0:00:00.000,0:00:04.930
[Talkmeister] Next, we will have zack presenting[br]"Debian in the Dark Ages of Free Software"

0:00:08.690,0:00:17.380
Can you hear me?

0:00:17.701,0:00:18.420
Better.

0:00:18.570,0:00:19.562
So, hello everyone.

0:00:19.683,0:00:21.382
Welcome again to DebConf, I guess.

0:00:21.522,0:00:25.360
It's a great pleasure to be back again [br]at one DebConf

0:00:25.443,0:00:28.447
and a great honor to be doing one [br]of the opening talks.

0:00:29.161,0:00:31.801
I confess I wasn't really expecting[br]that honor.

0:00:31.870,0:00:33.531
I just wanted to propose a session

0:00:33.559,0:00:36.281
which was supposed to be[br]a self held session

0:00:36.281,0:00:39.326
for those of us that think there are[br]some worries

0:00:39.520,0:00:43.320
about where the free software is going[br]in general.

0:00:43.763,0:00:47.651
And the role that distributions have to play[br]in the current state of affairs.

0:00:48.004,0:00:50.925
So this talk will be about a couple of[br]journeys at once.

0:00:51.216,0:00:54.286
The first journey is a journey[br]through emotions,

0:00:54.569,0:00:58.640
through good feelings about what[br]we have achieved in Free Software

0:00:58.921,0:01:01.800
over the past 15 to 20 or 30 years

0:01:01.824,0:01:03.860
depending on how long you've been[br]involved.

0:01:04.006,0:01:06.284
The second journey is essentially [br]my own journey[br]

0:01:06.284,0:01:08.122
through software freedom

0:01:08.132,0:01:10.840
from the day I started discovering[br]Free Software

0:01:10.866,0:01:13.240
and what I've ended up doing since then.

0:01:14.282,0:01:17.240
Starting with the positive news.

0:01:17.640,0:01:23.001
This is how I got involved myself[br]in free software in 1997.

0:01:23.050,0:01:25.140
I understand that there are people[br]in the room

0:01:25.161,0:01:27.920
who have been involved [br]since way earlier than that,

0:01:27.949,0:01:30.360
others that have been involved[br]since way later than that.

0:01:30.450,0:01:31.561
Well, that's my story.

0:01:32.046,0:01:34.802
I hope you'll find ??? points[br]with your own story.

0:01:35.560,0:01:41.041
When I started as a freshman in a computer[br]science class at university of Bologna,

0:01:41.360,0:01:43.600
that was a huge tiping point,

0:01:43.845,0:01:47.121
a huge hype point for the so-called[br]opensource movement.

0:01:47.560,0:01:52.040
That was the year the very influencial[br]essay by Eric Raymond has been published.

0:01:52.404,0:01:58.528
That was the year that Netscape decided to[br]opensource its own code.

0:01:58.870,0:02:01.210
That was the moment in the history of[br]free software

0:02:01.320,0:02:04.000
when people were trying to sell[br]to the industry

0:02:04.280,0:02:09.431
what free software was doing, and[br]I'm not using that word in a bad sense.

0:02:09.610,0:02:12.970
There was reasonable concern that[br]without involvement of the industry,

0:02:13.000,0:02:16.400
the free software movement wouldn't have[br]got far.

0:02:16.773,0:02:21.830
So they were trying to tell about free[br]software in an industry-friendly way.

0:02:22.283,0:02:25.203
Essentially, the rhetoric at the point[br]was that

0:02:25.475,0:02:28.800
if you do development of software[br]in the free software way,

0:02:28.960,0:02:31.230
in a more open way,[br]a more participative way,

0:02:31.360,0:02:35.693
you will end up having better software[br]and that by merely opening up you code

0:02:35.896,0:02:40.526
you'll have these flocks of programmers[br]coming to you project and end up helping you.

0:02:41.600,0:02:47.046
A few years later, I realised that[br]I personally didn't believe much in that idea:

0:02:47.261,0:02:50.390
it's only because your software is open[br]that it's gonna be better,

0:02:50.727,0:02:53.521
but it was a fair thing to try[br]at the time.

0:02:54.282,0:02:57.304
What I discovered a bit later is actually[br]what stuck in me

0:02:57.304,0:03:00.336
was essentially the philosophy[br]of free software.

0:03:00.336,0:03:05.360
The fact that computer user should be[br]in charge and in control of their own machine,

0:03:05.480,0:03:07.200
that should have some basic freedom.

0:03:07.241,0:03:10.770
You know about the 4 freedoms,[br]I'm not going to repeat them here,

0:03:11.067,0:03:15.922
but my personal point is that[br]the narrative of free software is something

0:03:15.960,0:03:17.890
that resonated with me a lot at the time.

0:03:18.201,0:03:21.334
As a student, I realised that by having[br]free software at my fingertip

0:03:21.334,0:03:22.847
as a computer science student,

0:03:22.847,0:03:25.898
I could debug any single layer[br]of the software stack

0:03:25.898,0:03:27.699
and look at how things are going.

0:03:27.699,0:03:32.000
I didn't have to trust the teacher on how[br]an operating system should be developed.

0:03:32.121,0:03:36.314
I was able to open up sched.c in[br]the linux kernel and have a look

0:03:36.314,0:03:41.440
at the actual scheduling algorithm that[br]was being implemented in the real kernel.

0:03:41.489,0:03:43.555
Not that I really got all of it at the time

0:03:43.555,0:03:46.781
but the possibility was just breathtaking[br]for me.

0:03:47.911,0:03:52.982
Later on, I ended up distilling[br]the main intuition of free software,

0:03:52.982,0:03:55.980
which is the one I used to explain[br]free software to people,

0:03:55.980,0:03:57.600
which is intuition of control.

0:03:57.832,0:04:03.303
So, I ended up believing that the main[br]reason why I've been involved in this movement

0:04:03.362,0:04:07.750
for about fifteen years is that I really believe[br]that every single computer user,

0:04:08.013,0:04:10.249
and that's a lot of people these days,

0:04:10.249,0:04:13.755
should be in control over[br]their own computations.

0:04:13.755,0:04:17.212
Everything you're doing with a device[br]which is mediated via software

0:04:17.440,0:04:20.840
is controled by someone,[br]either it is you or it is someone else.

0:04:21.539,0:04:24.350
And the best episode, the best narrative[br]to explain that to people

0:04:24.391,0:04:26.794
that they've been using for quite a while[br]is this passage

0:04:26.794,0:04:29.197
from the novel "Makers" by Cory Doctorow

0:04:29.197,0:04:31.600
which is a bit long so I'm not gonna read it in detail,

0:04:31.752,0:04:35.518
but essentially there is one character[br]of the novel which is Lester

0:04:35.760,0:04:39.580
which is explaining to another character[br]the importance of controling

0:04:39.580,0:04:41.370
your own devices, your own tools.

0:04:41.370,0:04:44.170
The first example he takes is the example[br]of a hammer,

0:04:44.170,0:04:45.840
a physical hammer,

0:04:45.840,0:04:48.570
and he goes on saying that if you own[br]a hammer,

0:04:48.750,0:04:50.940
essentially you could do[br]whatever you want with it.

0:04:50.980,0:04:53.303
You can use it for its main purpose,

0:04:53.303,0:04:55.426
or you can use it for something[br]completely different

0:04:55.426,0:04:58.520
which was not meant to be its original[br]purpose but it's you that decide.

0:04:58.952,0:05:03.886
He compares that another device[br]which is the "Disney in a box" in the novel

0:05:04.155,0:05:08.487
and Disney in this book is the big evil[br]villain which is oppressing people

0:05:08.770,0:05:14.271
and essentially Disney in a box is a[br]glorified 3D printer that can only print

0:05:15.131,0:05:17.960
what Disney wants it to print for that day.

0:05:18.632,0:05:22.071
One day, it will print a Goofy character,

0:05:22.071,0:05:24.600
another day it will print Donald Duck,

0:05:24.600,0:05:26.160
but it's not you who decides.

0:05:26.217,0:05:30.227
It's Disney that decides what the printer[br]is gonna print for you that day.

0:05:30.560,0:05:33.968
You own the device but you are[br]not in control of what the device does.

0:05:35.080,0:05:38.831
The big quote for me is that if you don't[br]control your life, you're miserable.

0:05:39.560,0:05:45.920
This notion of oppression is what has[br]been motivating me for all these years.

0:05:46.160,0:05:49.864
So the fact that if you are not in control[br]of your own computation,

0:05:49.864,0:05:52.088
then someone is oppressing you.

0:05:52.088,0:05:56.552
Someone usually is the person or the company[br]or whatever that has created the software,

0:05:56.588,0:05:59.210
that has the power to change that software[br]instead of you.

0:05:59.640,0:06:01.413
This is something that really ??? in me.

0:06:02.711,0:06:05.440
What was I doing at the time[br]with my computer?

0:06:05.440,0:06:08.049
Well I was doing pretty standard stuff.

0:06:08.049,0:06:10.596
I was using some hardware we had at the time

0:06:10.596,0:06:14.063
which was mostly desktops and[br]local network servers.

0:06:14.063,0:06:16.668
I didn't have a laptop because[br]it was really expensive for a student

0:06:16.668,0:06:19.003
so I did get a laptop much later.

0:06:19.003,0:06:21.930
I was doing some content production,[br]some content consumption.

0:06:21.983,0:06:24.836
The kind of content I did produce[br]at the time was mostly

0:06:24.836,0:06:29.139
office suites, desktop publishing[br]and this kind of stuffs.

0:06:29.139,0:06:32.352
I was doing some communication, some email,[br]some IRC, some newsgroup

0:06:32.352,0:06:34.705
which was really cool at the time[br]for geek communities.

0:06:34.705,0:06:36.946
And I was doing some software development[br]as a newbie

0:06:36.946,0:06:38.987
but it was what I was doing at the time.

0:06:38.987,0:06:41.941
I also did some content consumption,[br]some gaming

0:06:41.941,0:06:45.855
which are arguably some content that[br]someone else is producing for you to consume.

0:06:46.075,0:06:47.361
I was doing some web browsing.

0:06:47.461,0:06:49.147
Internet was not as popular as it is today,

0:06:49.147,0:06:52.743
but there were some websites[br]you could find interesting.

0:06:54.163,0:06:57.195
In that situation,[br]with this kind of computing,

0:06:57.565,0:07:02.507
the actual path to software freedom[br]and to control was fairly clear.

0:07:02.817,0:07:07.911
It was difficult, but it was fairly clear[br]to me as a new activist in free software.

0:07:08.057,0:07:11.674
What I should have done, what we all should[br]have done to actually liberate people

0:07:11.674,0:07:14.541
from the oppression of people controling[br]our own computation.

0:07:15.031,0:07:18.948
The idea is that while you have[br]a lot of pieces of proprietary software

0:07:18.948,0:07:22.883
which you do not control, what you need[br]to do is to replace

0:07:22.883,0:07:28.198
every such a component of proprietary[br]software with a free software equivalent.

0:07:28.658,0:07:31.626
Using some local application, some game,

0:07:31.626,0:07:34.054
we need to replace it[br]with an equivalent free game.

0:07:34.054,0:07:39.156
We were using some client-server software,[br]some mail ???, some mail client,

0:07:39.156,0:07:42.218
some mail server, some IRC client,[br]some IRC server.

0:07:42.218,0:07:47.395
What we needed to do to actually empower[br]people and liberate people was to rewrite

0:07:47.395,0:07:50.862
those pieces of software with free software equivalents.

0:07:50.942,0:07:55.360
It was difficult, because it was a lot of stuff[br]to be rewritten, but it was fairly clear.

0:07:55.599,0:07:56.561
The plan was clear.

0:07:56.842,0:08:01.560
And also, luckily, we also had, at the time,[br]all the heavy lifting was already in place.

0:08:01.880,0:08:04.570
The GNU project existed since[br]quite a while,

0:08:04.570,0:08:07.460
the Linux kernel existed already[br]and it was working.

0:08:07.460,0:08:11.963
So someone else with shoulders larger[br]than I had at the time

0:08:11.963,0:08:15.252
had already done a lot of work for me and me

0:08:15.252,0:08:19.458
and together with other free software activists,[br]what I had to focus on was to rewrite

0:08:19.458,0:08:24.294
proprietary application into equivalent[br]free software application, possibly better.

0:08:24.514,0:08:27.480
That was clear, was hard,[br]but it was fairly clear.

0:08:29.001,0:08:34.570
That's where, I think, the notion[br]of a free software project comes from.

0:08:34.813,0:08:39.955
We use very often this term of free[br]software project and I never ended up

0:08:39.955,0:08:44.323
really thinking about that before a few[br]years ago and I think the reason why

0:08:44.323,0:08:46.881
we call it free software project is that[br]there is an objective.

0:08:46.881,0:08:49.914
So there is a mission,[br]ideally a time-limited one,

0:08:49.914,0:08:54.916
and that mission is writing a replacement[br]for a proprietary application using

0:08:54.916,0:08:58.568
free software which is as good,[br]possibly better than the original.

0:08:59.338,0:09:03.612
Having a lot of free software projects[br]around gives rise to a lot of releases.

0:09:03.960,0:09:06.656
So what we were doing a lot at the time[br]in the 90s

0:09:06.656,0:09:10.092
was to actually manually install[br]software on our own machines.

0:09:10.092,0:09:14.679
To be fair, our lab was running[br]some Red Hat machines.

0:09:14.679,0:09:18.107
At the time there weren't that[br]many packages available and

0:09:18.107,0:09:21.895
we had to fairly often install stuff[br]by hand on the lab machines

0:09:21.895,0:09:25.033
in our own directories and also[br]on our computers at home.

0:09:25.033,0:09:27.633
This is a procedure you all know very well.

0:09:27.933,0:09:31.693
You download a tarball, you run "configure",[br]you run "make", you run "make install".

0:09:32.000,0:09:35.529
The first time I saw that, it was kind of[br]a magical recipe for me.

0:09:35.720,0:09:38.402
Just follow these steps and you will get[br]some software to play with.

0:09:38.672,0:09:41.890
Well, except that every single step[br]could fail, of course.

0:09:42.200,0:09:47.023
Let's keep aside for the moment the fact[br]that the website might be down but,

0:09:47.023,0:09:50.726
you run "configure" and you miss some software[br]you need to fetch from somewhere else.

0:09:50.726,0:09:54.480
You run "make", you encounter some[br]compilation problem.

0:09:54.791,0:09:58.428
You run "make install", maybe the path[br]will clash and so on and so forth.

0:09:58.658,0:10:03.160
The problem with this procedure for[br]installing software we are using by hand

0:10:03.880,0:10:05.801
is that you are essentially[br]conflicting roles.

0:10:06.600,0:10:09.800
You're mixing together the role of[br]software user,

0:10:10.201,0:10:13.480
the role of system administrator[br]and the role of software developper.

0:10:14.080,0:10:19.240
You need to have a little bit of all those skills[br]together to be able to enjoy software.

0:10:20.000,0:10:24.760
In a sense, a free software which works[br]like this is essentially a very elistist thing.

0:10:25.160,0:10:29.241
It's only an elite which have[br]all the needed skills who is able to enjoy

0:10:29.560,0:10:34.440
the benefits of free software and is able[br]to be in control of their own computation.

0:10:35.080,0:10:39.240
This is essentially the reason why distributions[br]much earlier had been invented.

0:10:39.720,0:10:42.960
We all know very well here[br]what distributions do,

0:10:43.320,0:10:48.204
they sit in between software developpers[br]and software users and make it easy for you

0:10:48.681,0:10:50.040
to actually use that software.

0:10:50.480,0:10:55.600
We do installer work, we create installers,[br]we create package managers,

0:10:55.752,0:11:00.040
we do all the integration work that make [br]different pieces of software work well together.

0:11:00.480,0:11:04.560
We actually make life easy for final users.

0:11:05.480,0:11:11.395
So, for me, something that I started believing[br]is that the ultimate mission of free software

0:11:11.395,0:11:16.521
distributions is to actually democratize[br]free software, to enable users

0:11:16.720,0:11:21.361
which do not have software development skills[br]or do not have system administration skills,

0:11:21.640,0:11:24.547
enable them to enjoy the benefit[br]of free software.

0:11:24.960,0:11:27.320
We offer very simple interface,

0:11:27.640,0:11:31.601
we offer the equivalent of what these days[br]are called appstores in which

0:11:31.760,0:11:34.640
with one click, you can just install[br]some software and

0:11:34.800,0:11:38.084
enjoy the benefit of that software,[br]in particular a free software.

0:11:39.760,0:11:43.320
This is for me the historical mission[br]of distributions.

0:11:44.440,0:11:49.244
Later on, in 1998, our lab decided[br]to switch to Debian

0:11:49.750,0:11:51.440
and I was really happy about that.

0:11:51.961,0:11:54.120
We switch from Red Hat to Debian and[br]I look out about this project,

0:11:54.120,0:11:58.430
I start learning what this project does[br]and I find out that not only

0:11:58.770,0:12:03.020
this project Debian was actually up to[br]the mission of empowering user

0:12:03.020,0:12:06.480
by making it easy for users[br]to use free software.

0:12:06.980,0:12:12.242
If you read the original announcement of[br]Ian Murdock announcing the Debian project,

0:12:12.242,0:12:15.880
we'll find this notion of being competitive[br]with proprietary operating systems

0:12:16.160,0:12:18.720
and it's really clear that the point is[br]empowering users.

0:12:19.403,0:12:24.560
I end up reading about this project and[br]not only I found out that their mission

0:12:24.887,0:12:28.480
they're up to is the mission I believe in,[br]but I found out that the key intuition there

0:12:28.760,0:12:30.720
is to make the project a community project.

0:12:31.230,0:12:34.121
Not only the target are the users [br]and empowering them,

0:12:34.280,0:12:37.840
but also the way to reach that objective[br]is creating, fostering a community

0:12:38.120,0:12:40.000
that will work together to that goal.

0:12:40.440,0:12:41.760
I got immediately hooked,

0:12:41.927,0:12:44.521
I vividly remember the moment[br]a collegue of mine, a student

0:12:44.680,0:12:47.441
explained to me the anatomy of[br]a Debian source package,

0:12:47.760,0:12:51.120
the fact that it was a .orig.tar.gz,[br]the fact that it was a diff.gz

0:12:51.150,0:12:53.881
with the differences with respect to upstream,[br]and all those metadata

0:12:53.881,0:12:56.481
that was really thrilling for me[br]from a technical point of view.

0:12:57.000,0:13:00.841
A few years later, I ended up joining[br]the nm-process.

0:13:01.369,0:13:05.480
I was doing some OCaml development[br]at the time, there were some libraries,

0:13:05.521,0:13:08.200
OCaml libraries in Debian, others were[br]missing and I said

0:13:08.440,0:13:12.960
"Ok, maybe I should help and create[br]some libraries for the project as well".

0:13:13.240,0:13:20.239
I went through nm and there are a few things[br]I've learned doing nm

0:13:20.280,0:13:23.360
and also in the subsequent ten years[br]or fifteen years or so.

0:13:23.840,0:13:29.720
One thing I've learned in all these years in[br]Debian is the importance of being principled.

0:13:30.520,0:13:34.761
Debian is a project that did not start[br]from only technical means

0:13:35.280,0:13:37.680
but also decided at some point that[br]they needed some guidance,

0:13:37.720,0:13:40.400
some clear guidance of what it should do[br]technically and what it shouldn't.

0:13:41.240,0:13:45.920
And an important document where we have[br]distilled this notion are the DFSG.

0:13:46.602,0:13:47.960
The Debian Free Software Guidance

0:13:48.161,0:13:50.961
which has been very influencial[br]on the free software movement as a whole.

0:13:51.320,0:13:54.280
They've been used as a base for[br]the open source definition as you know,

0:13:54.761,0:13:58.240
and what was very thrilling for me[br]is that commitment we had in Debian

0:13:58.521,0:14:03.057
in keeping the main archive completely[br]DFSG-free, keeping it completely free software.

0:14:03.601,0:14:08.167
This commitment is depicted here[br]by those fearsome character

0:14:08.668,0:14:14.400
and his owner on a couch and it's mediating[br]and triggering the NEW queue, supposedly,

0:14:14.801,0:14:17.479
and the NEW queue is not necessarily[br]the best way we could implement

0:14:17.837,0:14:21.641
a system which triage all the software[br]in the archive and to ensure it's DFSG-free

0:14:21.760,0:14:27.320
but it shows our commitment to actually[br]only follow the guidance we have set for ourselves.

0:14:27.600,0:14:29.480
It was really motivating for me.

0:14:29.760,0:14:32.121
The second thing I've learned and which[br]will come handy in a bit,

0:14:32.121,0:14:37.240
is the importance of the legal knowledge[br]and legal geeks in the free software movement.

0:14:37.440,0:14:42.910
Like it or not, free software as an ideal[br]is philosophical mean,

0:14:42.910,0:14:48.082
but its main implementation is through the[br]legal system, is through copyright licenses.

0:14:48.561,0:14:52.280
To really grok what's happening[br]in free software in general,

0:14:52.280,0:14:55.840
to understand where the free software[br]movement is going, figuring out and

0:14:55.840,0:14:59.321
really understand what's going on[br]in the legal system is very important.

0:15:00.200,0:15:03.920
In Debian, we know that pretty well,[br]that's a stumbling block for many people

0:15:04.061,0:15:06.120
when joining the Debian project.

0:15:06.120,0:15:09.960
It's something we insist people are at least[br]basically familiar with and

0:15:09.960,0:15:14.280
that's pretty characteristic[br]of the Debian project.

0:15:14.520,0:15:17.040
In the end, what I've learned is that

0:15:17.840,0:15:22.660
in this quest that I feel very much myself[br]against the oppression of someone else

0:15:22.660,0:15:27.361
controling your own computation,[br]law, if you hack around it smartly,

0:15:27.361,0:15:31.723
can be a very useful ally,[br]a very useful device to liberate users.

0:15:33.400,0:15:38.480
Time passes − there was supposed to be an[br]image here, which for some reason disappeared.

0:15:38.960,0:15:45.040
And, we might argue that, these days,[br]we have achieved a lot since that moment.

0:15:45.520,0:15:48.922
If I look around the industry or, in general,[br]if I look around computing

0:15:49.240,0:15:54.000
as people are doing that,[br]free software is a little bit everywhere.

0:15:54.680,0:15:58.400
In the industry, there are some stats[br]that claim that essentially

0:15:58.640,0:16:02.320
every single software product you find[br]on the market has, inside of it,

0:16:02.480,0:16:04.600
a little bit of free software code.

0:16:05.536,0:16:08.280
If you look at all the different application[br]stacks we have

0:16:08.642,0:16:11.280
from webservers to education to clients[br]to smartphones,

0:16:11.602,0:16:15.800
you find a lot of free software, free software[br]infrastructures that are everywhere.

0:16:16.400,0:16:19.400
So these are just some stats I figure out[br]in the recent years

0:16:19.880,0:16:24.600
and for instance if we look at one of the[br]key target market for Debian ???

0:16:24.960,0:16:30.441
we'll find out one website over ten[br]on the Internet in general is running Debian.

0:16:30.720,0:16:33.800
If we include also some of our most[br]popular derivatives such as Ubuntu,

0:16:34.121,0:16:35.960
we'll find that more than 20%[br]of the websites

0:16:36.240,0:16:38.560
are running something which comes[br]from our own work.

0:16:39.001,0:16:45.721
And some of the recent hype on free software[br]is coming from the Snowden revelation

0:16:45.721,0:16:50.000
and most people are starting to be concerned[br]about what the software they're using is doing

0:16:50.400,0:16:55.680
and is turning to free software and is turning[br]to stuff like Tails which is heavily Debian-based

0:16:55.680,0:17:00.685
to actually see in which way we can[br]help them foster their own security.

0:17:01.281,0:17:03.000
In some sense, we have achieved a lot.

0:17:03.680,0:17:08.041
In everything we do in computing,[br]there is a little bit of what we have done

0:17:08.041,0:17:10.800
in free software and also a little bit[br]of what we have done in Debian.

0:17:11.160,0:17:12.960
This is pretty impressive for me.

0:17:13.040,0:17:18.560
We're in a place where I wouldn't have[br]dreamed being when I started in 1997.

0:17:18.874,0:17:19.881
That's very impressive.

0:17:20.321,0:17:24.160
On the other hand, there are some reasons[br]of concerns

0:17:24.480,0:17:27.951
and this is the main thought[br]I wanted to share with you.

0:17:30.020,0:17:33.880
There are some technical reasons which[br]we discuss often in free software circles

0:17:34.240,0:17:40.040
like the fact that "Ok but most of these[br]platforms are not 100% free software".

0:17:40.841,0:17:42.560
If you look at smartphones for instance,

0:17:42.760,0:17:48.440
you will find a lot of non free code every here[br]and there and the point can be made that

0:17:48.440,0:17:52.041
either you have full control over[br]your own computation,

0:17:52.200,0:17:53.440
or you are not in control at all,

0:17:53.600,0:17:57.206
because if your software stack has a single layer[br]which is controlled by someone else,

0:17:57.206,0:18:01.200
and is mediating all your communication,[br]maybe you're not so sure

0:18:01.524,0:18:04.600
that you are the real owner and[br]the real controller for your own device.

0:18:04.760,0:18:07.157
That's a absolutely fair point.

0:18:07.160,0:18:11.241
We can make some more technical points[br]about for instance non free JavaScript.

0:18:11.321,0:18:14.520
More and more of our computations are[br]happening in our browsers

0:18:14.855,0:18:17.680
and are happening through code which is[br]delivered to our browser

0:18:17.680,0:18:21.081
by remote servers and this code[br]is not free at all.

0:18:21.160,0:18:26.840
I absolutely agree with that but the point[br]I want to focus on today is actually

0:18:27.040,0:18:29.561
what we call the cloud.

0:18:30.080,0:18:33.012
All my images are gone.

0:18:34.160,0:18:36.040
You had a very nice image there, sorry.

0:18:41.390,0:18:45.525
The remaining point and my main reason of[br]concern is what is being called the cloud.

0:18:45.961,0:18:48.640
Let allow me to be a bit generic here[br]for a moment.

0:18:49.040,0:18:51.360
I know there are very different parts[br]in what we call the cloud

0:18:51.720,0:18:53.800
and will be specific in all of them[br]in a bit.

0:18:54.320,0:18:56.680
But for now I want to focus on the [br]common trend that

0:18:56.720,0:18:58.880
the cloud is bringing to computing[br]these days.

0:18:59.481,0:19:03.897
Computing today, for most people, is not[br]much different from the kind of computing

0:19:03.960,0:19:05.962
I was doing fifteen years ago.

0:19:07.240,0:19:09.960
That's the kind of computing that we do[br]on very different hardware,

0:19:10.120,0:19:13.920
we have way more smartphones, way more[br]tablets than in the past and that's true.

0:19:14.320,0:19:19.160
But the kind of activities we do − producing[br]content, consuming content − is very similar.

0:19:19.720,0:19:23.721
The big difference is the kind of[br]technological stack we're using

0:19:24.040,0:19:25.880
and where the computations are happening.

0:19:26.280,0:19:30.440
For most people today, the kind of[br]office suites we use is no longer

0:19:30.640,0:19:33.687
a software which is installed on [br]your machine but it is Google Docs.

0:19:34.000,0:19:39.200
I'm an academic myself, I'm very often[br]forced to use some Google Docs applications

0:19:39.281,0:19:42.160
to work with others, otherwise I'm free[br]not to work with them,

0:19:42.640,0:19:45.881
because it's a technological choice[br]made by someone else.

0:19:46.320,0:19:49.240
For many people, e-mail, as you know,[br]just means GMail.

0:19:49.480,0:19:52.960
All our e-mails, even if your not[br]using GMail ourselves,

0:19:53.288,0:19:55.480
are passing through some GMail servers.

0:19:55.681,0:20:00.080
Asynchronous communications still exist,[br]but it is very often mediated

0:20:00.200,0:20:01.640
to software like Skype or GTalk.

0:20:02.000,0:20:02.940
And so on and so forth.

0:20:02.940,0:20:04.240
You have seen this list very often.

0:20:04.920,0:20:08.640
Consuming content, there as well,[br]we are still doing gaming,

0:20:08.640,0:20:11.400
we are still doing browsing but it's often[br]mediated by platforms

0:20:11.520,0:20:14.320
which are far away from us and just stream[br]content to us or,

0:20:14.480,0:20:19.320
in the specific case of web browsing,[br]they are more and more often hosted

0:20:19.320,0:20:23.120
by very few hosters in the world − which[br]we often refer to a walled garden −

0:20:23.240,0:20:25.644
that can do whatever they want[br]with our content.

0:20:27.040,0:20:30.720
The point here is not demonizing[br]those services.

0:20:30.960,0:20:35.307
People are using those services because[br]they are convenient and

0:20:35.560,0:20:39.525
there is a lot of network effect going on[br]that makes it easy for other people

0:20:39.601,0:20:41.360
to start using those services.

0:20:41.720,0:20:44.200
It's really not the point of demonizing[br]those services.

0:20:44.200,0:20:49.444
The point here is observing that interesting[br]computations that we are doing

0:20:49.720,0:20:53.200
as our job, as our life,

0:20:53.320,0:20:57.241
are no longer happening on our machines,[br]but are happening on other machines

0:20:57.390,0:21:02.400
which are far away from us and which[br]are not under our direct control.

0:21:02.641,0:21:09.200
In this context, for me, I confess, what[br]actually is the road to software freedom

0:21:09.200,0:21:12.040
and to control, to enable people[br]to control their own computation

0:21:12.120,0:21:13.080
is no longer clear.

0:21:13.400,0:21:17.080
It's no longer enough to say[br]"Well, we just need to rewrite

0:21:17.200,0:21:20.383
Google or Facebook or Twitter[br]in free software".

0:21:20.440,0:21:23.961
That's not enough, because even[br]if you do that, you have the problem

0:21:23.961,0:21:27.401
that when you are using a server[br]you don't know if the code it is running

0:21:27.521,0:21:31.658
is the one they claim it is running, so[br]that's a very difficult problem to solve.

0:21:31.880,0:21:35.320
And even if it were the case,[br]where do you deploy yourself

0:21:35.600,0:21:38.521
a Google-like architecture,[br]or a Facebook-like architecture?

0:21:38.800,0:21:40.000
You simply can't.

0:21:40.400,0:21:42.371
It is no longer enough to just say

0:21:42.521,0:21:45.200
"We just need to do[br]some software development,

0:21:45.320,0:21:47.802
we just need to make it better[br]than the alternative."

0:21:48.201,0:21:51.600
There is a real tricky combination between[br]software development

0:21:51.819,0:21:55.881
and software deployment which[br]not easy to see how to fix it.

0:21:56.380,0:21:59.520
At least for me, it's very ???

0:21:59.620,0:22:01.043
So, what about distros?

0:22:02.139,0:22:06.280
We are distro people, doing one [br]of the most popular distros in existence.

0:22:06.320,0:22:09.248
Are we winning or are we losing[br]in this situation?

0:22:09.400,0:22:13.165
How are we doing in terms of our efforts?

0:22:14.074,0:22:16.320
In a sense, we are very much winning.

0:22:17.521,0:22:20.800
A lot of our work is being used[br]to deploy those infrastructures.

0:22:21.320,0:22:24.080
A lot of the infrastructure[br]of the big companies are deploying

0:22:24.200,0:22:27.560
on top of free software, if not direct [br]on top of our very own systems,

0:22:27.640,0:22:31.560
maybe modified here and there where[br]they need to make things better

0:22:31.560,0:22:34.680
as it is their own right[br]given it's all free software.

0:22:35.400,0:22:36.680
In that sense, we're winning.

0:22:37.137,0:22:38.480
We're increasing market share,

0:22:38.642,0:22:40.920
??? are being used a lot[br]to make infrastructure.

0:22:41.400,0:22:46.240
But we are also losing in the sense that[br]we are really not empowering users

0:22:46.240,0:22:48.560
to be in control of [br]their own computations.

0:22:49.000,0:22:53.401
If our final users are the sysadmin[br]that are running those infrastructures,

0:22:53.720,0:22:55.202
for them we are doing great.

0:22:55.360,0:22:56.920
We are making them be sure

0:22:56.920,0:22:59.041
they are in control[br]of their own infrastructure.

0:22:59.320,0:23:01.640
But for the final users of those services,

0:23:01.960,0:23:04.562
we are really not empowering them[br]at the moment.

0:23:05.080,0:23:07.241
So what I call the free software dark ages,

0:23:07.280,0:23:13.241
which is an expression I actually borrowed[br]from Bradley Kuhn and I find it quite inspiring,

0:23:13.520,0:23:17.760
is a situation in which we win [br]on the end user market

0:23:17.760,0:23:21.681
so every single device out there[br]in the hand of people − desktop,

0:23:21.681,0:23:26.000
laptop, even smartphones where right now[br]we are not doing very well −

0:23:26.160,0:23:28.360
all of this is running free software.

0:23:28.440,0:23:29.720
All of that is running Debian.

0:23:30.080,0:23:34.600
So, total world domination as[br]we were talking about a long time ago.

0:23:34.760,0:23:39.080
But all interesting computations,[br]all the final user application

0:23:39.201,0:23:42.920
which is being used to bring on[br]with your digital life,

0:23:43.160,0:23:46.801
are no longer happening on your devices,[br]happening far away from you

0:23:47.040,0:23:50.363
on computer you do not control,[br]sometime with free software,

0:23:50.363,0:23:51.655
sometime with non free software.

0:23:52.001,0:23:54.126
But in any case, outside[br]of your own control.

0:23:54.440,0:23:58.493
In a sense, this is very worrysome for me[br]because we have this euphoria of saying

0:23:58.493,0:24:00.166
we are very popular.

0:24:00.246,0:24:06.600
We are winning the war − we were using a lot[br]of this war-like terminology when I started.

0:24:06.600,0:24:11.680
But the war we are winning seems to become[br]increasingly pointless

0:24:12.080,0:24:15.522
because it's not being useful to actually[br]empower users to be in control

0:24:15.800,0:24:17.240
of their own computation.

0:24:18.401,0:24:23.601
To make things worse, there seems to be[br]some cultural problems that might be

0:24:23.601,0:24:28.441
just a perception of mind, maybe I'm being[br]too pessimistic, but it seems to me that,

0:24:28.441,0:24:31.520
as developper communities,[br]as hacker communities,

0:24:31.640,0:24:34.440
we are becoming way more lenient,[br]way more lax

0:24:34.960,0:24:38.281
about the lack of control on the tools and[br]on infrastructure we use

0:24:38.600,0:24:39.760
to make free software.

0:24:40.280,0:24:43.880
More and more often we see free software[br]developed on non-free infrastructure,

0:24:43.880,0:24:47.280
meaning infrastructures which are built[br]using non-free software

0:24:47.360,0:24:50.600
and which are anyhow centralized[br]in the hand of a few hosters.

0:24:53.670,0:24:55.640
The new generation of developpers[br]which is coming up

0:24:55.720,0:24:57.080
seems to be totally fine with that.

0:24:57.320,0:25:00.520
I'm not gonna argue this point in much detail,[br]there is a great essay by Mako

0:25:00.880,0:25:03.880
that I encourage all of you to read,[br]"Free software needs free tools",

0:25:04.200,0:25:05.840
which actually make couple of points.

0:25:05.881,0:25:09.360
One is that by using non-free software[br]to make free software,

0:25:09.360,0:25:11.081
we are sending out a very bad message.

0:25:11.160,0:25:13.814
We are telling to the world that[br]free software is good for you,

0:25:13.814,0:25:16.480
that's why we are developing it,[br]but it's not good for us

0:25:16.561,0:25:18.480
because we are using non-free tools[br]to make it.

0:25:18.801,0:25:21.841
That's the kind of "catch 22"[br]in our advertising message,

0:25:21.920,0:25:26.120
but it's also making the software[br]we are creating indirectly less free,

0:25:26.401,0:25:29.880
because if the favorite way to contribute[br]to that free software

0:25:29.961,0:25:32.200
is using some non free infrastructure,[br]some non free tools,

0:25:32.320,0:25:35.840
indirectly we're making people [br]that only want to use free software

0:25:36.040,0:25:38.600
less apt to contribute to that software.

0:25:38.920,0:25:41.241
So I really recommend reading that essay.

0:25:41.560,0:25:44.800
But also technically, we are going back[br]to a sort of a cage problem,

0:25:44.920,0:25:49.640
which is also a problem which is called [br]"the problem of the bug that noone can fix"

0:25:49.680,0:25:53.640
by the FSF I think, and essentially[br]we're creating software stacks

0:25:53.680,0:25:57.400
in which some part of it is entirely[br]free software, that we can debug

0:25:57.481,0:26:02.320
and some other parts are non free software[br]or software run by someone else,

0:26:02.600,0:26:05.160
so we have lost the ability to debug[br]the full stack.

0:26:05.960,0:26:11.000
When I was starting to learn programming,[br]this idea that I could debug everything

0:26:11.280,0:26:14.801
from the end user I was writing myself[br]for an assignment

0:26:14.920,0:26:17.709
down to the kernel level[br]was just exciting for me.

0:26:18.040,0:26:20.520
We seem to be losing sight of this,[br]a little bit.

0:26:21.800,0:26:25.360
As a second cultural problem,[br]we seem to be losing sight of

0:26:25.720,0:26:29.160
how much help we could get from[br]the legal system

0:26:29.492,0:26:32.720
and from new legal solution that[br]we might be in need of finding.

0:26:33.041,0:26:38.361
An example of that is the post open[br]source software "POSS" debate

0:26:38.640,0:26:40.720
which some of you might have run into.

0:26:41.000,0:26:43.920
That's a debate which actually observes[br]that the new generation of

0:26:44.160,0:26:48.008
free software developpers actually[br]don't care about licenses.

0:26:48.523,0:26:51.363
They just want to kick out their code,[br]just put it on GitHub,

0:26:51.643,0:26:54.963
not declaring their license at all[br]and they're just fine with that.

0:26:55.363,0:27:00.883
They want to be ??? to have[br]the hassle of deciding first of all a license,

0:27:01.208,0:27:04.403
second of all also some governance[br]model for their projects.

0:27:04.483,0:27:09.163
They just want to be hacking and doing,[br]and not caring about those annoying details.

0:27:10.163,0:27:15.523
This could be intervetedly interpreted[br]in positive ways like says that

0:27:15.763,0:27:22.523
we want the right to work on the code and[br]to do whatever we want with that by default.

0:27:22.763,0:27:25.483
We do not want to be expliciting[br]which kind of rights we give and

0:27:25.643,0:27:27.603
that's a very positive interpretation[br]of this phenomenon.

0:27:28.043,0:27:32.083
But in the end, for now, it is creating[br]a huge bunch of code that

0:27:32.123,0:27:35.243
we could not use as free software yet.

0:27:35.443,0:27:38.643
For instance we cannot include in Debian[br]something that does not have a license at all.

0:27:39.523,0:27:43.043
A second example is the debate about[br]the non-freeness of AGPL.

0:27:43.723,0:27:49.923
If you look up the history of free software,[br]there is argument that GPL itself is not free.

0:27:50.003,0:27:52.723
It's an argument that was being used[br]twenty years ago

0:27:52.763,0:27:56.963
when the battle between copyleft and[br]liberalizing was very high, was very harsh.

0:27:57.403,0:27:59.402
And it's just recurring again.

0:27:59.843,0:28:04.123
So maybe for some syntactically[br]interpretation of our own guidance,

0:28:04.163,0:28:08.123
we could make the point that something[br]like the AGPL is non-free, maybe.

0:28:08.603,0:28:12.883
But the point is that the way we distribute[br]software to final users is really changing.

0:28:13.443,0:28:18.123
Twenty years ago or fifteen years ago,[br]the main way to enable some user to use

0:28:18.363,0:28:21.283
a piece of software was actually to make[br]a copy of that software and

0:28:21.283,0:28:25.163
give it to him or to her via the network[br]or some media.

0:28:25.563,0:28:29.163
And all those conveying,[br]that kind of conveying software is clearly

0:28:29.163,0:28:34.483
distribution and that kind of activity used[br]to trigger some sort of license clauses.

0:28:34.723,0:28:38.443
These days, a software is no longer[br]distributed that way, in large parts.

0:28:38.729,0:28:44.683
It's being used over the net and something[br]like the AGPL is the equivalent of triggering

0:28:44.803,0:28:48.203
some licensing condition via the main way[br]of distributing,

0:28:48.243,0:28:50.123
of giving access to some software.

0:28:50.603,0:28:53.453
I want to enter in details in this debate.

0:28:53.493,0:28:56.283
Those are just examples, for me they are[br]examples of the fact that

0:28:56.283,0:29:01.163
we are kind of losing faith in how much[br]the legal system and free software

0:29:01.203,0:29:02.403
are intertwined.

0:29:02.922,0:29:06.435
And this actually mixes very badly[br]with the situation in which

0:29:06.435,0:29:10.324
users are losing control because those[br]computations are moving away from them.

0:29:10.683,0:29:14.243
I think this situation, in general,[br]is not going to fix themselves

0:29:14.608,0:29:18.403
and we, as distribution people,[br]have a role to play in fixing it.

0:29:19.723,0:29:25.563
What could be a role for Debian in all this[br]computing situation we have these days.

0:29:27.813,0:29:31.883
The common trend in the so called cloud[br]seems to be that computations

0:29:31.925,0:29:33.844
are moving away from user devices.

0:29:34.643,0:29:37.763
We cannot just say[br]"Well just don't use anything cloudy",

0:29:38.043,0:29:40.083
because it is convenient, people will want[br]to use that.

0:29:40.243,0:29:41.563
We need to do something different.

0:29:41.921,0:29:46.291
As distribution people, we could do a lot,[br]I think, and I have a couple of thoughts

0:29:46.564,0:29:53.683
to share with you that are different[br]depending on the so called service model

0:29:53.683,0:29:54.643
of the cloud.

0:29:54.803,0:29:57.521
One of the first service model of the cloud[br]you might have heard about is

0:29:57.685,0:30:00.643
"Infrastructure as a Service" (IaaS) where[br]essentially you have servers that

0:30:00.849,0:30:05.643
give virtual machines to people and[br]essentially you get to administer

0:30:05.763,0:30:08.403
your own machine which is a virtual machine[br]on a virtual machine server

0:30:08.683,0:30:10.206
controlled by someone else.

0:30:10.444,0:30:14.243
This is potentially very good for people[br]because it is lowering the barrier

0:30:14.284,0:30:16.323
you need to have your own server.

0:30:16.643,0:30:21.083
When I first set up my own server[br]with friends, at the end of the 90's,

0:30:21.324,0:30:24.883
we had to buy some machine, to find[br]someone kind enough to host it,

0:30:25.003,0:30:27.523
pay the hosting fees and so on[br]and so forth.

0:30:27.723,0:30:31.963
It was something that was by far not at all[br]accessible to the random user.

0:30:32.323,0:30:37.045
These days, a lot of people can simply go[br]to some virtual machine provider, rent

0:30:37.283,0:30:42.044
a virtual machine with one-click button and[br]they have their own machine to administer.

0:30:42.364,0:30:46.164
Maybe they don't have the skill to[br]administer it, that's a different problem,

0:30:46.332,0:30:50.403
but you are definitely lowering the barrier[br]to access, to have you own server

0:30:50.682,0:30:52.563
and do your own remote computation.

0:30:53.083,0:30:56.363
As Debian, we are doing pretty well[br]in this area, I think.

0:30:56.483,0:31:00.206
We're offering technology like OpenStack[br]and other competitors of OpenStack,

0:31:00.403,0:31:04.603
which seems to be the market leader on[br]that market which are entirely free software.

0:31:04.884,0:31:09.963
But I think we should be investing more in[br]offering a trivial deployment experience

0:31:10.211,0:31:11.364
for Debian users.

0:31:11.403,0:31:15.765
We should make trivial for people[br]to have their own virtual machine servers.

0:31:15.765,0:31:19.203
If they are not computer geeks, they should[br]be able to flock together friends

0:31:19.443,0:31:25.523
which have system administration ability[br]and have their own local IaaS

0:31:25.723,0:31:30.843
and have their own virtual machine without[br]having to rely on big hosters provided

0:31:30.923,0:31:33.043
virtual machines to everyone in the world.

0:31:33.444,0:31:36.523
This is a great step to our autonomy.

0:31:36.603,0:31:40.683
As Debian, what is the best deployment[br]experience we can offer for people

0:31:40.724,0:31:43.323
that want to setup their own virtual[br]machine servers.

0:31:44.163,0:31:48.124
Then, there is another service model which[br]is called PaaS, "Platform as a Service".

0:31:48.163,0:31:51.724
This is a kind of service model in which[br]essentially you have hosters

0:31:51.724,0:31:55.724
of application engines, you develop[br]application targeting

0:31:55.724,0:31:58.723
specific application engine.

0:31:59.273,0:32:02.462
Sorry, application servers. You target[br]specific application servers.

0:32:02.462,0:32:04.813
An exemple of this is Google App Engine.

0:32:05.683,0:32:11.003
I think in some sense this service model[br]of the cloud is mostly orthogonal to

0:32:11.003,0:32:15.123
what we do as a distribution because either[br]you're using a full fledge distribution

0:32:15.363,0:32:19.243
and you do your own system administration,[br]or you are developping an application

0:32:19.363,0:32:22.403
for a specific application server and[br]you rely on someone else

0:32:22.403,0:32:23.963
to do that administration.

0:32:24.323,0:32:28.003
So, yes, I think it's mostly orthogonal[br]to what we do, but might also be

0:32:28.117,0:32:31.810
a symptom that there is a reject from the[br]application developper community,

0:32:31.883,0:32:36.723
a reject from the way they can target[br]distributions like Debian.

0:32:37.123,0:32:40.523
So if it is very difficult to have your own[br]application running properly on Debian

0:32:40.803,0:32:45.523
because we have old software, because we[br]change libraries, because we do not accept

0:32:45.802,0:32:48.483
multiple copies of the same libraries and[br]so on and so forth,

0:32:48.683,0:32:51.364
if it is too difficult for application[br]developpers to target Debian,

0:32:51.729,0:32:57.363
they might be more and more tempted[br]to target applications servers like PaaS.

0:32:57.763,0:33:02.363
So there might be something we could do[br]about this, here, like finding better synergies

0:33:02.763,0:33:06.443
between containerization technology,[br]we have some work being done in Debian,

0:33:06.443,0:33:12.251
and the way we usually develop some,[br]we usually maintain a distribution.

0:33:12.251,0:33:14.403
There might be something we could do[br]about this here.

0:33:14.733,0:33:18.881
Oh, and I didn't mention this, but I have[br]no specific answer to give to you,

0:33:18.923,0:33:22.683
just a train of thoughts I wanted to share[br]with you and what we could do

0:33:22.723,0:33:23.963
to improve the situation.

0:33:25.244,0:33:29.563
The final service model we have in the[br]cloud, which is I think worrysome

0:33:29.563,0:33:34.723
from the point of view of user control,[br]is SaaS, "Software as a Service".

0:33:35.043,0:33:39.283
There, essentially your own device,[br]your own computer only is thin client

0:33:39.664,0:33:43.563
and rely entirely on a remote server[br]to do your own computation.

0:33:43.843,0:33:48.208
We are back to the mainframe / thin client[br]distinction of the early days of computing

0:33:48.563,0:33:52.963
and here, there is a lot we could do,[br]I think, but also a lot we could not do.

0:33:53.290,0:33:55.963
Here, most of the work should come[br]from upstreams.

0:33:56.283,0:33:59.883
We need better free software and federated[br]replacement for

0:34:00.003,0:34:05.043
popular centralized proprietary applications[br]in which users can participate

0:34:05.203,0:34:07.843
in some kind of network by using[br]their own node.

0:34:08.203,0:34:12.763
This is work that should not come from[br]distribution itself, it should really come

0:34:12.763,0:34:14.843
from application developpers upstream.

0:34:15.203,0:34:18.083
But still, there are useful things[br]we could do here.

0:34:18.643,0:34:21.083
We already have a lot of building blocks.

0:34:21.523,0:34:26.883
We have stuff like Owncloud, Git-annex,[br]mediagoblin, pump.io, Yacy.

0:34:26.963,0:34:31.123
We have a lot of good building blocks,[br]most of them are not yet up to par

0:34:31.203,0:34:36.243
with the centralized proprietary equivalent,[br]but I'm confident we could get there.

0:34:36.604,0:34:42.883
What we lack is the equivalent ease of[br]deployment of these services on user machines.

0:34:43.385,0:34:50.003
In some sense, if we have democratized[br]the installation of software twenty years ago

0:34:50.203,0:34:56.643
with distributions, these days, to face the[br]challenge of control of our own computation,

0:34:56.883,0:35:00.723
we need to make it as easy as using a[br]package manager to install

0:35:01.003,0:35:04.043
your own nodes using those applications.

0:35:04.443,0:35:12.483
Ideally, everyone in the world without[br]nothing more than basic computer user skills

0:35:12.763,0:35:17.323
should be able to have its own machine[br]at home doing some anonymous browsing,

0:35:17.563,0:35:21.363
doing some mail handling, doing web hosting,[br]doing storage calendar,

0:35:21.563,0:35:24.323
doing encrypted peer to peer backup,[br]and so and so forth.

0:35:24.723,0:35:28.923
I'm maintaining my own mail server and it is[br]a user ???, I struggle myself

0:35:29.243,0:35:33.443
to keep up with the need of knowledge and[br]of surveillance that I need to make

0:35:33.523,0:35:39.203
to my own mail server to be able to run it[br]properly and I get blacklisted

0:35:39.243,0:35:41.562
from time to time from providers and[br]it's a pain.

0:35:41.923,0:35:47.091
Something that no one without having at least[br]some basic system administration ability

0:35:47.091,0:35:48.723
could do properly.

0:35:49.043,0:35:51.843
This is the thing we need,[br]the nut we need to crack.

0:35:52.163,0:35:55.963
We need to empower everyone out there[br]to have its own computer with

0:35:56.003,0:35:58.243
its own node of those services.

0:35:58.323,0:36:00.483
Of course, you are all thinking of[br]the FreedomBox now.

0:36:00.843,0:36:06.363
That's a great example of a project who[br]wants to tackle precisely that problem.

0:36:06.803,0:36:11.003
It's a project that's been announced by[br]Eben Moglen a few years ago at a Debconf

0:36:11.083,0:36:12.683
if my memory serves me well.

0:36:13.043,0:36:17.443
It's heavily based on Debian and it's doing[br]exactly that.

0:36:18.283,0:36:21.523
But my question from the Debian[br]point of view is:

0:36:21.603,0:36:25.123
maybe this project should not only be[br]a spin-off of Debian,

0:36:25.283,0:36:29.723
should not only be a derivative distribution[br]of Debian,

0:36:29.848,0:36:33.883
maybe we should think at making something[br]like this a first class citizen in Debian.

0:36:34.243,0:36:38.603
I don't know exactly what that means yet,[br]it's something we could think about

0:36:38.603,0:36:41.328
having the main administration interface[br]for Debian something

0:36:41.363,0:36:43.203
that targets these specific scenarios.

0:36:43.203,0:36:47.843
We could generalize that, we do not need[br]to target only specific plug devices

0:36:48.243,0:36:51.603
because people at home might have desktop[br]computers, might have media center.

0:36:51.763,0:36:55.924
They might want something like the[br]FreedomBox at home and

0:36:55.963,0:36:57.523
collaborate with others immediately.

0:36:57.963,0:37:02.443
My point here is that if our mission back[br]in the days was to

0:37:02.683,0:37:06.843
democratize free software by making it[br]easier to install software

0:37:06.949,0:37:10.483
on your machine, today our mission is to[br]democratize free software by making it

0:37:10.483,0:37:15.815
trivial to install some node of some[br]federation of free services on your machine.

0:37:18.243,0:37:20.963
Another thing we could do,[br]it is the last one for me today,

0:37:21.323,0:37:23.882
is to step in the free service debate.

0:37:24.323,0:37:28.244
When I started looking up these arguments[br]a few years back, I was surprised by

0:37:28.244,0:37:33.323
the fact that it's still not clear what[br]it does mean to be a free service.

0:37:34.723,0:37:37.243
When I started working on free software[br]fifteen years ago,

0:37:37.403,0:37:39.735
it was fairly clear what does[br]free software mean.

0:37:39.923,0:37:43.483
Sure, it was some terminology debate[br]between free software and open source

0:37:43.523,0:37:44.643
which still exists today,

0:37:44.926,0:37:48.443
but the basic freedoms, the basic rights[br]you should have to call something

0:37:48.511,0:37:51.083
free and open source was fairly clear.

0:37:51.443,0:37:54.643
That kind of intellectual debate had[br]already happened at the time.

0:37:54.963,0:37:59.490
Today, where the problem of computations[br]moving away from indivual user

0:37:59.770,0:38:02.643
is raging, there is no clear consensus[br]on that matter.

0:38:03.083,0:38:07.603
There is some great work, for instance[br]there is the Franklin Street statement on

0:38:07.603,0:38:10.445
free network service,[br]I think that's a full ???,

0:38:10.643,0:38:16.283
dating back to 2008, six years ago, in[br]which you find a lot of very useful

0:38:16.763,0:38:21.963
recommendations for users, for software[br]developpers and for system administrators

0:38:22.043,0:38:27.364
to make sure that you maximize your control[br]over your own computation on the network,

0:38:27.563,0:38:32.483
but they take no stance on what it does mean[br]to be a free service.

0:38:33.603,0:38:37.883
Is it enough to have something which is free,[br]do you need more specific license.

0:38:37.964,0:38:40.883
There are some recommendation[br]on that point, but still,

0:38:41.123,0:38:42.923
there are no clear answers[br]to this question.

0:38:45.083,0:38:50.803
There is another work by RMS in 2010[br]about Software as a Service or

0:38:51.123,0:38:53.163
"service as a software substitute"[br]as he calls it.

0:38:53.483,0:38:57.333
Here, essentially what you have is a main[br]recommandation about

0:38:57.403,0:39:00.844
not using Software as a Service at all.

0:39:01.363,0:39:04.724
Essentially there is a recommandation of[br]doing your own computation

0:39:04.763,0:39:06.243
on your own machines.

0:39:06.684,0:39:11.203
I think that might be a generally good[br]recommandation but it's not gonna scale,

0:39:11.646,0:39:14.483
it's not gonna be enough in my opinion[br]to convince people

0:39:14.525,0:39:16.603
not to use very convenient services.

0:39:16.923,0:39:21.123
Think we need more gradual and blurry[br]lines saying, encouraging people

0:39:21.271,0:39:25.644
to keep computation closer to them,[br]to rely on federation of friends of people

0:39:25.843,0:39:27.484
to do computation together.

0:39:27.563,0:39:31.371
And we, as distribution people, could[br]make easier for them to do so.

0:39:31.883,0:39:36.323
And then there is another work which is[br]"Network Services Aren't Free or Nonfree"

0:39:36.763,0:39:40.723
which is a couple of years later, still by RMS,[br]which essentially tries to walk the fine line

0:39:40.963,0:39:44.083
between what's the difference between[br]a pure service, so a service that

0:39:44.203,0:39:47.643
just for instance convey messages,[br]as opposed to a service which does

0:39:47.643,0:39:51.363
computation that could have been[br]done instead on your machine.

0:39:51.523,0:39:55.724
That's a very fine line to work, it's very[br]difficult to stay there and

0:39:55.724,0:40:01.049
what we might need there is a strong[br]opposition, actually, and we should try

0:40:01.291,0:40:05.203
to replace everything which is centralized[br]with federated equivalent and say that

0:40:05.203,0:40:08.945
we as free software people and distribution[br]people should work in that direction.

0:40:09.963,0:40:11.284
So what we could do in Debian.

0:40:11.284,0:40:13.611
Well, I think we should try to step[br]in this debate.

0:40:13.875,0:40:20.283
Surprisingly for me, we still have no clear[br]answer to what it means to be a free service

0:40:20.283,0:40:24.283
today and we have quite a bit of[br]experience in Debian

0:40:24.323,0:40:27.203
in leading debates in free sotfware.

0:40:27.203,0:40:29.764
We have created the DFSG which is being[br]used as an example for

0:40:29.964,0:40:34.123
many other communities, we have participated[br]in the GPLv3 discussion for instance.

0:40:34.283,0:40:39.124
Our decisions in terms of free license[br]are looked up by other projects.

0:40:39.363,0:40:43.003
So we might have the authority and[br]the reputation to step in this debate

0:40:43.211,0:40:46.569
and we also have a lot of technical[br]knowledge in the area.

0:40:47.043,0:40:52.803
Being a distribution commited to free software,[br]we know a thing or two not only about

0:40:53.043,0:40:58.763
software freedom, but also about how you[br]deploy software, how difficult it is

0:40:58.763,0:41:01.524
and how difficult it should be for people[br]to deploy free software.

0:41:01.843,0:41:06.083
So I think we are in just the sweet spot[br]to actually enter this debate

0:41:06.523,0:41:10.604
with the needed authority and make[br]a contribution to actually help people

0:41:10.803,0:41:13.566
realize what it means today[br]to use a free service.

0:41:14.963,0:41:16.283
The concluding question[br]I have for you is

0:41:16.523,0:41:21.324
"What's Debian take today[br]on liberating users?".

0:41:21.763,0:41:26.384
Would we be happy enough to have Debian[br]on every machine in the world

0:41:26.565,0:41:29.324
if people are using completely[br]remote services?

0:41:29.684,0:41:33.763
And if we were not, what should we do,[br]what should we be working on to change

0:41:34.003,0:41:37.963
that future which seems very much[br]the future that we have at hand.

0:41:39.524,0:41:47.083
Pictures are gone, so there was a cloud[br]on the left,

0:41:47.643,0:41:50.204
there was Debian here and a sun here.

0:41:50.523,0:41:54.364
LaTeX, beamer or Tikz or something[br]is playing tricks on me.

0:41:55.004,0:41:58.683
So that's all I have for you, I hope[br]I've given you some food for thoughts

0:41:58.924,0:42:03.324
for this week and if you have any question[br]or comments in these topics,

0:42:03.523,0:42:05.084
I'm very much happy to hear about that.

0:42:05.604,0:42:06.483
Thank's a lot.

0:42:06.643,0:42:15.563
[applause]

0:42:23.123,0:42:25.923
There seems to be a mic which is floating[br]around down there.

0:42:43.564,0:42:49.963
[Q] ??? quite a lot and quite brilliantly[br]about what cloud computing buzzwords

0:42:50.203,0:42:57.603
mean for free software, but I think what important[br]battle we are actually losing is ???

0:42:57.687,0:42:58.723
in the minds of people.

0:42:58.924,0:43:02.243
[Q] Why is it young developpers or[br]newcommers to free software

0:43:02.484,0:43:05.603
don't care about software being free?

0:43:07.043,0:43:11.603
[Q] Why don't they care about using non free[br]tools, why don't they care about

0:43:11.843,0:43:17.683
which license declare for their software[br]if any license is at all? and so on.

0:43:18.043,0:43:22.883
[Q] You mention that problem, but what do[br]we do about it? Do you have any ideas?

0:43:25.403,0:43:28.843
[Zack] Well, a friend of mine we asked[br]a similar question I think once answered

0:43:29.283,0:43:32.003
"What could they say more that[br]'Oh those young kids' ".

0:43:32.803,0:43:37.123
So, I don't know, maybe it's our fault,[br]maybe we have failed as a generation

0:43:37.163,0:43:40.883
to convey the importance that being[br]in control of our own computation had,

0:43:41.164,0:43:45.923
or maybe it's just that the public that[br]is open to coding and

0:43:45.923,0:43:49.643
hacking is much larger than in the past so[br]we are reaching out other communities.

0:43:50.331,0:43:54.523
It's very good for them to be coding because[br]I think every citizen in the world need

0:43:54.523,0:43:58.283
to have basic knowledge of coding to[br]understand what's happening in the world,

0:43:58.643,0:44:02.363
but maybe they just have different mission[br]than we had in the past.

0:44:04.969,0:44:07.403
So, very good question, I don't have[br]a very good answer, sorry.

0:44:10.803,0:44:11.923
[Q] Hello.

0:44:17.513,0:44:22.044
Thank you so much for the wonderful talk,[br]I think it's great to talk about these

0:44:22.284,0:44:29.403
political issues and I see there's a challenge[br]between the sort of very individual focus

0:44:29.563,0:44:34.166
of each person being able to use their own[br]computer as the wish which has its own values,

0:44:35.003,0:44:39.603
but there's a different sort of value that[br]relates to power structures in general.

0:44:40.003,0:44:46.163
So, we're talking about not just how free[br]is each individual person but whether

0:44:46.483,0:44:50.567
an entity like Twitter, Google or Facebook[br]or some these other services

0:44:50.603,0:44:57.164
is a very powerful entity that has power over[br]the majority of us who use their services.

0:44:57.884,0:45:02.963
And so, I wonder if and I'd like your[br]thoughts on thinking about it

0:45:03.083,0:45:09.764
less as a "Is this software free?" but[br]about "Who is in power in the community?"

0:45:09.843,0:45:16.371
and so in a democratic sense, you could have[br]the community that builds the tools together

0:45:16.371,0:45:23.404
as government structures or as mechanisms[br]for handling power that make the power

0:45:23.404,0:45:28.723
bottom-up and more democratic and maybe[br]that's more important than

0:45:28.723,0:45:31.443
the technical status of each[br]individual user.

0:45:32.563,0:45:36.449
[Zack] So, as a concerned citizen and also[br]as a political activist,

0:45:36.564,0:45:39.011
I very much share your concern.

0:45:39.403,0:45:46.484
I think we need to focus on what is in[br]reach on us as geeks in this circle

0:45:46.723,0:45:49.324
and have this kind of discussion[br]in a different circle.

0:45:49.607,0:45:54.803
So, as someone with activity in politics[br]and as a geek, I very much try

0:45:54.803,0:45:58.723
to actually explain to politicians and[br]to activists the role of

0:45:58.843,0:46:03.003
what we are doing here in very technical[br]ways and the impact that it has

0:46:03.083,0:46:04.363
on politics in general.

0:46:04.884,0:46:07.283
And I think the ??? the talk later on[br]this evening might have

0:46:07.363,0:46:09.163
a thing or two to say about that as well.

0:46:09.523,0:46:15.283
So from our part we need to understand it[br]in some sense even if

0:46:15.283,0:46:19.404
we advance a lot the status quo of user[br]control of technology

0:46:19.444,0:46:21.123
that we had thirty years ago.

0:46:21.403,0:46:24.483
We have also started to lag behind[br]many other areas.

0:46:24.803,0:46:28.804
Something that I wanted to mention before[br]but I fail to do so is that

0:46:28.843,0:46:32.724
when I was doing my computing in the[br]nineties, a lot of computations

0:46:32.843,0:46:34.803
were mediated by clearly defined[br]protocols.

0:46:35.403,0:46:40.083
So we had RFCs or equivalent documents[br]by other organisations which were like

0:46:40.243,0:46:45.364
clearly marked paths to how to collaborate[br]technically on the internet

0:46:45.603,0:46:47.219
and how to make software talk together.

0:46:47.763,0:46:54.644
In a sense, that culture of interoperability[br]of protocols has actually started lagging

0:46:54.763,0:46:57.403
behind a lot with respect[br]to popular technology.

0:46:57.803,0:47:02.923
So stuff like social networks, most of them[br]except the good ones that free software guys

0:47:03.163,0:47:08.764
try to build like pump.io or like diaspora,[br]well all those technologies started up

0:47:09.003,0:47:12.203
without any kind of interoperability[br]in mind.

0:47:12.723,0:47:18.123
So technically I think we need to push[br]again on the direction of interoperability

0:47:18.165,0:47:21.523
of protocols, and that's a technical[br]contribution that we could do that

0:47:21.603,0:47:22.804
will have an impact.

0:47:22.883,0:47:26.683
You know, code is law, as Lessig was saying,[br]and that would have a technical impact

0:47:26.723,0:47:28.363
on the power structures you mention.

0:47:28.803,0:47:30.603
That's my thought on this matter.

0:47:33.403,0:47:35.563
[Q] I have an answer.

0:47:35.563,0:47:36.563
Hello.

0:47:36.604,0:47:40.083
I have an answer, sort of an answer[br]to the previous question.

0:47:40.325,0:47:46.563
This is of course the heart of the difference[br]between free software and open source.

0:47:47.043,0:47:50.683
The difference between free software and[br]open source isn't nothing at all

0:47:50.883,0:47:53.043
and it's not about licenses.

0:47:53.483,0:47:55.603
It's about goals and aims.

0:47:56.043,0:48:05.245
Over the past decades, many of us have[br]chosen not to pick a fight with

0:48:05.524,0:48:11.963
open source people just for an easy life and,[br]you know, it's always easy to have somebody

0:48:12.402,0:48:17.204
who might share some of your goals and[br]to be able to collaborate with them.

0:48:17.723,0:48:20.818
But less and less is it becoming[br]the case that

0:48:21.723,0:48:27.446
the goals of people who are doing open[br]source are the same as the goals

0:48:27.643,0:48:29.283
of people doing free software.

0:48:29.803,0:48:35.124
You can see that very clearly[br]in the responses from people like

0:48:35.124,0:48:37.003
Google to things like the AGPL.

0:48:37.683,0:48:39.563
And there are a lot of examples.

0:48:41.011,0:48:45.203
So, one of the things that we can do[br]to try and bring some of

0:48:45.363,0:48:49.243
the new crop of developpers along with us[br]is to actually make it

0:48:49.324,0:48:51.483
a bit more of a fuss about…

0:48:52.243,0:48:55.118
You know, let's not come ??? all Stallman[br]about that,

0:48:55.118,0:49:00.403
Stallman is not the best PR guy, but I think[br]Debian can do a lot better than he can

0:49:00.626,0:49:03.243
and we've probably got[br]a lot more credibility.

0:49:03.923,0:49:06.043
And individually, we have as well.

0:49:06.443,0:49:15.004
What we need to do is we need to explain[br]our vision to those new developpers

0:49:15.244,0:49:21.843
who mostly are just being, you know, they[br]see an open source marketing machine

0:49:22.123,0:49:23.363
and we are something different.

0:49:25.583,0:49:26.485
[Zack] Thanks.

0:49:26.485,0:49:30.118
So there's not need to be questions and[br]answers, so if you have comments, feel free.

0:49:31.433,0:49:34.323
[Talkmeister] I think we're running short of[br]time and we need to take one more question.

0:49:35.003,0:49:37.843
So maybe one last or, Stefano, one last?

0:49:38.403,0:49:38.963
[Talkmeister] We can.

0:49:39.843,0:49:41.483
Ok, one last question or comment?

0:49:41.803,0:49:43.684
[Q] Just a quick comment if I may.

0:49:43.963,0:49:48.603
You talked about federated services and[br]facebook and dropbox and that sort of thing.

0:49:49.163,0:49:55.683
I think maybe the issue here is less about[br]federated services but is about identity.

0:49:56.564,0:50:02.043
If I have my own dropbox alike and you have[br]your own dropbox alike,

0:50:02.197,0:50:05.730
the problem is not that the two couldn't[br]talk to each other,

0:50:05.923,0:50:11.524
we have no way of negotiation of identity[br]authentication, access kind of problem.

0:50:11.923,0:50:14.203
I think maybe part of the answer to your[br]question is

0:50:14.323,0:50:18.043
"Can we come up with some way of allowing[br]federated identity management

0:50:18.243,0:50:22.243
for people in general and just us say".

0:50:23.163,0:50:27.363
[Zack] I think this is very much related[br]to what I was answering before to Aaron,

0:50:27.723,0:50:28.963
in the sense "yes we could".

0:50:29.405,0:50:33.486
We have shown in the past that we can[br]come up with very smart protocols

0:50:33.486,0:50:36.603
that allow people to technically[br]interoperate over the net.

0:50:36.963,0:50:40.644
But we are coming to late for that.

0:50:40.923,0:50:44.972
Those big entities which now have the power[br]to attract a lot of users to them

0:50:45.244,0:50:49.123
developped before those standard[br]that we could have used to make

0:50:49.404,0:50:53.163
smaller entities interoperate could[br]have been put in place.

0:50:53.529,0:50:57.243
So yes, I agree with you, there is technical[br]work to be done but in some sense

0:50:57.243,0:51:01.883
we are late in doing that work and[br]the question now is not only

0:51:02.124,0:51:05.923
"How could we do the technical work that[br]allows us to have smaller entities

0:51:05.923,0:51:10.203
that interoperate for authentication or[br]everything else?" and also

0:51:10.483,0:51:15.563
"How do we migrate from the status quo to[br]the ideal world that would be possible

0:51:15.603,0:51:17.643
if those standards existed[br]in the first place?".

0:51:18.206,0:51:22.483
So in a sense I think we are a bit late[br]and we have twice the work to be done

0:51:22.763,0:51:26.083
before reaching the optimal and more[br]federated situation which I think

0:51:26.243,0:51:27.443
would solve the problem.

0:51:29.963,0:51:31.444
So, thanks a lot.

0:51:31.603,0:51:39.833
[applause]