17:42:33 <isabela> i have another question on TLS, sorry
17:43:05 <isabela> even though is different from normal support and is a separate team
17:43:10 <isabela> they should have a limit as well
17:43:11 <isabela> right?
17:43:24 <nickm> are you talking about debian LTS or our LTS or what?
17:43:33 <nickm> i was talking about debian lts
17:43:33 <isabela> our
17:43:36 <nickm> ah.
17:43:40 <nickm> for ours, we could set the rules
17:43:56 <isabela> k
17:44:47 <nickm> though if linux distros don't like our rules, we will get packaged less, and/or worse, and/or make packagers angry.
17:44:51 <nickm> none of those would be great.
17:45:56 <isabela> ok
17:46:11 <isabela> maybe we could write our rules and give them a chance for feedback/questions
17:46:19 <isabela> before we announce them to the world
17:46:26 <isabela> should we do that for others too?
17:46:41 <nickm> others?
17:47:29 <isabela> like bsd ppl
17:47:37 <isabela> or just linux distros?
17:47:45 <nickm> bsd people, tails, whonix, etc
17:47:47 <nickm> yes
17:47:58 <nickm> whoever cares
17:48:08 <isabela> ok
17:49:31 <nickm> more for this meeting?
17:49:42 * dgoulet is good
17:49:49 <nickm> #endmeeting