FreedomBox 1.0
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
| Title |
| |
| Title of Series | ||
| Number of Parts | 90 | |
| Author | ||
| License | CC Attribution 2.0 Belgium: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor. | |
| Identifiers | 10.5446/40289 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language |
Content Metadata
| Subject Area | ||
| Genre | ||
| Abstract |
|
FOSDEM 201352 / 90
2
5
8
10
12
13
14
15
17
19
21
24
25
28
29
31
32
34
36
39
40
43
44
46
50
51
52
54
55
57
58
62
65
66
67
78
79
87
88
00:00
Software developerOpen sourceOpen setSoftwareCuboidSource codeXMLLecture/Conference
01:59
Server (computing)Online helpPower (physics)Graph (mathematics)Arithmetic meanProxy serverWordInformation privacyElectronic mailing listLecture/Conference
02:47
Graph (mathematics)State of matterExistenceDependent and independent variablesPower (physics)Goodness of fitMereologySoftwareForm (programming)Multiplication signMoving averageWeightData miningBitLecture/Conference
05:03
FacebookSoftwarePower (physics)CuboidRoutingReal numberComputer architectureInformation privacyComputer hardwareProgramming paradigmPoint cloudNumberMultiplication signStallman, RichardGraph (mathematics)Client (computing)FreewareSpacetimeServer (computing)WeightPoint (geometry)Virtueller ServerMereologyQuicksortPhysical lawSystem callBitExtension (kinesiology)EmailLecture/Conference
10:30
Internet der DingeSeries (mathematics)Computer hardwareQuicksortMultiplication signNeighbourhood (graph theory)Metropolitan area networkHecke operatorCuboidAsynchronous Transfer ModeSolid geometryWordCircleAddress spaceVirtual machineCue sportsAdditionRight angleSoftwareINTEGRALSingle-precision floating-point formatLecture/Conference
15:43
Video gameMultiplication signChaos (cosmogony)Arithmetic progressionImplementationRepresentation (politics)Projective planeSoftwareCuboidFunctional (mathematics)Cartesian coordinate systemMusical ensembleLecture/Conference
16:55
Key (cryptography)Data modelQuicksortEvent horizonBitFundamental theorem of algebraKey (cryptography)MereologyDiscrete element methodDevice driverInformationSocial classEndliche ModelltheorieElectronic signatureMultiplication signSoftwareComputer hardwareFirmwareOpen setBinary codeComputer animation
18:01
Data modelKey (cryptography)Different (Kate Ryan album)Form (programming)Cartesian coordinate systemEncryptionQuicksortValidity (statistics)BitIdentity managementWeb 2.0SoftwareKey (cryptography)Endliche ModelltheorieComputer animationLecture/ConferenceMeeting/Interview
18:54
Device driverFirmwareComputer hardwareServer (computing)Binary fileComputer hardwareCuboidMultiplication signProjective planeSoftwareLatent heatCommitment schemePrototypeAxiom of choiceMereologyQuicksortSoftware developerDecision theorySet (mathematics)Right angleComputer animation
19:58
Data modelMobile appContent (media)Electronic signatureForm (programming)Cross-correlationKey (cryptography)Identity managementAuthenticationQuicksortEndliche ModelltheorieKey (cryptography)Computer hardwareReal numberTheory of relativityCoroutineFreewareCuboidWeb 2.0Software developerElement (mathematics)Open setCross-correlationMobile appLecture/ConferenceComputer animation
21:19
Kernel (computing)Public key certificateInteractive televisionProduct (business)Form (programming)Server (computing)Multiplication signElectronic signatureBitFunctional (mathematics)Web browserMathematicsContent (media)Connected spaceInformationQuicksortFrustrationSocial classAreaMoment (mathematics)Web 2.0Process (computing)Module (mathematics)Standard deviationAuthenticationKey (cryptography)Set (mathematics)AlgorithmEncryptionPatch (Unix)CryptographyCountingSphereLecture/Conference
23:41
Computer hardwareComputer hardwareCuboidInformationContext awarenessOpen setQuicksortConfiguration spaceProjective planeScaling (geometry)CodecComputer animation
24:22
CuboidQuicksortFilm editingComputer hardwareBitProjective planeLecture/Conference
25:10
BootingKernel (computing)File systemNetwork socketInterface (computing)Server (computing)Power (physics)CuboidComputer hardwareInternet service providerOrder (biology)Binary codeInterface (computing)System on a chipVideo gameAsynchronous Transfer ModeCoprocessorWhiteboardComputer animationLecture/Conference
25:59
Derivation (linguistics)Default (computer science)ComputerServer (computing)Router (computing)Computing platformInterface (computing)Computer networkServer (computing)Interface (computing)SoftwareWeb serviceCuboidFiber bundleComputer hardwarePlug-in (computing)Derivation (linguistics)ImplementationConfiguration spaceProduct (business)Default (computer science)Data storage deviceContext awarenessWhiteboardTheory of relativityComputer animation
26:57
BootingData storage deviceInterface (computing)ExpressionProcess capability indexPlastikkarteMobile WebLaptopQuadrilateralCore dumpSynchronous dynamic random-access memoryBinary fileNetwork topologyComputer hardwareQuicksortProjective planeOpen setWhiteboardRing (mathematics)MereologySinc functionCuboidReverse engineeringLecture/ConferenceComputer animation
27:40
Web serviceComputer hardwareCodeGraphics processing unitGoodness of fitBinary codeWhiteboardOpen sourceOpen setComputing platformPrototypeExistenceProcess (computing)SubsetFormal languageSocial classCodeQuicksortProfil (magazine)CoprocessorMultiplication signCuboidPoint (geometry)LaptopDirect numerical simulationDirection (geometry)Programming paradigmData conversionArithmetic progressionMoment (mathematics)Projective planeStack (abstract data type)ImplementationEmailWeb 2.0Device driverServer (computing)WordLecture/Conference
30:22
Computer hardwareData modelObject (grammar)Open setCuboidGroup actionSpeichermodellOrder (biology)User interfaceComputer hardwareEndliche ModelltheorieSoftwareMereologyArmServer (computing)Configuration spaceSet (mathematics)Text editorComputer fileCoprocessorWeightPower (physics)BuildingRight angleFreewareJSONXMLComputer animationLecture/Conference
32:18
Computer hardwareUser interfaceContent (media)User interfaceScripting languageMenu (computing)Right angleFunctional (mathematics)Group actionCuboidProjective planeData structureUsabilityElement (mathematics)Server (computing)SoftwareHeegaard splittingMedical imagingBitEncapsulation (object-oriented programming)Computer fileInformationPhysical systemConfiguration spaceState of matterMeta elementPlastikkarteCore dumpSet (mathematics)Key (cryptography)Connected spaceSphereOpen setComputer animationLecture/Conference
33:49
User interfaceForm (programming)Flow separationAdditionMultiplication signServer (computing)Front and back endsUser interfaceWeb 2.0Data structureForm (programming)Cartesian coordinate systemCuboidDebuggerTotal S.A.Medical imagingQuicksortBuildingFlow separationDirectory serviceLibrary (computing)PiSet (mathematics)System administratorWeb serviceInformationInclusion mapComputer animationLecture/Conference
35:54
Computer hardwareBuildingComputer-generated imageryIdentity managementClient (computing)Virtuelles privates NetzwerkKey (cryptography)Service (economics)Client (computing)Identity managementComputing platformMultiplication signMeta elementTouch typingFrictionLinearizationQuicksortChainInformationOpen setComputer hardwareModal logicServer (computing)MereologySphereThresholding (image processing)Electronic mailing listCuboidAdditionCartesian coordinate systemBranch (computer science)NumberInstance (computer science)WikiOrder (biology)Web 2.0Utility softwareProcess (computing)Case moddingWeb browserInclusion mapPoint (geometry)SoftwareConnected spacePublic key certificateView (database)Different (Kate Ryan album)Computer animation
39:18
Server (computing)Set (mathematics)Information securitySoftwareInformation privacyWeb 2.0InternetworkingDatabase transactionRule of inferenceRight angleClient (computing)RandomizationPoint (geometry)Instance (computer science)Lecture/Conference
40:05
Information privacyModul <Datentyp>Online helpMachine visionMachine visionContent (media)WebsiteQuicksortOnline helpAdditionUser interfaceThresholding (image processing)BitMultiplication signServer (computing)Medical imagingSoftwareKey (cryptography)Configuration spaceInformation privacyWeb 2.0Scripting languageMereologyWikiFunctional (mathematics)Group actionData structureComputer animation
41:34
Self-organizationWordMultiplication signClosed setFreewareSoftwarePhysical systemCuboidComputerTrailCartesian coordinate systemOpen setComputer hardwareKey (cryptography)Projective planeBitTelecommunicationCountingOrder (biology)HookingUniverse (mathematics)Moment (mathematics)QuicksortComputerCodeStack (abstract data type)MereologyWeb 2.0RoutingServer (computing)Connectivity (graph theory)BuildingStreaming mediaPower (physics)Fraction (mathematics)AuthenticationInformation privacyProxy serverRoboticsPoint (geometry)CAN busArithmetic meanLecture/ConferenceXMLComputer animationMeeting/Interview
47:56
Multiplication signState of matterQuicksortCommunications protocolDistribution (mathematics)SoftwareTouch typingLikelihood functionMathematicsNoise (electronics)CuboidExtension (kinesiology)Repository (publishing)Drop (liquid)Scripting languageCivil engineeringMereologyContext awarenessSound effectPhysical systemMachine visionXMLComputer animationLecture/Conference
Transcript: English(auto-generated)
00:01
OK, please take your seats. And the next two speakers, the first of this afternoon, are two great software developers, and open source developers, actually. And they have a great experience in open source that I don't think I have to introduce themselves. And they are going to talk about FreedomBox.
00:21
Thank you. Good afternoon. It's a pleasure to be here always, and an honor always. And it's a particular pleasure and honor for me to be here with Bedale.
00:41
I thought we might have to go on alternating years for a little while longer. But I think we're close enough that we can actually be here productively in the same place. Two years ago, I wanted to say we should do this.
01:00
A year ago, Bedale said, here's how we're going to do this. I would love to be able to say, we did this. Now we can retire and do other things. Because there are a lot of other things that need to be done, also. We can't say that. What we can say is, we're getting ready to use.
01:22
I think that's really important, because in the two years since I came here to say we should do this, every government on Earth has started doing all the things we were afraid of. There were governments on Earth that were doing things we were afraid of two years ago, in fact, four, when I started wondering whether we
01:42
would be able to do this. But now we are living in a world that we foresaw, and which is happening faster than we thought it would. Between the beginning of 2011 and now, the United States government decided that it wanted, as one White House official put it to me
02:04
last March, we've decided we need a robust social graph of the United States. That means the most powerful government on Earth keeping a list of everybody every American knows. And as you are all aware, what the United States
02:24
government will do at home, it will, of course, do everywhere else in the world. And it will help lots of other people to do, too, as long as they all share in the very limited meaning that intelligence services use when they use the word share.
02:40
The democracies are very complacent about extinguishing the privacy and freedom in the net, because the democracies believe that they are just. And they think that they should be trusted. And they think that the infrastructure of despotism
03:00
is OK to make, because we're trusting them with it and they're going to be fine. But a network is now coming into existence around the world which is capable of making despotism impregnable. And that means that everything that is happening to the net, because these guys think
03:23
that they need a robust social graph of the United States in the interests of good and trust us with it, the infrastructure we are laying down is a time bomb for every human being. Because anywhere where despotism comes with this on its side, it will stay.
03:46
And darkness will be growing bit by bit, however long it takes. And it won't roll back once it arrives. So what we need to do in response to the way
04:00
the states are taking advantage of the commercial surveillance of the net that commerce is building, that's the form in which it happens, right? Business finds a way to do something to data mine customers, and governments take advantage of it. And once again, you can't slow down the pace of innovation
04:23
very well. And so business goes on innovating, and power catches up behind. And if it's power you can trust, then you empower it today. But what happens next is sooner or later, darkness comes and it never rolls back again.
04:40
And we have enough friends and colleagues making freedom or trying to make freedom in parts of the world that are already pretty dark that we know what we're talking about when we talk about that. When we were together here in 2011 saying we ought to do this, we were already aware that we had friends and colleagues
05:00
in the street around the world who were in danger because they were going to demonstrations organized on Facebook carrying iPhones. In Bahrain, the darkness hasn't let up yet. And it never will, maybe, if everything about the network makes all that listening and all that predicting
05:21
possible. So Freedom Box is really how we rebuild the net to extend Gilmore's law just a little bit. John Gilmore and I have been working together since we were teenagers. And when he said censorship is regarded by the net as malfunction and the net routes around it,
05:44
he was speaking the dream of our childhood, the one we have to make real right now or else. The net has to regard despotism as malfunction and route around it. And the only way that can happen is if we not only make routing smart and give privacy
06:04
tools to people, but we also make it possible for them to use those tools in their real lives. Freedom Box says the world's going to fill up with cheap, small hardware that is wonderful. And some of it is going to be totally free hardware
06:22
that we have been looking for for decades. And we're going to use that hardware to give people the ability to provide services to one another in federated ways that secure communications and enable sharing that doesn't watch you share for the benefit of whoever happens
06:42
to have power where you are. Freedom Box says we can do all that and Freedom Box says we have to. This marks for me personally 20 years of working with Richard Stallman. And I know what those 20 years were spent trying to do,
07:03
which was to make sure that freedom came to software and that software came to everywhere with freedom in it so that when we had free software everywhere we could turn freedom on.
07:20
At the time that I started doing this work, making freedom come to software meant dealing with companies that wanted all software to be unfree, particularly one. That effort failed. We got what we needed. We got free software into everything
07:42
that dominated the server client paradigm in which the Microsoft monopoly was forged. Now we are working as hard as we can to make sure free software is in everything in the cloud to mobile architecture, which is replacing it.
08:00
But when they talk about the cloud, they mean virtualized servers. When we talk about the cloud, we mean a large, large number, hundreds of millions and then billions of devices running free software whose purpose is to keep the net federated rather than centralized so that power
08:23
that becomes despotic doesn't become immortal. I think it's important to go over that much part of why now because what we wanna say today is now we are beginning to have software ready
08:41
that we have to start teaching people to use. Not because we're done. We're not even really altogether ready to say we're done beginning. But because from the beginning, we have to be helping people to adopt.
09:00
So the space between now and what I think of as 1.0, which is getting really narrow, is the space in which it's time to start helping people become users. All over the world, people are beginning to understand the problem who aren't geeks like us.
09:22
2012 was not only the year the United States government decided it needed a robust social graph of the United States, and the Indian government said it wanted to listen to every telephone call and SMS and email message in India as soon as possible. And everybody sort of agreed
09:41
on how we're all gonna talk about privacy law together so it doesn't really hurt Google, which is the real subject of the front of the New York Times today. The last part of our time, 2012, while we were trying to get to 1.0
10:00
was also the year that all around the world, ordinary people who don't lead geeky kinds of lives began to understand why they needed some geek to come and rescue them. Because everything starts to get looking pretty sketchy, even to people who don't follow technology, they just use it in their lives or are used by it.
10:21
And 2012 was the year that people around the world who don't do software for a living suddenly began to understand that the net was using them at least as much as they were using it. And that's gonna be a powerful political subject around the world for particularly younger net focused people, no matter whether they live
10:43
in democratic or undemocratic societies. And as they get there and try and use politics to preserve freedom for themselves, they're going to discover that politics won't do it for them. Politics is too slow, politics is too corrupt,
11:01
politics is too incapable of helping them to get what they need in their daily lives right away. Politics will talk about it for 20 years and then do something. We know that, that's why the Free Software Movement was not a movement of lawyers, it was a movement of people making stuff happen right now.
11:21
And what a few good people who wanna make stuff happen right now can do is our subject always, all the time. We've made an extraordinary amount of what a few people who wanna get something done can do everywhere you look. That's why FOSDEM is such a wonderful experience
11:41
every single year. Bedale's gonna talk to you in a minute about what just a few people who really wanna get something done have gotten done in FreedomBox, but I need to say that none of it would have gotten done without Bedale, whose leadership has made all of this possible. And then there's Nick and James
12:00
and a bunch of other people, all of whom deserve the praise of the human race and statues of solid gold that they're not gonna get. And then there's me trying to figure out how to keep the postal address working
12:22
and stuff like that, which is not interesting or important. What is really interesting now is that this is a time when in addition to a few really good people to make free software, we need you to go out and be a few good people teaching people how to use it and why it's so important.
12:43
We're gonna offer, as Bedale is gonna show you, a stack that is ready for people to use. It doesn't do all the things it needs to do, but it does things for people that will make their lives better now. And it creates demand for more of itself as good technology does.
13:02
And we need to start helping the circle of adoption go past geek. And for that purpose, integration with Debian is crucial because we can go a lot of places. And the wonderful things that are happening with hardware, so wonderful, I leave Bedale to hold them up
13:21
and show them to you. The wonderful things that are happening with all that cheap, free hardware that's going to take over the world are now upon us. There is stuff people can hold in their hands and use. And of course, every old machine sitting under a bed is a freedom box. And soon every Android powered dishwasher,
13:42
of which there will be some on sale shortly, the oven is coming to the United States this summer, right? We need every coffee pot to be a secure telephone. When that happens, despotism will not be immortal. So we need to meet the internet of things with freedom for the internet of things.
14:04
And that's what I think freedom box defines itself to be. And the target I need to go at with all my friends who are doing this wonderful work. And now it's about how do we start getting people
14:20
to expect that maybe the 12 year old in their neighborhood knows how to turn a coffee pot into a secure telephone. We gotta go out and meet the users. We gotta go out and start being in each one, teach one mode pretty soon. That's where we need to be on the day we open up because the guys on the other side,
14:42
they're getting near their chalk marks too. And if they arrive on their chalk marks before we have taught enough people around the world how to stay free, then a rainy day is coming. That's my last word on this subject and I'm gonna turn it over to the people
15:02
who really do the work represented by the man who makes it actually happen. So this really is the problem of co-presenting with Eben Moglen.
15:21
He's a heck of a tough act to follow. But I'll be completely blunt and say that the series of talks that Eben began making three or four years ago that sort of culminated in his opening keynote here two years ago are really sort of what inspired me to believe that this work is something that was worth doing
15:42
and worth trying to make time for. Unfortunately, given some of the chaos in my personal life over the last year, it's been more a question of trying to make time for it than actually succeeding at making time for it. But despite that, I and various other folks who, some of whom Eben's already have named, have made a lot of progress.
16:01
And we are, as Eben said, getting much, much closer than we've been in the past as something that we think might actually be useful for people other than the purely geeky. So let me, before I dive in, also mention that the session immediately following this one in this same lecture hall is something
16:20
that if you haven't noticed already, you might wanna hang around for. Representatives of some of the major federated, distributed social network and voice networking projects are gonna band together to talk about some of the issues that affect all of us as we try to actually implement useful applications on top of the layer of functionality
16:43
that those of us in FreedomBox are trying to deliver. For whatever reason, I've been invited to participate in that discussion, so I'll be staying around, and I hope some of you will be able to stay around too. So when I was asked by the Fosdom folks to answer some interview questions before this event,
17:03
one of the things I was asked was what are the things that have been challenging? What's caused you to be in the situation where you have not actually hit something that you're willing to call a 1.0 milestone yet? And I think that really breaks down into two sort of fundamental issues. One is something I talked about a little bit here
17:21
on Sunday afternoon at last year's Fosdom, and that's that I've spent entirely too much time dealing with binary blobs, firmware device drivers and so forth for the hardware devices that were the best in class things available at the time that turned out to not be nearly as friendly
17:40
to free software as we would have liked. And the other thing is that it turns out it's actually been fairly difficult to implement our intended sort of person model around open PGP keys, which are the kinds of keys used by GPG and so forth. In part, this has to do with the fact that key and signature information tends to be presented
18:03
and used by different applications in way too many different forms, and this complicates the problem. But there's also a really interesting conceptual disconnect between the way most of the sort of encrypted commerce world around HTTPS works with this notion that checking a key really means
18:23
checking to see if some trusted third party agrees that that's a valid key, not actually checking that that particular identity is who it says it is and so forth. And so there's sort of a conceptual disconnect between the centralized CA cartel model of key validation
18:41
and the way we think about validating trust in sort of the network distributed model that our GPG-based web of trust is all about. I'll talk about that a little bit more as we go forward. So with respect to binary blobs, I just freely admit that we spent too much of our time over the last couple of years dealing with this.
19:02
Part of that is that we made a set of commitments to people who participated in our early Kickstarter campaign, which was used to raise money to get the foundation started and sort of get this project off the ground, around sort of the fact that we would deliver specific piece of hardware to them with useful software on it.
19:20
And that meant that as freedom-loving people, the idea of delivering even a prototype of a freedom box that had embedded in it GPL violations just wasn't gonna happen. So in hindsight, the decision that Eben and I jointly made early on to treat this purely as a software project and to not try and engage
19:40
in specific hardware development in support of the freedom box activity, I think is still probably the right choice. But in 2020 hindsight, had we chosen from the beginning to set out by building a cute little piece of hardware that was actually freedom-respecting, a lot of the time that we spent spinning our wheels
20:01
dealing with binary blobs and existing hardware devices might very well have been better spent in other ways. And personal regrets aside, I think that's worth sort of pointing to as something that we spent way too much time on. As I mentioned, basing the idea of our identity
20:21
and authentication model on OpenPGP keys, I still think is absolutely the right thing to do and we are very, very close to having all the pieces in place to make this real. You know, the advantage of it is really that today, the web of trust represented by those of us here and elsewhere who actually use OpenPGP keys
20:43
in our routine free software development activities is really the best evolved distributed web of trust that exists today and there is a high correlation, I believe, between people who might be early adopters of something like freedom box and the related technology elements that we're working on and the people who already participate
21:01
in the OpenPGP key using community. The problem, of course, is that most existing apps, as I mentioned, assume this CA model of the world and there is a really significant conceptual disconnect and this is harder than it seems to fix. You know, I originally thought it would be really easy to go replace one of the Apache authentication modules
21:24
with something that interfaced with MonkeySphere and it turns out that thanks to work from DKG that I'll talk a little bit more about later, we're actually very close to having that all working but that's been harder than you'd think because the question you're trying to ask
21:40
when a browser delivers an X.509 certificate to a web server is not exactly the same question you're asking when you try to check and see whether a key is valid in the GPG web of trust and it's as much an issue of working through those conceptual disconnects as actually working through the technical interactions
22:00
that has caused us to have to spend a lot of time. It's also really frustrating that key and signature and certificate content now sort of is presented to the world in so many different forms and this is, I think, a function of the fact that we as a using community, not just those of us in this room but in general, users of information technology products
22:24
around the world just haven't made as much use of encryption as would lead us to be sufficiently frustrated about this to drive to a tighter, more coherent set of standards and frankly, one of the frustrations for me personally is that this is an area where I am marginally technically qualified.
22:43
I did once take some mathematics of crypto algorithm classes and there were moments in my personal history where those little glimmers of understanding flickered through and I went, oh, wow, gee, that really does work and I understand the math.
23:00
Those neurons have now completely and thoroughly shriveled and died and I trust and count on people much better at this than I am to make sure that we get it right. Those people, it's kind of like, we joke in the Linux kernel community that if you know how to get patches accepted into the kernel.org mainstream, you have job offers available to you.
23:22
The same problem is true in the crypto world. If you actually know how to frog key and certificate information successfully, you're busy and even if you're passionate about something and making time for it can be very difficult. Okay, so what can I say about sort of where we are with respect to binary blobs and all of that?
23:43
One of the things that Evan sort of mentioned in his opening is that by choosing to do all of the work that we're doing within the context of the Debian project, in essence, anything that's capable of running Debian is capable of being a freedom box and some of the work that we're engaged in right now is intended to restructure the way we deliver
24:03
package dependency and configuration information so that that's actually literally true and not just a really good idea. But despite that, we had to choose something to use as an initial hardware target. I think many of you have seen these before. This is a dream plug from GlobalScale Technologies. It's the hardware target we chose
24:22
as our initial primary target. It's actually a pretty reasonable piece of hardware sporting one of my brand new cut vinyl freedom box logo decals that I have piles of with me if you want one, see me afterwards, hint, hint. But there are other things that people have been very interested in trying to use. I'll mention just briefly sort of what's going on
24:42
with a couple of these projects. I assume most of you have heard of Raspberry Pi. Is there anybody here who hasn't? And will admit it. I bet not many of you though have heard about Navina. Is there anybody who has? Yes, yeah, okay, we have a few friends in the audience.
25:01
I'll talk a little bit more about that because it's something I'm really excited about. Okay, so the dream plug, I'm not gonna belabor the details. It's a cute little Marvell Kirkwood based box. It draws very little power. It's actually not nearly as big as it looks because that's just power supply. That's the actual plug server. This really can run off a few AA batteries
25:20
for a lot of hours and provide useful services over a Wi-Fi interface in that mode. So it's not a bad piece of hardware. It's just that I wish I hadn't spent as many hours of my life in the last two years dealing with its imperfections as I have. Raspberry Pi, not a great freedom-loving piece of hardware.
25:42
What's wonderful about it is that it's so cheap. What's not wonderful about it is that the processor, SOC that was chosen, does essentially require that you use binary blobs in order to make things work. However, it's just too cheap to ignore. It's $35, low power board, ARM-based processor,
26:01
capable of supporting some storage. It runs a Debian derivative by default. We have to be sort of clear about that. A Raspbian's not exactly Debian, but it's pretty close. And so a lot of folks have expressed great interest in using this. It's completely plausible as a personal server. It's not really ideally suited for being used
26:21
in a network routing kind of configuration because it's got one, at most, one network interface built in. But with the usual USB device plugin trickery, you can do things with that. And there's at least one company that I'm aware of in Europe that's interested in trying to deliver a product and services bundle using the FreedomBox reference implementation,
26:43
this kind of hardware, and some associated web services. So I think we're going to see things happening here. Nick Daly, who works with me on Freedom Maker and related things for FreedomBox, has a Raspberry Pi board, and we're working on trying to get some build stuff for that fully integrated into our tree.
27:02
This, however, is the thing that I'm currently immensely excited about. And I think, in all fairness, that some of the features that are part of this board are things that the activity of the FreedomBox Foundation has at least partially inspired. All right, does Bunny's name ring a bell to most of you?
27:22
Do you recognize who he is? It's the guy who was involved in reverse engineering, the original Xbox, and is the guy behind the Chumbee and other interesting sort of open hardware projects since then. He recently blogged about the fact that he's been working for several months on the design of a completely
27:40
open hardware laptop implementation. And this is one of the five boards in existence that are the early prototypes of that hardware. I found out about this very early on and had the opportunity to make some suggestions to him about, in particular, trying to make sure that at least one of the subsets of the hardware
28:00
that's on this board could be put to good use by people who don't want to run any binary blobs. The processor he's chosen, the Freescale i.MX6, has a wonderful 3D graphics accelerator, for example, that is completely relevant for a laptop implementation, not terribly interesting for a freedom box. That accelerator does not have an open source
28:23
3D acceleration driver at the moment. There's some interesting conversations I had with folks at Linux Conference Australia in Canberra last week suggest that there's actually more progress in the direction of opening up even that blob than I would have guessed. The thing that's interesting to me about this board is not only is it completely freedom-preserving
28:44
in that there are no binary blobs necessary to make this board work at all, it's got a lot of other neat little hardware features that matter to us, but one of the things that is really cool about this is it's so much faster from a processing performance standpoint than the Kirkwood chips that are in things
29:02
like the Dreamplug that it might actually be possible with hardware platforms like this for us to be able to consider including big steaming piles of PHP code in a freedom box implementation because we might finally have the ability to successfully jail or virtualize those
29:20
to keep the higher sort of risk profile of successful penetration, particularly when many of us working on this project don't understand that class of language and its coding paradigms as well as some people do from being as big a risk as it might otherwise be. Frankly, the processors that are in devices like this
29:41
are really good at running traditional services, the kinds of things like DNS and web serving and email relaying and so forth. They're not as well-suited for running really big stacks of web services kinds of code. So I'm actually really excited about this board. As I said, it's essentially an open laptop project,
30:02
but surprise, surprise, the hardware is also really capable of providing a really good platform for something like freedom box. Okay, so that's enough about hardware. I think the point I really wanted to make is that after spending the last couple of years spending way too much time dealing with things
30:21
like binary blobs, the fact that our work and other things happening in the world have helped to inspire the creation of this and other credible, completely open hardware designs that are freedom respecting is a really exciting transition. So what does it mean to get to something
30:41
that we're willing to call a 1.0 release of freedom box? Well, for me, there are really three objectives that we have to accomplish in order to be willing to call something a 1.0 release. The first is that we have to make the software that we're building more accessible to people who would identify themselves as being less geeky.
31:02
And what I mean by that is I go back to when I was talking last year, I mentioned the four pillars of the work that we were doing at the foundation and that I thought the most important one of those pillars was work on user interface and user design. We have to make it possible for people to configure and use free software solutions
31:20
without necessarily having to become as adept at running pure text editors and whacking away at config files directly as some of us are capable of doing. The second thing, as I mentioned, is we have to broaden the set of available hardware targets because not everybody wants to go buy something like this.
31:41
There are parts of the world where using plug servers based on ARM processor technology that are extremely low power is actually one of the more compelling motivations for wanting to play with this stuff, but there are other places where there's just enough existing x86 PC hardware sitting around that buying new stuff doesn't matter.
32:00
And then finally, I think we have to successfully establish services that are capable of using this OpenPGP key-based trust model before I'm willing to call anything that we release something like a 1.0. So in order to accomplish that, the things that we're really focused on right now
32:20
are trying to take the plinth user interface that James and others have built for use with FreedomBox and finish that, get it into packaged form in Debian so that we can use it more easily than is currently possible to flesh out the rest of the action scripts and menu structure elements to actually make it a usable user interface
32:42
for the functionality we want to deliver in this release and generally just sort of get it from a science for a project to something that's actually generally useful on an average day. The second thing is that we're restructuring the FreedomMaker tool so that instead of encapsulating lots of package knowledge information
33:00
alongside the tools for creating actual images for SD cards and so forth, that that's split out and that we actually deliver a set of Debian metapackages so that you can take any existing Debian system and do something as simple as an apt-get FreedomBox and bring in the core set of packages,
33:20
services, and related configuration files to be able to participate in a FreedomBox network. And then the last piece is that we've been doing quite a bit of work on getting the Monkeysphere tool which, again, I mentioned last year is the thing that allows us to connect open PGP key information into other services that you don't normally think of as using that
33:43
to get that into a state where we can actually make productive use of it. So with Plinth, if you haven't heard me talk about this before or haven't had a chance to see this, this is a web front-end user interface for FreedomBox. It supports this very pluggable infrastructure
34:00
for forms and menus so that the idea of the pluggable structure is that over time as we add additional applications and services to the FreedomBox builds, those can each come with additional fragments of information about how to configure themselves that all gets aggregated together dynamically
34:21
into the administrative user interface. This was built by James Basile using Python on top of Cherry Pi. I don't know how many of you, how many of you are familiar with Cherry Pi? Yeah, I wasn't. And so trying to understand how that works to figure out how to do the right things to get it configured and all has been
34:41
sort of an interesting challenge for me, but we're almost there. Uses an interesting little separate Python class library for doing privilege separation. Once again, these are things that I'm very adept at doing in C and trying to make sure that I understand all of the issues and how these things should properly be packaged and configured and Python has been sort of a real pain in the rear.
35:03
If any of you are a complete and total Python fanatics and would like to dive in and help me make sure I'm not making mistakes there, I'd love to hear from you. And you know, sort of in the classic, let's sit around in a hack fest and work on this and do it in a sort of modern web services kind of way,
35:22
Plinth was unfortunately not really designed to be packaged. You know, of course you'd wanna run it directly out of the directory you checked out from Git and so forth and so working through those things and restructuring it so that it can be a reasonably policy compliant Debian package has been interesting, I'm almost done with that though and hopefully I'll be able to push that set of commits
35:42
sometime this week and do an initial upload to Debian. That of course means it's too late for inclusion in Debian Weezy but not to be worried. I don't let those things worry me. For image building, to date we've mostly been using a tool called Freedom Maker that I wrote a year and a half or so ago.
36:01
Nick Daly who's also here, where are you sitting Nick? Yeah, here he is. Nick is the other guy who deserves sort of hero worship for making things work here. He's been maintaining something we call the Shiny Branch. That branch incorporates a number of additional applications that we are sort of interested in having people play with but have not been merged back into our main branch and are not committed
36:22
for inclusion in future releases. The thing as I've mentioned a couple times though is that we're in the process of restructuring Freedom Maker so that it encapsulates sort of less package dependency information and instead uses Debian meta packages because that's one of the ways of making all of this
36:41
sort of work with less friction on more hardware platforms. I keep talking about MonkeySphere. How many of you have not heard of MonkeySphere before? Go look it up. I've said that every time I've talked about Freedom Maker. I think MonkeySphere, not everybody is totally thrilled
37:02
with the way MonkeySphere is designed and worked but it's one of those sort of necessary pieces of glue to connect different world views together. It was originally developed for use with SSH and in fact my initial indoctrination into MonkeySphere was that if you want to actually do things like push commits into the FreedomBox foundations,
37:22
web, icky wicky instance, you have to be on IPv6 and using MonkeySphere in order to authenticate to the servers that those things are hosted on. But Dan Gilmore has recently resurfaced and is finally working on something that we've been talking about for a while which is resurrecting and updating the Apache mod GNU TLS
37:43
so that it can do MonkeySphere callouts and we're actually the point now where we can successfully validate a client browser's X.509 certificate against an open PGP identity using this chain. There's a bunch more work to do here. If any of you find this particular kind of work
38:02
really intriguing, get in touch with me. I'll get you in touch with Dan and hopefully we can push that through to conclusion very quickly as well. So with all of those things coming together fairly well and just sort of linear time amounts of work remaining to get things concluded,
38:20
I think we're narrowing in pretty well on sort of the short list of services that we actually hope to include in something like a 1.0 release. This list probably looks a little boring and that's sort of intentional because the emphasis for the 1.0 release is on taking the work that we've already done
38:41
and the things that we already use and making those more accessible to a broader audience of users. If it were anything other than boring to the geeky of us in the room, maybe we would be doing the wrong thing. We have all sorts of other services that we would like to work on and we would like to deliver and some of which people are working very hard on right now
39:03
but those are all things that'll be part of a 2.0 or a 3.0 or a 4.0 or whatever release. We don't consider them critical to cross a threshold of utility at which we can really start encouraging people to acquire personal servers, deploy software on them and start making use of it for themselves.
39:23
I won't go into great detail on any of these. I think it's all reasonably self-evident and we've talked about each of these pieces in the past but with this set of services, we will be to the point where someone deploying a personal server will be able to make use of it
39:40
to provide greater security to their set of existing client computing devices without having to modify those client devices to know a whole lot more about the world. Just running your web transactions through a privacy instance with the right rule set, for example, can dramatically increase your security and privacy on the internet from any random client PC today.
40:04
So how can you help? This is what I'm constantly being asked about. It really starts with being conscious about privacy and freedom and all the things you do. If you have not already created for yourself a GPG key and gotten yourself plugged into the global web of trust
40:21
based on GPG keys, now would be a great time to do that. There are a bunch of us here in the room who would be happy to sign keys if you already have them and all of those sorts of things. I mentioned that I'm just about to cross the threshold of having the plinth user interface structure packaged and sort of ready to be integrated
40:41
with the rest of the image. The thing that's still missing is filling in action scripts that actually do useful administrative things. We have a couple of examples like adding and removing users and frobbing some of the basic network configuration already done but we'd love to have help from people who are comfortable whacking around with a little bit of Python
41:01
and in particular if you are living in the intersection of people who grok Python and understand how Debian packages configure various things, I would love to hear from you and have your help. As we move towards adding additional services for future releases, we could still use more help in selecting appropriate Debian packages
41:21
and working out the configuration for those to make them actually deliver the functionality that's part of our vision and articulated on some of the content within the Wiki site that we're maintaining is being ways that we wanna work for the future. And of course, since the foundation is a nonprofit organization and all of us are volunteers
41:41
and yet still have various expenses, I'd be remiss if I didn't mention that financial donations are always welcome. If you'd like to learn more, these are the places to go to take a look. With that, I'm gonna thank you very much for your time and attention and hand back to Eben to say a few closing words.
42:09
So what is there to say that there was a time when we thought it's gonna rain. Not yet, but it's gonna rain.
42:22
And we went out to build a boat. And as you can see, it's still sort of a small boat. And there's not a lot of people working on it. And the wonderful thing is there doesn't have to be a lot of people working on it because so many people work on so much wonderful stuff and you just have to hook it all together.
42:44
But any free software project needs users in order to grow and we really need users because what we're doing here feels like the climax of the free software movement
43:01
at the moment when it becomes the freedom for everybody in a technological universe movement that we always knew it was going to have to convert into. Beedel has, I think, heartbreakingly shown just how hard it is to set out to do a thing
43:21
even now that involves free hardware because lots of little bits of gum got pushed into the hardware system as it began to be the software and hardware are always mixed together system. And we need free hardware. We set it a decade ago and we still need it.
43:42
But the beauty is it's finally arriving. And it's going to accelerate very rapidly in quality and power. And one of the reasons that is happening is the world is filling up with recyclable hardware that is modular and you can redesign it
44:01
as you take it apart and put it back together again. All over the world, the waste stream now includes wonderful hardware. And recycling that hardware and rebuilding it into appropriate technology that costs a fraction of what it would cost to design and build if all the components weren't already being thrown away
44:24
is going to have some very interesting consequences for us. So we are building our boats out of better and better parts because parts are filling up waste cans all over the world that are really wonderful and that can be put together in freedom.
44:40
And in the same way that the free software movement has ironically been benefiting by the effort made by people who didn't care about freedom very much at all, that's gonna happen in a hardware world that cares about freedom even less and is gonna be helping freedom to grow even more soon. Software we can always make.
45:00
Changing habits for users around the world is not a thing that we can count on anybody else to do for us. All the habit building around the world is now being done by people who are teaching people to carry robots in their pockets that don't work for them and that inform on them all the time. Which is a really scary thing to teach people.
45:22
What Bedell says about what you can do with just a little privacy and a little SSH proxying is the really key point because we can actually help people to free themselves with respect to the robots they are carrying. The FTC is gonna talk in the United States about applications really should have a do not track box
45:43
you can check and it should really mean something and they're gonna talk about that for five years and then maybe nothing will happen and then they'll talk about it five more years more and then nothing will happen. And those of you who work with the European Commission and the European Parliament and heroically strive to make this society's efforts correspond to freedom
46:04
know that that's exactly what will happen here too. And meantime, we can be using cheap boxes and boxes under people's beds and thrown away computers to give them something that will actually connect up with the devices they carry and help to make those devices more safe. In an editorial in the communications of the ACM
46:23
out this month, I refer to these as little personal coal scrubbers that improve the air everybody breathes and cost nothing to run. We can't count on people to put out clean air we can make air purifiers and teach people how to use them. That's what we're gonna have to do.
46:41
So as Bedell says, there's places to get involved if you wanna help us put stuff together so that we can make an open PGP key authenticate a route between people to do anything a web server can do. And after that, anything, any TCP connecting service can do and so on out. Or you can decide that what you'd like to do
47:01
is get a computer out from under a bed, put a Debian free software stack on it, hand it to somebody and show them how to use it. And suddenly a whole bunch of devices that are hurting people stop hurting them. And we hope it's painless. And if we teach enough people how to do that and they demand more for more people that they know
47:22
so they can do it too, then we are starting to clean up an ecological nightmare before it poisons us all. There's just a few of us and we need only a little more and only a little more code and only a little more money and only a little more time and only a little more sleep.
47:48
But the world needs a lot more freedom and you have to help us get it. Thank you very much.
48:03
I think there is time for just one question or two if they are answered quickly. Anybody wants to ask a question?
48:22
Okay, so I have a question. Have any other distributions behind Debian got in touch with you or do you know of any other distribution that would package parts of freedom? There's a lot of background noise and I can't hear you. I have to start again. Sorry, okay, so again, have any other distributions gotten in touch with you
48:41
about packaging freedom box parts? We've talked to all sorts of people but honestly, so far all of the activity within the context of what we're doing around the foundation has been done with Debian. We're certainly completely open to the idea that these ideas and protocols and technology
49:03
can and should be implementable and deployable anywhere and all of the software that is being created and all of the glue scripts and everything else are in publicly visible Git repository so there's nothing to stop people from doing that but to date, I'm not aware of anybody actually doing it.
49:23
Hello, you talk of freedom box here. You talk of freedom box as a tool to prevent people getting hurt but in your vision, to what extent would you see it as a tool for hurting the baddies? How far do you see it as a tool
49:41
for subversion, civil disobedience and counteracting the system? It's not going to be more than a drop in the bucket. The problem is that without that drop in the bucket, unfreedom is so much easier. I think about this as merely defensive activity
50:05
by and large intended to prevent disaster. A slightly better functioning net substantially reduces the likelihood that the net becomes part of the major tools of unfreedom in the 21st century.
50:22
We're moving to a world in which government's ability to employ big data changes the relationship between citizen and state drastically. We need to find ways to help people live their lives as though in freedom.
50:41
That doesn't bring freedom. That creates the opportunity. It's not a situation that will be unfamiliar to anybody who has thought about the path of Polish politics in the late 20th century and its effect. Living as though in freedom helps to produce freedom but it isn't freedom.
51:01
It's just a way of saving us from tyranny. That's what I would say about that.