[00:00.860 --> 00:05.500]  We're all from the Electronic Frontier Foundation that we've been working on for the last few years.
[00:07.160 --> 00:10.180]  And it's getting better and better as, you know, the days go on.
[00:11.100 --> 00:13.880]  So just by way of introduction, I'm Adi Kamdar.
[00:14.000 --> 00:15.580]  I'm an activist at EFF.
[00:16.360 --> 00:17.020]  Do you want to go down?
[00:17.200 --> 00:17.360]  Sure.
[00:17.660 --> 00:18.540]  I'm Nate Cardozo.
[00:18.740 --> 00:20.300]  I'm one of the staff attorneys at EFF.
[00:20.520 --> 00:24.500]  And one of the several things that I work on is the Coders Rights Project.
[00:24.500 --> 00:32.980]  So I defend hackers, coders, innovators, programmers, try and keep you guys out of trouble and try and help you get out of trouble when you get in it.
[00:33.120 --> 00:35.380]  And that's part of what I'm on this panel for.
[00:37.020 --> 00:37.980]  Hi, and...
[00:39.840 --> 00:40.560]  Is it working?
[00:41.180 --> 00:41.620]  Exactly.
[00:42.720 --> 00:44.560]  And I'm Ranga Krishnan.
[00:44.840 --> 00:46.780]  I'm a Technology Fellow at the EFF.
[00:47.080 --> 00:52.820]  And I actually joined just last year to work specifically on this project.
[00:52.820 --> 00:57.800]  And it has been very exciting and I hope to show where we are with it today.
[00:59.180 --> 01:04.900]  So the Open Wireless Project, it has a lot of roots.
[01:05.340 --> 01:14.560]  But one of the issues that this is trying to solve is one that we may have all experienced, that I definitely experienced.
[01:14.980 --> 01:20.340]  In fact, the other day I was at home and my Internet went out.
[01:20.340 --> 01:25.160]  And I checked the available networks and there were 15 networks.
[01:26.500 --> 01:28.300]  And most of them were locked.
[01:29.480 --> 01:32.780]  And I couldn't access anything, which was unfortunate.
[01:33.980 --> 01:38.080]  And I did see one unlocked network and it was called Big Blue House Guest.
[01:39.020 --> 01:41.880]  And I knew exactly which house this was on my street.
[01:42.480 --> 01:43.780]  And I'm like, great.
[01:43.960 --> 01:47.440]  You guys are gonna allow me to use your Internet.
[01:47.680 --> 01:48.940]  Thank you, friendly neighbors.
[01:49.940 --> 01:52.960]  Unfortunately, I ran into this right afterwards.
[01:53.200 --> 01:54.620]  A nice captive portal.
[01:55.460 --> 01:59.440]  And it says, you know, enter the guest account password, which I didn't know.
[01:59.560 --> 02:02.720]  And I didn't really want to go over and knock on the door and ask for the password.
[02:04.320 --> 02:15.560]  This is the sort of issue where, you know, we're living in a world where walking down the street, sitting at home, in the office, whatever, we have a bunch of wireless networks available.
[02:15.560 --> 02:17.980]  Yet none of them are actually accessible.
[02:19.320 --> 02:22.340]  And that seems inefficient to us.
[02:22.720 --> 02:31.200]  There's another story from around a couple years ago when we were launching this project, where Hurricane Sandy happened.
[02:31.500 --> 02:33.160]  How many of you guys are from around here?
[02:34.300 --> 02:35.860]  I can't see anything.
[02:36.120 --> 02:37.660]  So I'm assuming there's some hands up.
[02:38.460 --> 02:40.740]  I realized that was useless.
[02:42.060 --> 02:46.240]  Well, I'm sorry you guys had to experience Hurricane Sandy, because it seemed awful.
[02:46.440 --> 02:57.000]  But one of the things that came out of that is with the power outages and with the outages of service, we saw things like this.
[02:57.140 --> 02:58.580]  This was a friend who tweeted this out.
[02:58.920 --> 03:00.620]  Awesome picture of Internet infrastructure.
[03:00.760 --> 03:03.760]  New Yorkers idling for free working Wi-Fi outside of Starbucks.
[03:04.500 --> 03:06.520]  And this is actually right around the corner.
[03:07.900 --> 03:14.060]  Starbucks was one of the few places that was offering free Wi-Fi that was working and that people could actually access.
[03:14.060 --> 03:18.020]  And, you know, these were people who needed to get in contact with people.
[03:18.560 --> 03:19.740]  Needed to check their email.
[03:19.900 --> 03:23.980]  Needed to send out, you know, I'm okay, I don't know, Twitter updates.
[03:25.260 --> 03:27.960]  These were people who needed to use the Internet and yet they couldn't.
[03:29.140 --> 03:37.940]  And the one place that was offering some sort of open Wi-Fi, some sort of free Wi-Fi, was Starbucks.
[03:38.540 --> 03:42.720]  So, with this in mind, we started the open wireless movement.
[03:42.960 --> 03:48.320]  And the point of the open wireless movement is to encourage open wireless.
[03:48.540 --> 03:51.800]  It's to create a world where open wireless is the norm.
[03:53.520 --> 03:59.160]  And, you know, we're imagining a world, a future, with ubiquitous open Internet.
[04:01.060 --> 04:02.140]  Why would you do this?
[04:02.220 --> 04:04.160]  Why would you open up your wireless networks?
[04:06.280 --> 04:07.220]  It's a good question.
[04:08.000 --> 04:09.600]  Part of it could be because of sharing.
[04:10.160 --> 04:12.100]  Sharing is actually a very important thing.
[04:12.540 --> 04:15.060]  You know, not all of us are using all of our bandwidth all the time.
[04:15.060 --> 04:16.820]  Not all of us are using our networks all the time.
[04:19.000 --> 04:23.080]  We've been in situations where we needed to get on the Internet and we couldn't.
[04:23.260 --> 04:25.360]  And it would be nice if someone shared their Internet with us.
[04:27.320 --> 04:27.680]  We...
[04:27.680 --> 04:32.640]  EFF actually has a board member, Bruce Schneier, who runs an open wireless network.
[04:32.820 --> 04:33.540]  And it says...
[04:33.540 --> 04:34.920]  I like his quote a lot.
[04:35.240 --> 04:39.040]  He says, to me, it's basic politeness to run an open wireless network.
[04:39.920 --> 04:44.060]  Providing Internet access to gas is kind of like providing heat and electricity or a hot cup of tea.
[04:44.060 --> 04:45.420]  And who doesn't like that?
[04:46.260 --> 04:50.340]  But besides access, why open wireless?
[04:50.740 --> 04:52.100]  Access can't be the only issue.
[04:52.220 --> 04:56.520]  There are a lot of groups and people advocating for access to the Internet.
[04:56.740 --> 04:59.800]  And that's an issue that a lot of people are dealing with right now.
[05:00.220 --> 05:02.960]  And open wireless by no means is the be-all end-all of this.
[05:03.120 --> 05:05.880]  But it is a step in the right direction.
[05:06.340 --> 05:08.400]  One of the big reasons is privacy.
[05:09.240 --> 05:12.440]  Now, how can opening up your network be conducive to privacy?
[05:12.440 --> 05:17.660]  And this is something we're actually going to talk about quite a bit from a legal perspective and a technical perspective.
[05:19.300 --> 05:27.040]  But one reason is because right now when we're mobile, right now when we're roving around, we're tied to these things.
[05:27.140 --> 05:28.140]  We're tied to our smartphones.
[05:30.180 --> 05:33.200]  And when we want to access data, when we want to access the Internet.
[05:33.620 --> 05:39.580]  And these smartphones use networks offered by folks like these.
[05:41.840 --> 05:43.520]  And that's kind of an issue, right?
[05:43.640 --> 05:54.560]  If we want to be mobile, if we want to use devices, if we want to connect to the Internet and we're not close to a wireless network that we have access to, we have to pay a contract.
[05:54.700 --> 05:56.200]  We have to use these sorts of services.
[05:56.660 --> 05:59.520]  And our smartphones are essentially tracking devices.
[06:00.140 --> 06:11.020]  So wouldn't it be nice if the privacy world we imagine actually fostered the sort of innovation that we're seeking?
[06:11.760 --> 06:21.360]  If it actually fostered innovation that didn't have to rely on these few big players and could actually take advantage of all of these wireless networks running through us right now.
[06:21.960 --> 06:23.940]  And this could be any sort of innovation.
[06:24.300 --> 06:26.240]  Here are four random pictures I found.
[06:26.840 --> 06:28.300]  If you want to connect your dog.
[06:28.600 --> 06:31.920]  If you want to have a watch that does something smart.
[06:32.020 --> 06:33.480]  If you want a face mask.
[06:33.640 --> 06:35.500]  I'm not actually quite sure what that is, but it looked kind of cool.
[06:37.520 --> 06:39.560]  And it could be connected to the wireless Internet.
[06:39.800 --> 06:40.600]  And wouldn't that be great?
[06:42.260 --> 06:42.660]  Potentially.
[06:42.920 --> 06:43.480]  I don't know.
[06:43.480 --> 06:47.420]  But these are the sorts of innovations that you can't necessarily use out in the open.
[06:47.700 --> 06:55.520]  Because, you know, AT&T and Verizon and so have this sort of domain covered.
[06:57.260 --> 07:00.400]  One of the big questions that people ask is, all right, fine.
[07:01.480 --> 07:04.120]  I'm maybe okay with opening up my wireless Internet.
[07:04.320 --> 07:06.600]  I may be okay with opening up even a part of it.
[07:08.380 --> 07:12.360]  But doesn't this, doesn't opening up my wireless network just reward freeloaders?
[07:13.060 --> 07:18.980]  Can't people just get on and do whatever they want with it and not pay for anything?
[07:19.420 --> 07:20.280]  And doesn't that suck?
[07:21.040 --> 07:22.480]  And our answer is no.
[07:23.880 --> 07:27.980]  You know, this is the sort of world that we want to live in.
[07:28.080 --> 07:40.260]  We're trying to imagine a future where having that sort of culture of sharing and having that sort of culture of openness is something that we can actually foster and actually go towards.
[07:41.380 --> 07:46.200]  So we're trying to fight against this mentality that this is a bad thing.
[07:46.340 --> 07:47.600]  That this is my wireless Internet.
[07:47.740 --> 07:50.180]  I need to lock it down and have it be mine, mine, mine.
[07:50.960 --> 07:58.700]  Because we think with sharing, with open wireless, it's sort of a rising tides effect and everybody just benefits.
[07:59.420 --> 08:03.000]  So with that, we're going to get into the fun legal matters.
[08:05.200 --> 08:05.580]  That's me.
[08:06.500 --> 08:08.060]  As I said earlier, I'm...
[08:08.060 --> 08:08.840]  Is this on?
[08:08.920 --> 08:09.300]  Can you hear me?
[08:09.720 --> 08:09.860]  Yeah.
[08:10.500 --> 08:10.700]  Great.
[08:12.060 --> 08:13.560]  As I said earlier, I'm Nate Cardozo.
[08:13.620 --> 08:15.140]  I'm a staff attorney at EFF.
[08:15.760 --> 08:19.420]  And one of the things I do is counsel the open wireless project.
[08:19.420 --> 08:20.740]  I also do coders rights.
[08:21.000 --> 08:23.900]  So now is my quick little lawyer stuff disclaimer.
[08:24.120 --> 08:24.780]  I am a lawyer.
[08:25.220 --> 08:26.560]  I'm not your lawyer.
[08:26.840 --> 08:29.060]  And nothing I say will change that.
[08:29.400 --> 08:32.980]  Unless you're already one of my clients, in which case I'm not talking to you right now.
[08:34.280 --> 08:36.740]  So I will not be giving legal advice today.
[08:36.740 --> 08:44.680]  If you come up to the microphone after we're done talking, which we will invite you to do, and you ask a question that calls for legal advice, this is not a confidential setting.
[08:44.680 --> 08:46.680]  So I can't give it to you in this setting.
[08:46.860 --> 08:50.200]  I will give you my card and we can talk about it later in a confidential setting.
[08:50.380 --> 08:51.820]  But this isn't one of those.
[08:52.420 --> 08:52.600]  Okay.
[08:53.580 --> 08:53.980]  Copyright.
[08:54.700 --> 08:55.880]  There are several...
[08:55.880 --> 09:03.520]  I'm going to address sort of three, sort of a little more than three sets of reasons why open wireless isn't a bad idea.
[09:04.840 --> 09:08.620]  Copyright is one of the objections that a lot of people have to open wireless.
[09:08.940 --> 09:10.560]  And what do I mean by that?
[09:10.560 --> 09:24.160]  If I torrent a whole ton of movies without using any sort of protection online, chances are I'm going to get a DMCA notice or six strikes notice or some sort of complaint, right?
[09:25.580 --> 09:30.900]  If I don't do that and I open my wireless network to others, is that a problem?
[09:31.680 --> 09:32.200]  No.
[09:32.580 --> 09:33.760]  I don't think it is.
[09:34.520 --> 09:37.540]  And there's a number of reasons why that's the case.
[09:39.060 --> 09:47.740]  If you're not the person doing bad stuff online, especially with copyright, you're not going to be held responsible.
[09:48.780 --> 09:56.400]  In copyright, there's one way that you can be held responsible if you're not the one doing the bad thing, copyright infringement.
[09:56.560 --> 09:57.920]  And that's secondary liability.
[09:58.120 --> 09:59.220]  And that comes in two flavors.
[10:00.100 --> 10:02.960]  Contributory liability and vicarious liability.
[10:03.720 --> 10:22.340]  These are doctrines which are extremely fuzzy and have a lot of ins and outs, but what we need to know about them for these purposes is that contributory liability requires essentially that the network operator have knowledge of what's going on and caused or induced that behavior.
[10:22.340 --> 10:34.320]  And for vicarious liability, the network operator has to have the right and ability to control the copyright infringement going on and receive direct financial benefit from the copyright infringement.
[10:34.600 --> 10:42.520]  Luckily, with open wireless, those factors are not satisfied for either contributory or vicarious liability.
[10:43.140 --> 10:55.240]  If I open the wireless network at my house, you know, I don't have actual knowledge that anyone is using it to do copyright infringement unless they come tell me.
[10:55.440 --> 10:58.280]  And I'm certainly not inducing them to do so.
[10:58.420 --> 11:00.600]  I'm not saying this is...
[11:00.600 --> 11:09.960]  You know, unless I put a big sign up in front of my house that you connect to openwireless.org SSID and use it to torrent your movies, I'm not inducing copyright infringement.
[11:10.140 --> 11:10.760]  Don't do that.
[11:10.760 --> 11:12.160]  That's a bad idea.
[11:13.380 --> 11:15.560]  Vicarious liability is similarly the same.
[11:16.440 --> 11:23.160]  If I'm a neutral network operator, I don't have the right and ability to control what people are doing on my network.
[11:24.160 --> 11:27.540]  And I'm not receiving a direct financial benefit from it either.
[11:28.680 --> 11:36.380]  If you have an open wireless network at your house, chances are you are not receiving a direct financial benefit from that.
[11:36.980 --> 11:40.220]  But there's some other things that can help us out here too.
[11:40.440 --> 11:45.600]  And this is one of the aspects of the DMCA, which is not entirely bad.
[11:46.000 --> 11:48.220]  The DMCA... and there's the code number.
[11:48.340 --> 11:52.340]  I was told that maybe if there are lawyers in the crowd, they can get CLE credit for this.
[11:52.480 --> 11:58.820]  So that's why I actually included the case... there's some case titles later down and the statute.
[11:58.820 --> 12:01.540]  So there's section 512 of the DMCA.
[12:02.680 --> 12:04.500]  Creates a safe harbor for network providers.
[12:04.860 --> 12:06.260]  This is completely optional.
[12:06.360 --> 12:09.680]  Network providers do not have to take advantage of it if they don't want.
[12:09.840 --> 12:11.180]  But it's often a good idea.
[12:11.300 --> 12:14.300]  And we have a plan for how they can do this with open wireless.
[12:15.580 --> 12:18.160]  We've crafted a sample policy.
[12:18.560 --> 12:30.880]  One of the requirements for the safe harbor, if you want to take advantage of the DMCA safe harbor provision, is that you have a policy, a DMCA policy, and we have a sample for you.
[12:30.880 --> 12:38.240]  And the policy says that, you know, you can't use this for copyright infringement.
[12:38.720 --> 12:43.320]  And that you... the network operator will terminate repeat infringers.
[12:44.180 --> 12:49.380]  And the other part of the safe harbor is that it must accommodate rights holders' technical measures.
[12:49.800 --> 12:52.140]  Those technical measures must be jointly agreed upon.
[12:52.240 --> 12:54.360]  That's never happened, so we don't really have to worry about that.
[12:55.340 --> 12:57.180]  So, safe harbor.
[12:57.180 --> 13:09.980]  However, you... in order... oh, 512M, subsection M of section 512, says that the network provider does not need to affirmatively monitor the network to qualify for safe harbor.
[13:10.160 --> 13:12.460]  And that it even comes under the nice little subheading privacy.
[13:13.340 --> 13:15.720]  So, what is safe harbor in general?
[13:16.200 --> 13:24.280]  It is a way that will provide... it is a way to provide immunity to the network provider to DMCA complaints.
[13:24.900 --> 13:42.200]  We think that the sample policy that we've crafted at openwireless.org and running an open wireless network at home or at your small business... and if you post this policy or if you refer to it, which we think we're doing by having your SSID say openwireless.org,
[13:42.300 --> 13:46.540]  and the policy is right there at openwireless.org, we... there's a good argument.
[13:46.540 --> 13:49.740]  It is untested... that... that would qualify.
[13:49.960 --> 13:53.200]  And you would not be liable under the DMCA.
[13:54.860 --> 13:55.580]  Negligence.
[13:57.360 --> 14:01.640]  Some lawyers... and I won't call them trolls, but they're trolls...
[14:01.640 --> 14:10.120]  have... have decided that the very act of running an open wireless network at your house is negligent.
[14:11.840 --> 14:20.860]  Luckily, they're wrong, and the two courts who have addressed it directly have told them that they're wrong in very, very strong language.
[14:21.540 --> 14:25.460]  Running an open wireless network is not negligence.
[14:26.000 --> 14:26.640]  Why?
[14:27.020 --> 14:32.420]  I mean, probably because that's what wireless was originally meant to do, right?
[14:32.660 --> 14:38.400]  It's not a breach of the duty of care if that's the standard, which is completely true.
[14:40.300 --> 14:40.820]  Risk.
[14:41.220 --> 14:48.560]  This is... this is something... this is one of the more strong objections to open wireless.
[14:48.900 --> 15:01.340]  When we... after Andy Greenberg's story on this project came out a couple of weeks ago, the Reddit thread was dominated by people saying, if I run an open wireless network at my house, I'm going to get SWATed, right?
[15:01.440 --> 15:09.260]  I'm going to get a SWAT team busting down my door for people downloading child porn and doing whatever else bad people do on the Internet.
[15:10.940 --> 15:14.000]  That is not completely out of the question.
[15:14.000 --> 15:19.040]  However, a lot of people run open wireless today, and it doesn't...
[15:19.040 --> 15:22.200]  the SWAT team scenario doesn't happen very often.
[15:22.560 --> 15:28.620]  And the more people who run open wireless, the less likely this is going to be.
[15:29.140 --> 15:32.660]  So there's a little bit of a first mover problem, but luckily you're not going to be the first mover.
[15:32.660 --> 15:34.340]  The first mover was years and years ago.
[15:36.340 --> 15:44.980]  If... you know, if you're running open wireless and police comes to investigate and they see that you're running open wireless, you know, maybe that would help you out a little bit.
[15:45.740 --> 15:52.040]  You know, sometimes police don't do that thorough of an investigation before they break down your door, so that's not a... that's not surefire.
[15:53.200 --> 15:57.040]  But there are some other things we can do to minimize... to mitigate the risk here.
[15:57.720 --> 16:18.080]  And one of those things is in a future release of the software that Ranga is going to talk about, we're going to have the option of routing all of the guest network traffic over a VPN of your choice or over Tor, and that will not lead anyone back to your IP.
[16:18.080 --> 16:26.320]  Obviously this is... I'm not recommending that you do your child porn surfing over Tor or a VPN.
[16:26.440 --> 16:28.200]  In fact, I recommend that you do not do that.
[16:29.080 --> 16:35.160]  But hopefully most people don't... don't do really horrible things online.
[16:36.660 --> 16:44.700]  You know, I've been running an open wireless network at my house for years, and I've never been swatted, but, you know, I don't think I'm unique in that.
[16:44.800 --> 16:49.300]  So I think the risk... the risk is there, but it is overstated.
[16:50.320 --> 16:55.740]  There are some other things which I don't really have a slide for, which I do want to talk about, as Adi said.
[16:57.040 --> 17:02.900]  How does open wireless increase your privacy and the privacy of others?
[17:03.740 --> 17:15.600]  Right now there's a... I wouldn't say it's a legal theory, but there's a sort of prevailing attitude that a... a person's identity can be determined simply by their IP address.
[17:16.120 --> 17:17.300]  We think that's wrong.
[17:17.520 --> 17:22.720]  We think an IP address can sometimes be a proxy for identity, but often isn't.
[17:23.000 --> 17:25.720]  And that is absolutely not conclusive.
[17:25.920 --> 17:27.740]  And courts are starting to come around to this.
[17:28.260 --> 17:37.940]  It's... it's been tested most often in the copyright troll situation, where, you know, Malibu Media sues a 79-year-old grandmother for downloading unspeakable title.
[17:39.120 --> 17:43.180]  And the 79-year-old grandmother says, look, I wasn't the one who downloaded unspeakable title.
[17:44.040 --> 17:45.620]  It couldn't possibly have been me.
[17:45.780 --> 17:48.440]  And they say, well, you're the subscriber on the IP address.
[17:48.600 --> 17:50.640]  And she said, yeah, but that doesn't mean it was me.
[17:51.000 --> 18:01.480]  So the more people who do open wireless, the harder it is going to be as a legal matter to claim that an IP address is your identity and that your identity is your IP address.
[18:02.740 --> 18:04.780]  We... we don't think that's a good thing.
[18:06.900 --> 18:18.140]  Secondly, this sort of ubiquitous tracking that is permissible technically due to the fact that we all have a tracking device in our pockets.
[18:18.540 --> 18:26.640]  And as Adi alluded to, the network is controlled generally by a very small handful of very large providers.
[18:27.440 --> 18:29.520]  Open wireless can go a long way to change that.
[18:29.520 --> 18:38.720]  And the sort of ubiquitous tracking that something like an AT&T smartphone allows will be a lot harder if I'm hopping from...
[18:38.720 --> 18:41.860]  hopping anonymously from one wireless network to another.
[18:43.360 --> 18:47.220]  And, you know, ubiquitous tracking of everybody is not good, right?
[18:47.320 --> 18:53.920]  One of the things that we do at EFF is we sue the NSA to make them stop tracking all of us all the time.
[18:54.520 --> 19:02.320]  Ubiquitous open wireless is not a panacea here, but it will make their job more difficult, which we think is good.
[19:04.340 --> 19:11.500]  You know, I guess the analogy here is it's creating a bigger haystack, which is good for you and good for the world.
[19:12.680 --> 19:14.220]  And that's the end of lawyer stuff.
[19:16.820 --> 19:17.860]  Should I move there?
[19:18.100 --> 19:19.140]  Yeah, you should.
[19:19.860 --> 19:21.260]  Okay, you can move the slides.
[19:21.400 --> 19:21.740]  I'll do this.
[19:25.460 --> 19:39.100]  So, I'm going to talk about where all of this thinking at the EFF led, which was that the concepts were great, but people who even wanted to deploy this open wireless had some technical challenges.
[19:39.520 --> 19:48.520]  And so, in order to make this real, we needed to address those technical challenges and make it easy for people to do what they really want to do.
[19:48.900 --> 19:51.060]  So, that's the motivation.
[19:51.520 --> 20:01.560]  And so, we decided we would create a router firmware that people could put onto existing hardware that they could get off the shelf.
[20:03.560 --> 20:05.440]  And that would be the start.
[20:05.440 --> 20:10.040]  And then we would see the kind of interest we got and then think about what the next steps are.
[20:11.160 --> 20:14.740]  So, we started off with, you know, how do we make sharing easy?
[20:15.660 --> 20:23.200]  But we realized that we needed to solve a couple of other problems, which is the security and speed.
[20:23.560 --> 20:26.660]  And so, my next slides are going to talk about that.
[20:26.820 --> 20:39.540]  But before that, I want to mention that the firmware image that we are launching today in conjunction with this conference is at that URL over there.
[20:40.520 --> 20:42.220]  You don't need to remember it.
[20:42.560 --> 20:49.160]  We'll have a blog post on the EFF website shortly, which will have details.
[20:49.620 --> 20:53.640]  If you're subscribed to the EFF Twitter feed, we'll be tweeting about it.
[20:53.640 --> 20:59.260]  So, if you're interested, we encourage you to download it and try it out.
[21:00.940 --> 21:01.460]  Go ahead.
[21:02.340 --> 21:09.660]  So, I wanted to explain, like, the three goals that I listed there for this router firmware, starting with security.
[21:10.000 --> 21:21.000]  And I put it up on top there because even though we started with sharing, we realized this was the biggest problem that we needed to solve.
[21:23.120 --> 21:31.320]  You know, the things that we needed to do for sharing were relatively small in comparison, whereas this was a huge problem that many people in industry have flagged.
[21:31.520 --> 21:36.080]  And there isn't really anybody with a solution today.
[21:36.540 --> 21:40.940]  And so, it's also the largest component of work that we need to do.
[21:40.940 --> 21:46.800]  And we see this as a multi-year or a multi-decade effort to really make this happen.
[21:49.120 --> 21:57.700]  So, basically, SOHO routers, the routers that most of you have in your home or small office, are hopelessly compromised.
[21:57.940 --> 22:09.480]  This has been documented over the course of the last year by many security firms, including our partner, Independent Security Evaluators, who've written a few articles about it.
[22:11.540 --> 22:19.080]  And one thing they noticed was even though they had published all this, there really wasn't much of a response from the manufacturer community.
[22:19.300 --> 22:28.580]  They were still shipping the same old things with the same old Linux kernels without the fixes and with the same broken web user interfaces.
[22:29.120 --> 22:55.720]  And so, what we decided is, you know, since a lot of the attacks do happen through the web UI that the routers exposed to users, we would develop an entirely new one, focused not only on security, but also on simplicity, because we wanted to make it really easy for people to use and set up open wireless networks.
[22:56.200 --> 22:59.120]  So, that was the first part of the initiative.
[22:59.120 --> 23:16.900]  I mentioned that we are collaborating with Independent Security Evaluators, but we are also working with other people in the technology, security industry, to help us, you know, pen test, audit, improve this router.
[23:19.100 --> 23:28.580]  And what we realized is one of the reasons that the router security is so bad is that the software on the routers is not updated.
[23:29.380 --> 23:38.340]  You know, your phone automatically updates its software, your apps automatically update, but router firmware is rarely set to do that.
[23:38.740 --> 23:45.700]  Some of the problems in the past were router updates could break routers because of their limited memory footprint.
[23:46.000 --> 23:47.680]  But today, they have plenty of memory.
[23:47.800 --> 23:49.100]  You can have two copies of images.
[23:49.360 --> 23:51.800]  You can, if you fail, you can go back to the old ones.
[23:51.940 --> 23:56.220]  So, these are problems that can be solved today, but are not widely solved.
[23:56.520 --> 24:22.380]  And so, we are focused on creating a secure software update mechanism for these routers, which is not only provide security by patching bugs, but we wanted to add an additional protection because, you know, of our concerns with privacy, concerns with overreach of government in targeting individuals,
[24:22.700 --> 24:31.280]  that the software update process itself could be a way in which people could be spied upon or their rights violated.
[24:31.280 --> 24:47.420]  And so, we decided that the way to address this was to use Tor, which is, you know, as many of you know, was another project that came out of the EFF, or was supported by the EFF, and we continue to support it.
[24:48.880 --> 25:06.320]  And the purpose of that was the router will fetch the information about what the most current software is, the metadata, that decides, you know, whether it is going to update and what the file that it needs to use to update is over the Tor network.
[25:06.680 --> 25:22.440]  And what this does is, you know, whatever server it is reaching out to, if there are intermediaries in between who can intercept that traffic, they cannot know who it is that is requesting this information, this metadata information.
[25:22.920 --> 25:41.820]  And so, in that case, EFF or whoever else, and, you know, one of the goals is to set an example of how this can be done so that other people can adopt this kind of practice, is then, you know, if you're requested to push a malware to a particular user,
[25:42.440 --> 25:45.200]  you can honestly say, we don't have a way to do that.
[25:45.200 --> 25:52.200]  We, we, there is no way in this system in which we can know that this is the person you're trying to reach.
[25:52.340 --> 25:54.940]  And the only way to send malware is to send it to everybody.
[25:55.440 --> 26:06.220]  And, and when one does that, it usually defeats the interest the government has, because then that bad malware would be, I mean, the malware would be detected very soon.
[26:06.620 --> 26:09.300]  And, and so they, there is no interest in doing this.
[26:09.300 --> 26:17.660]  So, in fact, this mechanism I'm talking about, we have it in our software that we're releasing today.
[26:19.920 --> 26:23.700]  And, we hope to make it more and more robust as, going forward.
[26:26.620 --> 26:36.440]  And, we are, one of the other aspects of security is being actually able to change the software in case there is a problem.
[26:37.600 --> 26:43.080]  Many routers use binary blobs, which, you know, can only be provided by the manufacturer.
[26:43.440 --> 26:45.480]  And, it creates two problems.
[26:45.600 --> 26:55.080]  One is, if there's a problem in the binary blob itself, nobody can fix it except the manufacturer, who has no interest because there's no business in this low-end router segment to be fixing these things.
[26:55.220 --> 27:01.920]  The team that built it has, you know, moved on and they're doing more interesting, more remunerative things.
[27:02.300 --> 27:03.880]  And, and there's nobody to do it.
[27:03.880 --> 27:10.020]  And, and secondly, um, once you freeze a binary blob, you freeze all of the software around it.
[27:10.100 --> 27:14.960]  So, you cannot even upgrade the Linux kernel around it if, you know, there were problems there.
[27:15.240 --> 27:28.580]  So, you know, it's very important to have, uh, uh, software that can be completely built from scratch, so that we can fix problems and really give ownership of the device, uh, to the, uh, homeowner.
[27:28.580 --> 27:30.920]  And that they are not beholden to anybody else.
[27:31.180 --> 27:39.180]  And in this, you know, we agree with our friends in the Free Software Foundation, who are also, uh, have a, um, uh, router initiative.
[27:39.560 --> 27:42.100]  Uh, and I encourage you to go and check that out.
[27:45.120 --> 27:47.480]  Uh, so the second thing is speed.
[27:47.700 --> 27:53.080]  And this is, um, people, let me just step back a little bit.
[27:53.080 --> 28:00.000]  So the security came out because people might say, hey, I don't want to have an open wireless channel because maybe I'm exposing myself to security problems.
[28:00.160 --> 28:01.520]  And I kind of skipped that part.
[28:01.720 --> 28:09.040]  And when we looked into that question, what we realized was, you know, routers were hopelessly broken, independent of having to open a wireless channel.
[28:09.300 --> 28:13.440]  And so, I mean, I, uh, so this story with speed is something similar.
[28:13.760 --> 28:22.440]  You know, people might say, hey, you know, what about when I want to use the Internet and you have a bunch of, you know, uh, guests that are, you know, downloading movies or something.
[28:22.600 --> 28:25.140]  And I, I get a terrible, uh, uh, experience.
[28:25.580 --> 28:38.240]  And so, uh, when we looked into this, uh, we realized that independent of having an open wireless network, people had terrible experience just if they were the only users.
[28:38.300 --> 28:45.660]  If they had, uh, you know, multiple applications, maybe a web browsing session, maybe voice, maybe some large download happening simultaneously.
[28:45.660 --> 28:56.720]  Typically, the latency-sensitive application, the voice, uh, or gaming, suffers significantly way beyond what is actually, uh, you know, required.
[28:56.940 --> 29:08.260]  And this is because of a problem called buffer bloat, where because memory is cheap, you know, buffers in devices have grown, and that tends to, uh, uh, buffer up big queues of packets.
[29:08.560 --> 29:12.340]  And then so that your latency-sensitive applications are not getting the performance they want.
[29:12.420 --> 29:14.080]  They're not getting that priority that's needed.
[29:14.080 --> 29:23.720]  And this is a problem that, uh, uh, uh, a few people in the community had pointed out, uh, uh, uh, networking community had pointed out a few years ago.
[29:24.040 --> 29:33.260]  And there's a project called CeroWRT, uh, uh, run by Dave Todd and, and Jim Geddes, that has been focused on, uh, fixing this problem.
[29:33.260 --> 29:48.020]  And fortunately for us, the, at this point, after about three years of research and implementation and bug fixing, they really are at a point where their work is, is complete, uh, at least in some aspects.
[29:48.020 --> 29:50.220]  And so we are leveraging that.
[29:50.420 --> 30:05.220]  And what that does is not only can we solve the problem of, uh, having a home user have a better experience, but those same algorithms, they help us share the Internet between the guest user and the host user in a better way.
[30:05.220 --> 30:08.720]  And so, um, we are, we are using CeroWRT for that.
[30:09.040 --> 30:18.760]  And then looking down into the future, uh, you know, we want to make, uh, uh, ISPs live up to their promise, those numbers they give you when they sell you the plans.
[30:19.340 --> 30:27.260]  Um, uh, uh, and, and so we are hoping to have software in there that will, uh, measure, uh, on a continuous basis.
[30:27.260 --> 30:32.100]  Because, you know, you can run speed test, a one shot test, but then you don't know what's happening in the middle of the night.
[30:32.140 --> 30:35.480]  Or you don't know when, you know, what's happening at other times that you may care about.
[30:36.200 --> 30:40.040]  And, you know, having a router do these, do this testing has a lot of benefits.
[30:40.280 --> 30:48.300]  So, uh, we could have a continuous sample of, you know, what performance you're getting and then show the ISP the chart saying, hey, this is what I'm actually getting.
[30:48.420 --> 30:49.720]  So we want to have that.
[30:49.860 --> 30:53.720]  That also helps with censorship because often censorship can be subtle.
[30:53.720 --> 30:59.920]  If you make a site slow, it's probably as good as censorship for a lot of purposes.
[31:00.240 --> 31:03.320]  So, um, we want to have, you know, include that as well.
[31:03.520 --> 31:05.060]  So those are the aspects for speed.
[31:05.280 --> 31:18.520]  And then coming to sharing where we actually started, uh, you know, people care about, uh, especially internationally, not exceeding a monthly bandwidth cap that their ISP might have.
[31:18.660 --> 31:22.640]  So they, they, they're happy to share the Internet, but they don't want to be in a position to pay more.
[31:22.640 --> 31:29.200]  And, uh, uh, and similarly, they, they, they care about their own, uh, Internet experience.
[31:29.580 --> 31:44.720]  And so what we, uh, are including in the interface to this router is a simple way, uh, for, uh, the hosts to set a, uh, bandwidth fraction that the guest users can get.
[31:44.720 --> 31:51.860]  And, um, also an easy way to monitor how much cumulative data has been used by the guest users.
[31:52.080 --> 31:57.420]  So if your cap is, you know, 300 gigabytes or something, you may say, okay, I'm happy to give away 50 gigabytes.
[31:57.780 --> 32:05.180]  And so we, we keep track of that for you and say, okay, now if that's exceeded, maybe you turn off your, um, uh, guest network.
[32:05.180 --> 32:08.460]  So moving to the next slide.
[32:08.720 --> 32:11.240]  So where are we with, with all these goals?
[32:11.980 --> 32:12.700]  Uh, go ahead.
[32:13.800 --> 32:15.220]  So here's the router.
[32:15.660 --> 32:21.520]  Uh, we, uh, decided we would hold back the details until, uh, uh, this conference.
[32:22.400 --> 32:25.540]  Primarily to allow all of you to buy this model.
[32:25.540 --> 32:30.240]  Uh, because, uh, it is an older model of router.
[32:30.480 --> 32:34.900]  It is a, a NETGEAR, uh, WNDR 3800.
[32:35.260 --> 32:37.120]  It's an 802-11N router.
[32:37.580 --> 32:40.340]  Uh, uh, it has been very popular, uh, among hackers.
[32:40.500 --> 32:45.180]  It's a very stable, you know, good piece of hardware, which supports completely open software.
[32:45.720 --> 32:53.540]  And, uh, uh, we chose that not only for those reasons, but that is also the router on which the Serovert project has built.
[32:53.540 --> 32:55.700]  And that is the only router that they support.
[32:55.960 --> 32:58.640]  So, uh, we, we have used that.
[32:59.220 --> 33:06.340]  And, uh, Serovert itself, as I mentioned, it's a, uh, what it is, is a derivative of open word focused on routing research.
[33:06.540 --> 33:10.520]  And they feedback their stuff into mainline, open word, and the kernel.
[33:11.040 --> 33:13.280]  And so we're using that as the software base.
[33:13.640 --> 33:20.600]  Having said that, I just want to say that this router we are launching with will not be the long-term router that we plan to support.
[33:20.700 --> 33:21.420]  It's an older router.
[33:21.420 --> 33:31.300]  There is much more powerful hardware that allows us to do many more interesting things, uh, that are actually available, but they were not quite available in the schedule that we wanted.
[33:31.680 --> 33:44.440]  So, we may move to, uh, something like a dual core, um, one gigahertz ARM processor based device in the future for a more widespread, uh, distribution of this, um, uh, product.
[33:44.440 --> 33:45.280]  So...
[33:46.160 --> 33:46.720]  So...
[33:46.720 --> 33:46.860]  Go ahead.
[33:47.580 --> 33:48.140]  Okay.
[33:48.880 --> 33:50.320]  So, actually...
[33:50.320 --> 33:54.280]  So, Adi, can I request you to do a demo since, uh, you're...
[33:54.280 --> 33:55.900]  I would be happy to.
[33:56.100 --> 33:56.380]  Okay.
[33:57.160 --> 33:58.820]  Let's make sure this works.
[33:59.860 --> 34:01.000]  Uh, so we're gonna do a demo.
[34:01.540 --> 34:04.420]  Um, uh, we're...
[34:04.420 --> 34:09.420]  We're gonna demonstrate the web UI for you guys because, um...
[34:13.950 --> 34:15.790]  It's so big, small.
[34:17.670 --> 34:18.150]  Great.
[34:18.670 --> 34:19.150]  Great.
[34:19.610 --> 34:21.170]  It's not mirrored, so this sucks.
[34:22.050 --> 34:22.530]  Um...
[34:22.530 --> 34:22.670]  Okay.
[34:22.910 --> 34:27.990]  So, um, um, we wanted to make this as easy as possible and as simple as possible.
[34:28.190 --> 34:29.410]  And, um...
[34:29.410 --> 34:30.670]  You can turn on mirroring if you want.
[34:30.830 --> 34:31.450]  I could.
[34:32.170 --> 34:33.290]  I'm not gonna do that.
[34:34.610 --> 34:35.090]  Um...
[34:35.990 --> 34:36.830]  Maybe I should.
[34:38.630 --> 34:39.530]  I think you should.
[34:42.070 --> 34:47.930]  So, what he's showing is the screen that you'll see when you first install the, uh, the firmware.
[34:48.470 --> 34:52.450]  And, uh, it'll, it'll basically ask you to set your administrator password.
[34:56.050 --> 34:56.530]  Um...
[34:56.530 --> 34:57.370]  It can be whatever you want.
[34:57.870 --> 34:59.870]  It's, uh, yeah, minimum eight character right now.
[35:02.330 --> 35:07.330]  And then it asks you to set the network SSID for your private Wi-Fi network.
[35:09.290 --> 35:17.950]  So, the, um, openwireless.org, you know, shared, uh, Wi-Fi network, uh, it has a fixed name as of now.
[35:18.050 --> 35:21.410]  We don't, we're not allowing that to be changed and that'll, you, you'll see that going forward.
[35:21.590 --> 35:22.410]  So, go ahead.
[35:22.610 --> 35:23.050]  Um...
[35:23.050 --> 35:26.870]  And then, uh, do you wanna explain why the passphrase is not...
[35:27.950 --> 35:28.750]  Not masked.
[35:28.750 --> 35:29.870]  Oh, okay.
[35:30.390 --> 35:38.730]  So, this, in, in the setup phase, we, we thought that it would be, uh, in a fairly private, um, environment.
[35:39.050 --> 35:46.350]  And that it would help people to, uh, in terms of remembering it or, or, you know, noting it down somewhere to have it actually be visible.
[35:46.730 --> 35:53.390]  Uh, when, in the future, you know, people actually log in using this administrative password, we do not show the password.
[35:55.870 --> 35:59.150]  Dice wearer rocks, by the way, if you don't know what that is.
[36:02.810 --> 36:03.930]  And, and that's it, right?
[36:04.030 --> 36:05.430]  Like, it, it, it's all set up.
[36:05.790 --> 36:08.110]  Um, and then this is the, this is the dashboard.
[36:08.410 --> 36:13.510]  So, um, um, right now it, it will show you, oh, hey, this is connected to the Internet.
[36:14.070 --> 36:17.210]  Um, uh, you have your private Wi-Fi network.
[36:17.210 --> 36:17.910]  We are at HOPE.
[36:18.310 --> 36:23.730]  Um, you have your open wireless network that you can toggle on and off, um, uh, if you want.
[36:23.950 --> 36:27.670]  Um, there's a setting down, oh, maybe it's in the other settings.
[36:28.190 --> 36:33.930]  Um, and, there are a bunch of settings.
[36:34.090 --> 36:35.210]  Ranga, do you want to walk through these?
[36:35.450 --> 36:35.890]  Yeah, yeah, sure.
[36:36.470 --> 36:40.030]  So, uh, first is, uh, you could change your password.
[36:40.390 --> 36:48.550]  Uh, the main, one of the main security features we, we decided would be useful here is, um, keep the web UI fairly simple.
[36:48.710 --> 36:52.410]  The number of things you can do, uh, are limited and for a purpose.
[36:52.730 --> 36:55.630]  Uh, and we believe that it's adequate for most people.
[36:55.930 --> 37:01.650]  Now, if you want to do more, um, then we, we decided it would be through SSH access.
[37:02.110 --> 37:07.530]  And, uh, for that, we allow people to upload their, uh, SSH public key.
[37:07.750 --> 37:13.330]  Uh, we did not want to allow password access because, again, you know, that's a, that's a potential attack.
[37:13.330 --> 37:19.990]  And so, uh, this, uh, actually you can click on that, you know, um, to, click SSH enable.
[37:22.710 --> 37:28.590]  So, basically, it brings up a window where you can enter your public key and, and, uh, and send it.
[37:28.830 --> 37:39.470]  And it will be stored, uh, and, uh, once you complete the login, uh, there, there is no further upload possible through the web interface.
[37:39.470 --> 37:43.970]  So, you have to SSH into the device to add more keys or change the key.
[37:44.210 --> 37:47.510]  Uh, and, and I think that makes it much, uh, more secure.
[37:47.910 --> 37:55.010]  Um, and, uh, the next, uh, uh, just, just a little bit further up.
[37:55.130 --> 37:57.290]  I mean, I just wanted to see, was there anything else there?
[37:57.530 --> 37:58.250]  I don't think so.
[37:58.430 --> 37:58.730]  Okay.
[37:59.270 --> 38:12.550]  So, uh, the ISP speeds there are things that, uh, we require users to enter to optimize the, uh, performance of the router, the queuing algorithms that are in there that we, uh, got from CeroWRT.
[38:12.790 --> 38:19.410]  Uh, and, uh, uh, you know, you could do this through speed test or, you know, maybe you have, uh, knowledge of it already.
[38:19.710 --> 38:21.970]  And so, you enter the upload and download speeds.
[38:22.370 --> 38:26.810]  Um, the private Wi-Fi network, you can, uh, you know, change the SSID.
[38:27.230 --> 38:30.110]  You can change, you know, change the usual Wi-Fi parameters over there.
[38:30.110 --> 38:41.830]  And, uh, as Nate mentioned, we plan to support, uh, uh, VPN Tor, uh, for all of the traffic, uh, on that private network, uh, as well as on the open wireless.
[38:42.050 --> 38:44.810]  Right now, that's grayed out because it hasn't been implemented yet.
[38:45.270 --> 38:54.150]  Uh, on the open wireless, the main feature, as I mentioned, is the ability to set the bandwidth fraction that you, uh, will limit the guest users to.
[38:54.150 --> 39:05.050]  Uh, uh, and then, uh, you can also set the maximum, uh, you know, megabytes or gigabytes of data that you're willing to allow guest users to use.
[39:05.530 --> 39:12.210]  And what happens is, if that number is exceeded, the open wireless interface automatically gets switched off.
[39:12.930 --> 39:20.310]  Uh, we're looking into having, you know, um, a more, uh, graceful mechanisms that maybe throttle it down and kind of stretch it out over the course of the month.
[39:20.630 --> 39:22.070]  But right now, it'll turn it off.
[39:22.070 --> 39:29.850]  And then, um, uh, the reset there is to reset the counter on that, uh, uh, you know, monthly cap.
[39:30.250 --> 39:38.750]  Uh, and then we will also sub, uh, support in the future, uh, you know, setting which day of the month you're using for that, uh, and have an auto reset.
[39:39.170 --> 39:44.010]  Uh, and then similarly, the, there are, you know, wireless features for that.
[39:44.170 --> 39:48.770]  And then there's one additional security feature, uh, for the open wireless.
[39:48.770 --> 40:04.350]  A lot of the criticism of open wireless in the media, which, you know, is somewhat misguided, but it's still a lot that is, uh, um, um, said out there about how using open wireless networks, uh, compromises your security because somebody could sniff the traffic.
[40:04.350 --> 40:14.210]  Now, the real solution for this is end-to-end security, uh, through HTTPS, uh, you know, or other solutions because that gives you true security through all the intermediate hops.
[40:14.210 --> 40:18.590]  It's not just about the question of getting security to the next hop, the Wi-Fi router.
[40:19.250 --> 40:24.530]  But since, you know, this is something people are concerned about, a lot of applications still stuff, send things in the clear.
[40:25.270 --> 40:33.130]  Um, uh, there's a mechanism to have secure open wireless without, um, having, um, authentication.
[40:33.150 --> 40:36.990]  And through a mechanism called EAPTLS, and we will be supporting that as well.
[40:40.800 --> 40:41.600]  And that's it.
[40:45.580 --> 40:48.920]  So, in summary, uh, you know, this is a hacker release.
[40:49.080 --> 40:52.660]  That's why we're, uh, here at HOPE, uh, to release it.
[40:52.800 --> 40:56.860]  Uh, it's, uh, and in my, um, um, terminology, that's pre-alpha.
[40:57.120 --> 40:59.940]  And the purpose of the release is really to draw in developers.
[41:00.300 --> 41:03.100]  Uh, and we have made the project very developer-friendly.
[41:03.280 --> 41:11.740]  And I, I have to credit my colleague, uh, uh, Jacob here for, for really, uh, making the project very usable, uh, uh, to new developers.
[41:11.960 --> 41:14.760]  Uh, and, uh, and that's our goal, really, to be here at the conference.
[41:14.760 --> 41:18.280]  To bring you, uh, uh, in and get you involved.
[41:18.540 --> 41:23.040]  Uh, and we can now really begin the process of, uh, hardening the security on the device.
[41:23.680 --> 41:26.960]  And, uh, as I mentioned, the hardware platform will change.
[41:27.100 --> 41:29.520]  So, you know, it is not an expensive platform.
[41:29.620 --> 41:30.400]  I encourage you to buy it.
[41:30.480 --> 41:31.880]  It's about $50 refurbished.
[41:32.300 --> 41:34.060]  Units are available on Newegg.
[41:34.440 --> 41:36.280]  Uh, but, you know, it may change in the future.
[41:36.860 --> 41:36.960]  So.
[41:37.640 --> 41:38.740]  And how can you help?
[41:38.980 --> 41:41.320]  Uh, so, you know, since security, security is so important.
[41:41.540 --> 41:46.160]  If you have expertise in security, uh, you can pen test, audit, uh, the software.
[41:46.360 --> 41:47.440]  Uh, it is on GitHub.
[41:47.900 --> 41:51.480]  Uh, it will be, uh, the, the location will be announced on the EFF website.
[41:51.520 --> 41:57.340]  Or, uh, you can go to the download URL that I, I, uh, pointed out earlier.
[41:57.880 --> 41:58.940]  You can extend it.
[41:59.040 --> 42:00.100]  The web UI is minimal.
[42:00.200 --> 42:01.320]  You may want to do something different.
[42:01.460 --> 42:02.780]  There may be a feature that's missing.
[42:03.060 --> 42:06.120]  Uh, please, uh, uh, suggest it, add it.
[42:06.260 --> 42:09.420]  Uh, and, and we'll be accepting pull requests on the GitHub.
[42:09.420 --> 42:15.060]  Uh, you can, uh, there are features that can be added to OpenWRT, the, the base that we're using itself.
[42:15.940 --> 42:22.220]  Two projects that I want to mention that'll really help is people want to know where these open wireless access points are.
[42:22.360 --> 42:26.200]  So you, you, especially where it's most useful is you go to another city and you want to connect.
[42:26.280 --> 42:27.240]  So you want to know where they are.
[42:27.240 --> 42:32.380]  And some of the other systems, you know, you might have heard of Fawn or, uh, you know, Comcast Xfinity.
[42:32.620 --> 42:34.440]  They try to provide this kind of map.
[42:34.880 --> 42:45.100]  If one of you is interested in trying to create a map and, uh, creating a system where people can easily find out where to get these, uh, open wireless access points, you know, please come talk to me.
[42:45.100 --> 42:46.400]  I think that would be an awesome project.
[42:47.120 --> 42:54.700]  Uh, another is an open wireless app, uh, uh, instead of, you know, just a web UI that we have today.
[42:54.880 --> 42:59.520]  And there's some additional cool things that can be done if we have apps on Android and, and iOS.
[42:59.520 --> 43:02.180]  And if you're interested in that also, please, please come talk to me.
[43:02.380 --> 43:20.460]  And then in terms of deployment, you know, that's where we can really use a lot of help is, is, uh, you know, organizing maybe neighborhoods where a large number of people deploy open wireless is a great way, uh, to really, uh, make it useful, uh, rather than it just be scattered over a large area.
[43:20.820 --> 43:23.440]  So neighborhoods or your city or town wants to do it.
[43:23.540 --> 43:27.040]  Please come talk to us and we'll, we'll try to see what we can do to help make it happen.
[43:27.760 --> 43:28.140]  Great.
[43:29.840 --> 43:31.240]  Um, and, and that's that.
[43:31.360 --> 43:34.520]  That's, that's the project that we wanted to, uh, announce and talk about.
[43:34.680 --> 43:37.460]  And we're, we'd love to open it up to questions if anybody has any.
[43:37.620 --> 43:39.680]  I think we have 11 minutes for questions.
[43:39.720 --> 43:40.400]  Yeah, we have 11 minutes.
[43:40.580 --> 43:40.880]  Thank you.
[43:46.720 --> 43:52.360]  So, uh, yes, line up behind that, uh, that halo of a silhouette that I can't see anything.
[43:52.720 --> 43:53.120]  Hi.
[43:54.000 --> 43:54.400]  Hello.
[43:54.560 --> 43:54.720]  Hello.
[43:55.120 --> 43:58.260]  Are you using one radio or two radio in that, radios in that AP?
[43:58.880 --> 43:59.240]  Directo.
[43:59.820 --> 44:01.320]  And you allow them to be on the same band?
[44:02.900 --> 44:05.340]  So there's a five gigahertz radio and there's a two gigahertz.
[44:05.400 --> 44:05.540]  Right.
[44:05.700 --> 44:05.800]  Okay.
[44:05.920 --> 44:06.880]  So you have to have them separate.
[44:07.060 --> 44:07.180]  Yeah.
[44:07.820 --> 44:10.480]  Um, is there any reason why you're not running multiple SSID?
[44:12.160 --> 44:13.560]  Uh, I didn't understand the question.
[44:13.680 --> 44:14.600]  Multiple SSID meaning?
[44:14.600 --> 44:19.300]  You can run more than one SID, compartment them out into different lands.
[44:20.680 --> 44:22.800]  So we are, we have multiple SSIDs, right?
[44:22.920 --> 44:24.320]  There's an open wireless.org SSID.
[44:24.700 --> 44:26.500]  Yeah, I mean, I know it's an open WRT.
[44:26.620 --> 44:28.160]  I wrote the patent on it in 1999.
[44:29.680 --> 44:35.300]  So, uh, I think part of the answer is we do allow them to be on the same band, if that's what we're wrong.
[44:35.400 --> 44:36.460]  But not both radios.
[44:36.680 --> 44:38.560]  Both SIDs on the same radio is fine.
[44:38.880 --> 44:40.740]  Both radios on the same band, bad.
[44:41.280 --> 44:43.320]  And if I have to explain that to you, you're a non-starter.
[44:45.740 --> 44:51.200]  Um, secondly, is there any chance you're going to use, like, DNSSEC to sign that SSH key?
[44:51.460 --> 44:57.000]  So, you don't have that, you know, I've never seen SSH key before.
[44:57.240 --> 45:00.320]  If it has to change, you don't get the security warning.
[45:00.880 --> 45:03.160]  I mean, everybody in this room can deal with that.
[45:03.340 --> 45:04.020]  They know what they're doing.
[45:04.280 --> 45:07.880]  But if you start putting this out to homeowners, they're not going to know what that means.
[45:08.600 --> 45:08.960]  Right?
[45:09.180 --> 45:13.120]  I changed my SSH key and all of a sudden I can't log in anymore because I get the security warning.
[45:13.320 --> 45:14.920]  They're not supposed to use SSH key.
[45:15.160 --> 45:18.180]  But if you DNSSEC sign it, it works.
[45:18.580 --> 45:22.020]  For the consumer release, there will be a way of bumping.
[45:22.340 --> 45:24.220]  That's Peter Eckersley, our technology director.
[45:24.420 --> 45:24.600]  Okay.
[45:25.100 --> 45:26.620]  I mean, I like the idea.
[45:26.820 --> 45:27.820]  It's a great legal hack.
[45:28.040 --> 45:29.420]  I don't see a lot of technology here.
[45:29.580 --> 45:30.800]  But I understand the legal hack.
[45:31.860 --> 45:34.600]  So, we will be able to support DNSSEC on here.
[45:35.000 --> 45:38.320]  Because several, you know, one of the features that they have...
[45:38.320 --> 45:38.620]  Dave Tate.
[45:38.820 --> 45:38.920]  Yay.
[45:39.320 --> 45:39.400]  Right?
[45:42.200 --> 45:44.040]  You might look at doing this with FreeBSD.
[45:44.860 --> 45:45.220]  Thanks.
[45:45.640 --> 45:47.680]  If you want to help us do that, that would be great.
[45:47.860 --> 45:49.740]  You know, this is the best platform we could find right now.
[45:53.250 --> 45:53.610]  Hi.
[45:53.790 --> 46:03.470]  I wondered if you could speak to the issue of terms of service with the major ISPs not allowing you to share bandwidth.
[46:04.790 --> 46:08.830]  Well, I'm not going to give legal advice to you.
[46:09.650 --> 46:10.130]  Of course not.
[46:10.130 --> 46:12.290]  A lot of...
[46:13.430 --> 46:14.670]  I'll save that one in reserve.
[46:15.030 --> 46:15.150]  Okay.
[46:15.210 --> 46:16.290]  We can talk about it offline.
[46:18.990 --> 46:19.470]  Hey.
[46:19.810 --> 46:26.010]  So, the main point is to obfuscate multiple clients via having them all share the same IP address.
[46:26.710 --> 46:28.270]  If I understand this correctly.
[46:28.470 --> 46:29.430]  One of them.
[46:29.650 --> 46:29.850]  One of them.
[46:30.530 --> 46:35.590]  So, the EFF has done really nice research in the past on something called Panopticlic.
[46:35.990 --> 46:44.990]  Which is a project that identifies certain actors on the Internet via fingerprinting their web browser client in various ways.
[46:44.990 --> 46:57.010]  And so, your own research is actually pointing towards another vector in which a user can be identified even if they are sharing the same wireless connection and so on.
[46:57.370 --> 47:01.810]  And I'm not saying this in order to take away from this very promising project.
[47:01.810 --> 47:03.570]  But I think it would be interesting.
[47:03.890 --> 47:21.730]  Have you considered adding a proxy or some sort of DPI thing on the router level that would do some work to prevent the sort of Panopticlic style identification of clients?
[47:23.470 --> 47:25.750]  There are several problems with that approach.
[47:25.750 --> 47:38.270]  One of them is the wiretap act that make it not legally very good if the person connecting to the network doesn't know and doesn't consent to that approach.
[47:40.030 --> 47:42.050]  It's not one of the things we are thinking about right now.
[47:42.790 --> 47:43.410]  But one of the...
[47:43.410 --> 47:53.710]  So, Nadeem raises the point of Panopticlic which is that many, if not most, desktop web browsers are uniquely identifiable or damn close to it.
[47:53.710 --> 47:58.890]  One of the issues with Panopticlic though is something like this is not particularly identifiable.
[48:00.010 --> 48:04.170]  And that mobile is one of the use cases for open wireless.
[48:04.750 --> 48:16.890]  So, yeah, even though you can fingerprint a desktop browser with quite a bit of... with many bits of entropy, fingerprinting an iOS or Android device is a lot harder.
[48:18.270 --> 48:22.310]  Another part of the answer is that defending against fingerprinting is really hard.
[48:22.310 --> 48:24.230]  It's not impossible maybe.
[48:24.470 --> 48:25.110]  We're going to try it.
[48:25.110 --> 48:27.410]  But the better place to be trying to do it is inside the client.
[48:27.910 --> 48:34.090]  You have a better chance with a browser extension or browser patches actually against the, you would call, browser code.
[48:34.210 --> 48:43.770]  And you're using on the network effectively man-in-the-middle attacking some subset of the traffic and trying to hide the revealing nature of the endpoint.
[48:44.330 --> 48:51.150]  And one of the points of the open wireless project is that, as Peter raises, we're not man-in-the-middling your connection.
[48:52.030 --> 49:00.950]  Open wireless as it exists right now at cafes, at the Hotel Pennsylvania, at the hotel that we're staying at, man-in-the-middle is your connection, right?
[49:01.070 --> 49:03.350]  When you first... that's the captive portal problem.
[49:03.550 --> 49:06.830]  The captive portal problem is really bad and breaks a lot of stuff.
[49:07.090 --> 49:13.330]  It trains users to click, you know, this certificate is not... it's not correct.
[49:13.970 --> 49:24.330]  And we don't want... the open wireless project will not man-in-the-middle connections, either in a captive portal situation or DPI to try and get rid of the browser fingerprint.
[49:25.050 --> 49:27.170]  It's one of the policy positions that we've taken here.
[49:30.170 --> 49:43.710]  My question is, if I'm a user of, you know, this and I get a DMC notice about something I know I did not download and it was through my open thing, you say, oh, there's a legal precedent for this.
[49:43.830 --> 49:46.790]  So what's my next step?
[49:47.050 --> 49:51.130]  And is that... I'm not asking a personal legal advice question.
[49:51.210 --> 49:52.130]  I'm asking a general question.
[49:52.530 --> 49:56.050]  And is that practical for regular users?
[49:56.190 --> 49:59.750]  So if you say, oh, write a letter to your... demand letter to your ISP.
[49:59.910 --> 50:02.170]  I mean, like, is this something a regular user can actually do?
[50:02.830 --> 50:04.310]  Or is this something you have to call a lawyer for?
[50:05.130 --> 50:07.230]  It's probably something you have to call a lawyer for.
[50:07.230 --> 50:14.630]  Whenever you get a lawsuit or a threat of a lawsuit, calling a lawyer is one of the steps that I, as a lawyer, would probably recommend you do.
[50:14.890 --> 50:16.930]  So you should have a lawyer on retainer if you're using...
[50:16.930 --> 50:17.170]  No!
[50:17.870 --> 50:21.310]  I mean, should all of you have a lawyer on retainer?
[50:21.390 --> 50:22.710]  Yeah, it's probably a good idea, right?
[50:22.950 --> 50:25.630]  Just in general, whether you're using open wireless or not.
[50:26.310 --> 50:27.870]  I'm just meaning it increases the cost.
[50:28.070 --> 50:29.270]  Full employment for lawyers.
[50:29.750 --> 50:30.150]  Significantly.
[50:32.330 --> 50:37.110]  But there are... EFF maintains a list of lawyers who do this sort of work.
[50:37.110 --> 50:43.730]  So, you know, info at EFF.org is a great email address that you should have at the tip of your fingers.
[50:43.930 --> 50:51.450]  Do you anticipate having, like, some form things that people can be like, if you get a DMCA, here's the open wireless, you know, form letter you can send your ISP.
[50:51.590 --> 50:52.950]  That's something that we might consider.
[50:52.950 --> 50:53.170]  Yeah.
[50:53.170 --> 50:53.450]  Okay.
[50:54.770 --> 50:55.650]  There's also...
[50:56.210 --> 50:58.050]  I'm not a lawyer, but I'm going to say this.
[50:58.110 --> 51:12.990]  There's a certain kind of case that EFF is always looking to hear about, which is, if you are really not a file sharer, you know, or you've never downloaded anything, like, law and network history is super clean, you're being harassed because you're running from the wireless that work with DMCA.
[51:13.710 --> 51:15.250]  Email that email address, please.
[51:17.670 --> 51:18.330]  Excuse me.
[51:18.410 --> 51:19.610]  Yeah, I'm...
[51:21.370 --> 51:23.590]  My question is a little tangential to that.
[51:23.590 --> 51:26.770]  Maybe tech has already answered it, and I...
[51:26.770 --> 51:28.850]  It was too condensed for me or something.
[51:29.650 --> 51:36.730]  But, chiefly, when I have a user, or, you know, I guess, I don't typically...
[51:37.430 --> 51:43.430]  You know, a few people are wicked, but very many people are gullible or heedless.
[51:43.570 --> 51:57.410]  They click on a link that I wouldn't touch with a barge pole, and I end up infected with a rootkit or viruses or all kinds of vermin, and I, you know, and I, you know, and...
[51:57.410 --> 52:02.730]  You know, the guy will have, you know, you know, he would be perfectly innocent.
[52:03.170 --> 52:10.370]  He won't have been trying to do anything, you know, shady or anything like that.
[52:10.490 --> 52:15.890]  But I end up with my computer in jeopardy.
[52:15.890 --> 52:21.150]  And, you know, and I wonder, and I wonder about situations like that.
[52:21.310 --> 52:25.890]  How you can separate partition, you know, uses...
[52:27.950 --> 52:32.370]  Uses that might be hurtful that are...
[52:32.370 --> 52:38.160]  Have no malicious or, you know, shadowy intent from the user.
[52:38.910 --> 52:44.470]  He just clicked on something that, you know, that he shouldn't have.
[52:44.470 --> 52:46.170]  And I end up with the debris.
[52:46.910 --> 52:49.470]  You want to talk about how the networks are partitioned?
[52:49.570 --> 52:49.950]  Yeah.
[52:50.630 --> 53:07.050]  So, you know, one of the reasons to provide the VPN option or the TAR option for both the private and the open one is if you choose, let's say, to have a VPN on the open one, all your traffic would be completely separate from the guest traffic.
[53:07.230 --> 53:15.470]  So if he was doing something, you know, that could potentially be problematic with law enforcement, they would see that, okay, that came from a different address.
[53:15.630 --> 53:18.090]  I'm talking about the integrity of my CPU.
[53:18.710 --> 53:23.130]  He gets this message and it contains a virus or something like that.
[53:23.330 --> 53:30.090]  So there shouldn't be any reason that traffic from the open wireless is even getting into your network, right?
[53:30.290 --> 53:31.710]  So they're separated.
[53:31.710 --> 53:38.890]  There is no way that the guest users are able to see devices in your home.
[53:39.370 --> 53:40.630]  No, no, I'm not...
[53:41.550 --> 53:43.710]  I'm not talking about privacy.
[53:43.970 --> 53:48.030]  I'm talking about, like, this guy downloaded something.
[53:49.110 --> 53:51.550]  That'll affect his computer but not yours.
[53:52.050 --> 53:54.150]  It'll affect his machine, not your machine.
[53:55.150 --> 53:56.710]  It's going to pass through the router.
[53:56.850 --> 53:57.970]  The router is not going to be affected.
[53:58.130 --> 53:58.870]  They're not going to be affected.
[53:59.190 --> 54:00.250]  I mean, they're going to be affected.
[54:00.390 --> 54:01.590]  You're never going to see it.
[54:02.770 --> 54:04.550]  We have time for one more question.
[54:04.610 --> 54:09.270]  But I want to hear what you guys have to say and come find us afterwards.
[54:10.190 --> 54:11.550]  Okay, I'll try to make it a good one then.
[54:12.130 --> 54:14.710]  I just want to know, so this is a Tor client.
[54:15.030 --> 54:22.050]  Have you considered adding an option for Tor transit or exit nodes to increase the Tor network bandwidth total?
[54:23.250 --> 54:25.070]  So it would be very easy to do that.
[54:25.410 --> 54:26.930]  And, you know, we...
[54:26.930 --> 54:31.130]  From the SSH access, you can do that because the Tor binary is in there.
[54:31.250 --> 54:32.930]  You can set it up to do that.
[54:33.090 --> 54:39.150]  We won't have it set up like that by default because it uses up people's bandwidth and we want people to make the choice to do that.
[54:39.290 --> 54:40.390]  And we will make it easier for them.
[54:41.090 --> 54:43.990]  And in terms of exit node, we're probably not going to have that as an option.
[54:45.150 --> 54:50.930]  Running a Tor exit node, the Tor project benefits from...
[54:50.930 --> 55:00.350]  From large, low latency, high bandwidth exit nodes running on real hardware, not Soho router on a DSL.
[55:01.090 --> 55:06.150]  A whole bunch of really small, really slow Tor exit nodes is the last thing that Tor needs.
[55:07.710 --> 55:10.710]  It might be good for privacy, but it doesn't work with the spec.
[55:10.710 --> 55:14.210]  So, you know, but it is probably good as a bridge.
[55:14.390 --> 55:24.130]  So, you know, if you are traveling remote, you know, outside the country and you want to first hop off of your own home network and then, you know, access Tor, you know, it's useful for those kinds of things.
[55:25.190 --> 55:29.350]  So, as, as Adi said, unfortunately, this is the last question.
[55:29.430 --> 55:30.810]  That was the last question that we can take.
[55:31.570 --> 55:41.310]  But in terms of coming and finding us after, EFF is having a speakeasy at a bar called American Whiskey, which is like a block and a half away, starting in four minutes.
[55:41.650 --> 55:42.070]  Four minutes.
[55:42.850 --> 55:43.930]  And we will be there.
[55:44.370 --> 55:45.870]  Drinks are not on us.
[55:45.990 --> 55:46.630]  They are on you.
[55:46.630 --> 55:50.350]  So, bring a couple of bucks and let's have a beer.
[55:51.630 --> 55:52.390]  Thank you, guys.