20:03:28 <isabela> #startmeeting moat + tor launcher
20:03:28 <MeetBot> Meeting started Wed Dec 20 20:03:28 2017 UTC.  The chair is isabela. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:03:28 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
20:03:35 <isabela> alright!
20:04:03 <isabela> we have a few bugs to cover
20:04:17 <isabela> isis: do you want to share udpates on those tickets we mentioned in the email
20:04:34 <isabela> maybe we can start with the one mcs sent
20:05:10 <isis> #24636 and #24637?
20:05:32 <isis> or #24689
20:05:37 <isabela> #24432
20:07:19 <isis> oh, i had actually forgotten about that one, oops
20:07:53 <isis> there's something funny happening with the last layer of redirection between apache and the moat server
20:08:31 <isis> i added more logging to the server, so now i should be able to go see what's actually being requested and why it's not found
20:09:28 <isabela> i guess my question would be
20:09:56 <isabela> realistic speaking with holidays and everything else, how are we with the server side
20:10:34 <isabela> because i dont even know how much time mcs has on their side
20:10:45 <isabela> giving they need to still get this on alpha etc
20:11:00 <isis> do we have a date for the alpha?
20:11:45 <mcs> on our side, we are about ready to put the code out for review. we have a few loose ends but they can be handled via followup tickets.
20:12:13 <mcs> we would really like a server that people can test against though!
20:12:53 <isabela> yes, so my question is, to have something we can test
20:13:03 <isabela> and the time to do it so
20:13:55 <isis> so make it not listen through meek, or just fix the current tunneling?
20:14:27 <isabela> i would prefer if we can have somethign that is in a build by end of jan
20:14:39 <isabela> i dont know what is necessary on server side for it to be working
20:14:46 <mcs> gk has the exact date for the next Tor Browser alpha, but I think it is mid-January
20:14:47 <isabela> what is left on these bugs to do it
20:14:55 <isabela> yes is mid jan
20:15:52 <isabela> jan 23 is the release of alpha 8.0a1
20:16:16 <isabela> and the idea was to have it working on that at least
20:16:21 <mcs> I’d strongly prefer we make things work over meek if possible since we have coded it that way inside Tor Launcher ;)
20:16:23 <isabela> so we can test stuff too
20:16:29 <isis> so it looks like just fixing the redirections for #24432 and specifying what moat does if it doesn't have enough bridges #24637 and fixing what it does if there are no bridges
20:16:55 <mcs> I don’t think I can ask people to review our client code without something to use it with on the server side….
20:17:17 <isabela> yeah
20:17:38 <isabela> so thinking of that, to give people tiem to review etc and then having time to get it in the release etc
20:17:54 <isabela> we are running pretty close ere
20:17:58 <isabela> *here
20:18:13 <isis> i think i can get it done before i leave friday
20:18:31 <mcs> That would be very awesome!
20:18:39 <isabela> ok, but in case you cant, what would be a plan b?
20:18:50 <isabela> i just want to make sure we plan for it
20:20:00 <isis> uh… plan b would be either find someone else to do it, or wait until i come back on 3 january (i think… i don't actually have a return flight yet)
20:20:08 <mcs> Also, is there anything that needs to happen on the deployment side for BridgeDB? We have never used meek in front of it before, right?
20:20:44 <isabela> isis: i dont think we can find someone else to do it tho, not now at this time. but i think if you wuold be able to do it when you are back that could be plan b
20:20:45 <mcs> (maybe something can be done in parallel if some infrastructure setup needs to be done by someone who is not Isis)
20:21:11 <isis> it it has never used meek before , no
20:21:38 <isis> but it's already set up
20:22:06 <mcs> OK. I wasn’t sure if that was “real stuff” but that is good news.
20:22:13 <isis> (as described on #24432 and #16650)
20:22:17 <mcs> So it is a simple matter of writing some code :)
20:22:45 <mcs> (after debugging I guess)
20:23:17 <isis> one bad thing is that it means BridgeDB now requires a maintainer who can do Twisted Python and Golang
20:23:25 <isis> which probably isn't the most common of skill combos
20:23:27 <isabela> isis: how we catch up on friday before you leave to know how things are going?
20:24:12 <mcs> FYI, brade and I will be afk on Friday and we will only be available a little next week
20:24:20 <mcs> But we will check email.
20:24:20 <isabela> yes
20:24:36 <isabela> i will be away too i have an appointment with immigration that day
20:24:46 <isis> i will be afk friday as well, but i could be around a little bit
20:24:54 <isabela> but i just want to have the right expectations for january
20:25:02 <isabela> because if this is not done
20:25:11 <isabela> and we will have to figure out what to do in january
20:25:13 <isis> i can send out an update tomorrow night via email, or just tagging everyone in irc
20:25:26 <isabela> email would be better
20:25:30 <mcs> please send via email since I may not be on IRC :)
20:25:39 <isis> okay! will do
20:25:42 <mcs> thx
20:25:58 <isabela> isis: i would appreciate over communication on it because it will help me understand what to expect for january
20:26:31 <isis> definitely
20:26:39 <isabela> thank you
20:26:56 <isis> sorry for the rush right before the holidays
20:27:16 <isis> i had completely forgotten that #24432 still had a problem :/
20:28:47 <mcs> Maybe you will have time for #24636 and #24637 as well? I think those should be straightforward to fix.
20:28:52 <mcs> (one i just a spec change)
20:28:57 <mcs> s/i/is/
20:29:37 <isis> yeah, those should be easy
20:30:34 <isabela> cool - so you will look over these 3 tickets and get back to us via email during the rest of the week
20:32:03 <isis> oh, the reason for #24637 that it only ever returned 1 bridge to kathy wis that if it doesn't have enough bridges total (i'm assuming you only generated 200 descriptors) then it tries to cut down on the amount it hands out per request
20:32:17 <isis> it does that for all the distributors, not just moat
20:32:30 <mcs> ah, okay. makes sense.
20:32:38 <isis> on a normal system with thousands of bridges it'll hand out three
20:32:42 <mcs> so we should be prepared to handle 3, usually?
20:32:48 <mcs> (or N I suppose)
20:32:50 <isis> yep!
20:32:57 <mcs> OK
20:33:06 <isis> we can set it to a different number if we want, there's a config option
20:33:46 <mcs> Our client code only uses the first one at the moment but we will fix that soon.
20:34:26 <isis> oh okay cool
20:36:43 <isis> was there anything else to catch up on?
20:37:07 <isabela> not from me, not sure if mcs has any other questions
20:37:26 <mcs> I am good for now.
20:37:34 <mcs> It is exciting to see this come together!
20:37:42 <isabela> cool!
20:37:47 <isabela> i will stop the bot then
20:37:53 <isabela> #endmeeting