[00:00]
For those of you who don't know me, I'm Chris Foster, I'm 
president of the Free Software Society and I thank you for coming 
today. We are absolutely completely honored to have the executive 
director of the Free Software Foundation, an active member in the 
software development community and the boss of Free Software 
bosses pretty much, Mr Bradley Kuhn. [applause]

[00:28]
So... This talk is probably different, than most of the talks that 
happen here in any kind of academic setting. It tends to be the 
case that in computer science and in the sciences in general, 
people tend to avoid political and ethical issues surrounding the 
field. And in the Free Software community, for decades now, we 
have focused on the ethical and political implications of 
software. These days a lot of people focus on  the software that 
our movement produced, the free software that exist embodied today 
in GNU+Linux systems or the various BSD systems that are 
available, they think about it in terms of its reliability, its 
dependability, and how functional it is. But we continue in the 
Free Software Movement to focus on the ethical and political 
implications of the software and what rights people have. In fact 
we believe that the issues of technology and how good technology 
is and how functional it is are ultimately secondary to the 
questions of what kind of community is surrounding that 
software. And we focus on the idea of augmenting the community of 
software users and programmers, and fostering cooperation and good 
will between programmers and users alike. And that continues to be 
the issue we focus most on.

[02:02]
What I want to demonstrate to you tonight through some personal 
anecdotes, through some philosophical consideration and through a 
little bit of historical perspective about the movement, is to 
show you how Free Software can build and create cooperating 
communities of software developers and users and how by direct 
contrast proprietary software, the functional opposite of Free 
Software, tears down and impedes the creation of cooperating 
communities. Unfortunately, because English has a bug in it, I 
have to start by defining terms. It's always unfortunate when you 
have to start a conversation by defining your terms.  Well, 
English has more words than any other language on the planet, it 
still only has one adjective for two very different concepts. When 
we're talking about free software we're talking about free as in 
freedom not necessarily in price.  So think about free speech, not 
free beer. I know that's hard for college students to think about 
and avoid the free beer aspect of what we do but for the evening I 
hope you'll focus on the ideas of free speech rather than free 
handouts or free beer. Other languages of course make this 
distinction clear, in Spanish there is the adjective "libre" and 
the adjective "gratis" which are referring to these two different 
types of free. We don't have that in English unfortunately.

[03:36]
So, Free Software is defined in terms of what freedoms you, a 
particular user, receive with the software. And it's always 
defined in the context of specific users because a given software 
system could be Free Software for one person and the exact same 
system might not be free as in freedom for another person.

[04:00]
The first freedom that you as a user get with Free Software is the 
freedom to run and use the program for any particular purpose that 
you can think of. This includes any purpose that the original 
programmer who wrote the software or the original distributor of 
the software might not even have considered. This freedom is often 
these day restricted even with proprietary software. It used to be 
the case that all software by and large gave you this freedom, but 
these days, you can get versions of Microsoft Frontpage (for 
example) that forbid you from creating any web pages with it that 
say negative things about Microsoft. So even that freedom to run 
and use the program for any purpose doesn't exist sometimes with 
proprietary software. With Free Software you always get that 
right.  You can use the software for anything you can think of, 
including exercising your free speech right to say how much you 
hate the Free Software Foundation or something like that.

[04:56]
The second freedom that you get is the freedom to help yourself by 
improving, changing, modifying or otherwise adapting the software 
to suit your own needs. And of course, if you are a software 
programmer, you can take direct benefit from this freedom. Because 
you understand how software works and you can look at the software 
source code (which, by the way, is a prerequisite for this 
freedom) because even making the most trivial adaptations or 
modifications to computer software without its source code -- the 
original "stuff" the programmer generated -- is virtually 
impossible. So all Free Software by prerequisite has to come with 
the source code, so that you can modify that source code to 
improve the software to suit your needs. This freedom, as I said, 
directly benefits programmers but it also indirectly benefits 
non-programmers -- people who don't know how to write 
software. This is because they are allowed to ask anyone they 
want, to modify or adapt the software on their behalf. And an 
interesting thing comes up with regard to Free Software in this 
case. You have this same freedom whether you choose to exercise it 
in a noncommercial sense or in a commercial sense. So -- if you 
have, rather, if you're a software user but not a programmer and 
you have a Free Software program that doesn't work quite right for 
you, it has a bug or is missing a feature that you need, you can 
ask a friend as an altruistic act to improve or fix the software 
for you. But you can also -- if you can't convince a friend of 
yours to fix it without paying them -- you can hire anyone you'd 
like on the competitive free market to fix the software. You can 
hire someone to do the work for you. So this freedom, and all the 
other freedoms I'm going to talk about, always work in both a 
noncommercial and a commercial sense. And the Free Software 
licenses that we've produced and ones that we've approved as 
things we think are Free Software licenses make sure that the 
noncommercial and the commercial freedoms are treated equally -- 
that you have an equal right to make money from free software and 
you haven an equal right to engage in various activities 
non-commercially. 

[07:13]
The third freedom that you get is the freedom to help other 
people, by sharing your software. This, again, can be done 
altruistically or as part of a busyness model. You can share Free 
Software. If your friend sees you running a copy of a free program 
you can give them a copy, and make a copy immediately for them and 
distribute it to them. All Free Software licenses give you this 
right. But you also have the freedom to go into the commercial 
business of sharing software. We have a specific name in our 
economic culture for commercial sharing of software. It's usually 
called "distribution" or "putting software for sale". With Free 
Software you can charge any price you would like. If you can get 
people to pay you that price for the software, you are permitted 
to sell it at that price. You can of course put it for sale at any 
price you like but you may not get an opportunity to sell a copy 
of gcc for $1,000,000.  (Although MontaVista seems to be able to 
get $15K a copy for the same stuff you can download from our 
website.) So there is substantial money that you can make from 
distributing Free Software. Red Hat, for example, early in their 
business model, made a lot of money selling copies of 
distributions of GNU/Linux, especially before high speed 
connections were common among their user base. They were able to 
sell box sets for around $40 or $50. And that price was set by 
what the market would bear and what they felt they could get for 
it. Free Software has no requirements on you about how much or how 
little you can charge for distribution. So that's your third 
freedom, you freedom to help other people by sharing the software.

[08:49]
The fourth and final key freedom that you get with Free Software 
is the freedom to help the community at large, effectively by 
combining the previous two freedoms. If you modify or adapt a Free 
Software program to suit your needs, you have the right, if it's 
Free Software, to then distribute that modified or improved 
version to other people if you'd like. Again, this can be done in 
an altruistic way or as a commercial endeavor. A lot of Free 
Software gets developed by people posting a patch to a mailing 
list, a discussion list about the development of a particular 
software program, and encouraging the original developers to 
include that patch or improvement into the typical version of the 
software. And incrementally the software gets better by people 
altruistically contributing changes. But that's not the only way 
you can help your community. You can also go into business -- 
selling you services as someone who can modify or improve Free 
Software or adapt it to suit the needs of your clients. So that's 
again a very important aspect (that I really can't stress enough_ 
that commercial and noncommercial endeavors are treated as fully 
equal by Free Software licenses. If it's Free Software you have 
both a noncommercial and a commercial right to copy, share, modify 
and redistribute the software as much as you'd like.

[10:15]
Now by contrast you can define proprietary software, which is the 
term we use for the "opposite" of Free Software, as software that 
restricts most or all of these freedoms. As I mentioned, sometimes 
the freedom to run or use the program is often restricted by 
various licensing terms that prohibit you from doing certain 
things with the software. The freedom to modify or improve the 
software is usually restricted in two different ways with 
proprietary software. First, the licensing agreement that you are 
required to sign or agree to upon opening the package (a legally 
dubious tactic, by the way) might say: "You are not permitted to 
modify or reverse engineer this software to figure out how it 
works." But they also put a very practical constraint on your 
ability to modify or improve the software. Software is a very 
special kind of copyrighted work in our culture. Most copyrighted 
works (in fact in pretty much all other copyrighted works except 
for software) come in their transparent form.  Until the days when 
we have the chips in our brain which you can download a whole 
novel or motion picture into your head, you actually have to 
experience the piece of literature or the motion picture up on the 
screen, in its [native] form actually presented under 
copyright. So if you read a novel, even if it's by William 
Faulkner and very dense, you actually get to see the words, 
word-by-word, that composes the literary work. If you go to see a 
film -- even if it's directed by David Lynch and completely 
bizarre and you can't figure out -- you at least get to see, 
frame-by-frame and sound-by-sound, what the image and what the 
vision is and what the effect his work has.  So you always get, in 
other copyrighted works, presented with the material itself -- not 
so with software. When a proprietary software company presents 
their software to you, you get it only in the form that the 
computer can understand, the binary "ones and zeros" that execute 
to make the program act. You don't actually get the true 
copyrighted, original, reproduced work, as you do when you get a 
copy of a motion picture or when you get a copy of a musical song 
or when you get a copy of a novel. The original work that is in 
fact under copyright is the source code of the software. But, 
because computer software has this special thing where, 
by-and-large, it needs to be compiled into a binary of some form 
(or into a byte-code of some form) before it can be executed by 
the computer, they can distribute to you only those binary- or 
byte-code-compiled versions of the software, and not distribute 
the source code to you. So they have, of course, the licensing 
restriction against reverse engineering and trying to figure out 
how the software works.   But, they also have this practical 
constraint that really is a happenstance of the way software 
happens to work. It's almost by historical accident that you only 
get the binaries of software and not the source code. It's just 
the way software happens to work.

[13:20]
Now, of course, your freedom to help others is severely restricted 
because of the licensing agreement that you have to agree to; it 
prohibits you from sharing. You immediately have to agree, upon 
opening a package of a proprietary software program, that you're 
not going to help anyone else by sharing a copy with them. Even 
though it's zero-marginal-cost to make a copy,  and this takes a 
little bit of electricity and some hard drive space. So that 
freedom to help others by sharing is restricted with proprietary 
software again by the licensing terms. And, of course, since you 
have neither of those two freedoms, you don't have the freedom to 
help the community by sharing improved versions, by the nature of 
not having the source code, and not having the licensing 
permission to do these things.

[14:04]
So, I've laid out these two different worlds of Free Software and 
proprietary software, and how the rights that you get are set up 
in each of the worlds. But I haven't really shown you yet, at all, 
why these particular freedoms are important. In our society we 
frequently give up what are some "so called freedom" because there 
is just no reason for us to have that particular "freedom", and 
society gains some benefit from us not having it. For example, we 
don't have the "freedom" in our society to drive an automobile in 
any direction we want at any speed we want at any time we want on 
any road we want. We've decided as a culture that those aren't 
really freedoms that are of any consequence --  that it's a much 
better society when we have stop signs and stop lights and traffic 
laws because the safety of the people is more important than a 
particular drivers right to speed down the highway or down the 
road at any speed, ignoring all stop signs and stoplights and so 
forth. So we've decided as a culture that that thing -- driving in 
any direction we want at any speed we want at any time we want -- 
isn't something that matters. And we've decided as a society we're 
just going to give that up and let it go. So when I look at this 
question -- of why these freedoms are important-- I look at it 
from the terms of what harm is caused by not having these 
freedoms. If you don't have the right to do those things that are 
laid out you're allowed to do with Free Software, who is hurt and 
in what way are they hurt?

[15:39]
And a lot of you, I think who are probably my age or younger -- 
you know when you're 30 or younger -- grew up in this age where we 
have both Free Software around and proprietary software 
around. You probably all heard of GNU+Linux, although you probably 
heard it under the name Linux. You probably even used it in one of 
your classes and so forth. And you probably also used the stuff 
from the monopolist down the block, that is available to you under 
some licensing terms, and you probably haven't thought about this 
question at all, I would guess. I certainly didn't as I entered my 
career and I was really skeptical about these ideas of freedom 
that I read about on the FSF's web page and so forth.  And, I 
actually spent 14 years of my career thinking about the 
question. And, I actually came to the same conclusion that the 
founders of our organization did in the end, because of specific 
experiences that I had in the industry that began to show me that 
there was real harm caused, by proprietary software and by the 
practice of propagating proprietary software in our information 
technology industry.

[16:47]
And this 14 year journey begins, for me, when I got my first 
computer. I got a computer that probably by this time you guys 
remember only as an antique. A Commodore 64 was my first 
computer. The older people in the crowd seem to know what that is, 
but the younger people might not. That's just things you download 
ROM-s for, right, in these days? And it didn't come with a whole 
lot of software. It had BASIC that was burned onto the ROM chip in 
the computer. So you got the BASIC programming language with it 
(well, Commodore's version of BASIC at least). And if you wanted 
more software you had to get some peripheral device to plug into 
the machine.  So the first thing I got was a little tape deck, 
that --, you actually had a tape deck that you record programs, on 
which took forever to load a program. It would take hours to load 
a single game or something (well, you know, it seemed like hours I 
was of course 10 years old at the time so anything seemed like 
hours if it was more than 5 minutes). And I got sick of that, so 
my parents finally bought me the 1541 floppy drive unit, which was 
about that long and that high and probably that wide [shows size 
with hands], something like that. It was a pretty big device! It 
weighted more than my laptop sitting over there does now. But it 
was, in some sense, a portable device. You could carry it under 
your arm, at least. It was a little bit heavy. So I got this 1541 
floppy drive. And so I started looking for software. And all the 
software was available in the store, if you could get it. But I 
found there was a lot of other people who had Commodore 64s. And I 
went to this club in my neighborhood for Commodore 64 users. And I 
found other kids my age who also had a Commodore 64 and a 1541 
floppy drive, and we started to hang out together. And in our age, 
10 or 11 years old, we didn't see the computer as anything 
different than the rest of our toys. So we had an attitude about 
the machine and the programs for it that were very similar to the 
attitude our parents had instilled in us, regarding how to deal 
with our toys. And there was two key rules that went around my 
neighborhood about how you ought to play with your toys.  The 
first rule was that you should always share you toys. You should 
always make sure that everybody got a turn. That nobody was left 
out and didn't have a turn to play with the best action figure or 
the best ship in the action figure's universe or whatever. You 
made sure everybody got a turn to play with the toys in equal 
measure. The second thing we were supposed to do, and this was a 
really big deal when I was a kid, was we were supposed to invent 
games that could include everybody's toys. And this was a 
particular problem because, at the time, Star Wars figures were 
probably the largest selling toy. But there was also a whole set 
of knock-offs of "generic universe" space action figures. And they 
were about half the price of Star Wars figures so the kids whose 
parents didn't have as much money in the neighborhood bought them, 
you know, the generic space universe action figure. And the kids 
who had a little bit more money got the real Star Wars official 
Galoob rubber-stamped versions of Luke Skywalker and Darth Vador 
and so forth. And there was this big argument in my community 
about, you know, playing with the kids who didn't have the real 
action figures. And our parents were always saying: "Well, you've 
got to play with those kids too and let them be part of your group 
even though you have different action figures". So this rule was 
developed that you tried to include everybody's toys and make sure 
that everybody's toys could come in to play in the particular 
thing you were doing.

[20:08]
So, here we are with our computers. And we treat them 
fundamentally the same way. And we found both rules applied. First 
of all, we found that everybody could get a turn because there was 
a really neat thing you could do. If you took two of these 1541 
floppy drives and you plugged one to the other and ran a cable to 
the Commodore 64 unit you could make copies of pretty much any 
disk that you had. So every game that we had could be copied. And 
we could share it with all of our friends. And we discovered that 
we were actually engaged in the sharing activity, and it was a lot 
easier than it was with single physical-object toys because you 
didn't have to pass them around to take turns. When everybody got 
a copy, you could play the game together and you could all go home 
and play it separately as well. So that [sharing] rule fit well 
with what we were doing with our computers. Of course the second 
rule, by proxy, fit because to actually make these copies you had 
to bring your 1541 drive over to someone's house. You couldn't 
make the copies by yourself. You needed two disk drives to 
actually get it done at any [reasonable] speed.  With one drive, 
you had to swap disks all the time -- swapping them back and 
forth.  So you'd bring your 1541 drive over to make copies of 
disks. You had [in fact] invented a game that included everybody's 
toys and everybody got to share. Our parents actually discovered 
we played a lot nicer with our computers than we did with our Star 
Trek toys or Star Wars toys. So everything seemed to work out 
ok. And it was going ok for a few months.

[21:31]
And then we started to discover that not everybody had the same 
philosophy about sharing that we did. I discovered this one month 
when I went to the Commodore 64 club and one of the adults in the 
club (there were a few as it turned out) was presenting a video 
game that she had just got called "Little Computer People". And 
there was this little person that lived inside your computer and 
you had to take care of them and deliver their newspaper and make 
sure they got fed every day and had to feed their dog as well. And 
the coolest part about the game for me, at the age of 10, was that 
it remembered the last time you ran the game by checking the 
system clock. And if you took a while to run it, it would compl-- 
the little person would complain at you that you hadn't been by to 
visit in a long time and they were wondering what the news was and 
so forth. This was just the greatest game I'd ever seen. And she 
did a little demo of how she played it and so forth. And met her 
little person that she was taking care of at her computer. And I 
went up afterwards and said: "I found this is just the best game 
I've ever seen. Can I please get a copy so I can play it at home?" 
And her eyes darted a little bit. And my first thought was maybe 
she didn't have time that day to make the copy because there were 
not a bunch of drives around to do it. And I said, I said "Look if 
you are too busy today my parents can drive me over to you house 
tomorrow -- it's Saturday -- and we can hook up the drives and it 
only takes a few minutes and I can get a copy." And she still 
didn't want to do it. So I, like 10-year-olds tend to do, kept 
pressuring her, asking her: "Could you please get me a copy of 
this game?". And finally she looked at me and said: "I can't make 
a copy for you; it's illegal". And this was kind of strange to 
me. So I described to her that my friends did this all the time 
and it was no big deal and it couldn't possibly be illegal. After 
all, you know, the police never came over to our house and said we 
weren't allowed to do this. And it's what our parents told to us 
to do and after all we're sharing. And she stopped me and said: 
"You and your friends are criminals." In fact she had a specific 
word for the kind of criminals we were. She said that we were 
"pirates". At 10 years old, I had a pretty clear image of what a 
pirate was. So as she's telling me that I'm a pirate my mind 
flashed to this imaginary Scooby-Doo episode in my head. Where the 
villain of the week are "s-s-s-software p-p-pirates". And I 
imagine my friends and I as being chased around by the Mystery 
Gang and finally catching us in the end. And Fred (the 
ascot-wearing alpha male of the Mystery Machine) says: "Look, they 
did it by hooking two 1541 drives together." And we turned around 
and we'd take our masks of and say: "We would've gotten away with 
it, too, if it wasn't for you darn kids!" This image of what we 
were doing didn't fit with the reality. We seemed to be doing what 
our parents wanted us to do. We seemed to be sharing and helping 
each other. That seemed right to me but I couldn't really explain 
that to Eveyln at 10 years old so I did what most self-respecting 
10-year-olds would do and said that she was really mean and 
stormed out of the meeting and didn't come back.

[24:25]
But it continued in my mind that there were people in the world 
who didn't think sharing made sense. And I really didn't 
understand why people felt this way and why the world was 
structured this way. But the experience stayed with me until about 
10 years later, uh, 9 years later, when I got my first programming 
job. And I began to experience that, in the real world,  just like 
in the kids world of video games, the feeling about sharing was 
pretty much the same.   But the consequences were a lot bigger. My 
first programming job was for a blood typing laboratory. They did 
mostly paternity suits in those days of people that-- One woman 
accuses the guy being the father and they all get blood-tested and 
we did calculations to figure out the probability that he was the 
father or wasn't the father. And we wrote and fixed the software 
system [for it].   And underneath we used the Informix database 
products as the whole database that stored all the blood lab 
results and so forth. We were running this on an AT&T system V 
release 3.2 on an AT&T 3V2 600, you know, a big iron UNIX 
machine. And I was excited that I got to use a real UNIX system 
for the first time in my life. I even enjoyed learning the vi text 
editor, that's how excited I was. And after a couple of weeks I 
started getting assignments to work on various programs and of 
course every program we wrote had to, underneath, talk to the 
Informix database system, we did it through a library writing 
embedding SQL statements in the C code. So every program you wrote 
linked against the ESQL-C library provided by Informix. So after a 
couple of weeks I started to discover various bugs in the software 
I was working with. I couldn't figure out what I was doing wrong 
in various cases but my programs didn't seem to work 
correctly. And I spend 4 or 5 days, maybe a week, trying to figure 
out why it was that a particular program wasn't functioning. And I 
done programming long enough (since I was a kid) that I knew, if 
you have a piece of code that doesn't work, get it down to the 
smallest number of lines that can produce the bug so that you can 
isolate it and try and figure out what's going on. So I did that 
and still couldn't find the problem. I went to one of my 
co-workers, you know, somebody much older -- probably 10 years or 
15 years older than me -- and said "Greg, I'm trying to figure out 
why this software doesn't work. Could you look at my code and see 
what I've done wrong?" And he looked at my code and audited it 
line by line (this little short example I had that exhibited the 
bug) and said: "Everything looks fine. I don't think it's actually 
a bug in your program. I think it's a bug in the underlying 
Informix library." I said: "Well that really can't be possible 
because we got the Informix stuff in a shrink-wrapped box with a 
manual. So how can it possibly have any bugs in it?" [crowd 
laughs] And he tells me "No, no, no. I've been working with 
Informix software for 10 years; there's tons of bugs. But 
fortunately this particular client site has a tech-support 
contract. So you can call up and report the bug and explain it to 
them. You can even e-mail your code example to them. And they'll 
look into it and get back to and tell you how to fix it or how to 
work around it if they don't have a fix for it." I said "Ok, 
great." So I called up the tech-support line and, this is 
programmer tech-support, so it's not the, you know, Tier-1 
tech-support, but actual programmers on the other end you get to 
talk to, which is, you know, pretty good support, all things 
considered. At least we thought so.   I explained the bug to this 
guy and he took it down, and I e-mailed him the little code 
example and he looked it over. And a couple days later he called 
me back and said: "Yeah, this is a bug in the library on the AT&T 
3b2 platform and I don't know when it's going to get fixed. We 
don't have a work around at this time. Could I just close out the 
bug report?" I said "Well, I really need a fix, you know. I've 
been dealing with this for a while and I really want to find a way 
to fix it. I mean, maybe we can talk through it; we can figure it 
out. I know the 3b2 platform is a little quirky. You know, it was 
hard to get things compiled and didn't have sockets, you 
know. This is bizarre, I know. And so forth. So I said: "Well, 
maybe we can help figure out what's going on." He says, he says 
"No I'm pretty sure--". And I asked: "Why is it 3b2 specific, you 
know, is it fixed on other platforms?" He goes: "Oh yeah, it's 
been fixed on-- this particular bug has been fixed on SunOS for a 
couple of years but there has never been a 3b2 back port of that 
fix." I said: "Well, maybe, maybe we can do something about the 
3b2." He says "No, there's nothing that can be done. I just need 
to close out the bug report. Is that ok?"

[28:55]
And this experience was pretty surreal. As I was saying, we had 
become kind of experts on this platform. It was quirky -- it was 
pre System VR4 and we had gotten used to the way of getting 
various things working on the platform and how it differed from 
other UNIX-like systems. So I probably didn't know enough about 
the platform to fix it, but Greg probably did, or somebody on the 
site probably did. But I didn't even bother to ask him for this 
code because I had learned, 9 or 10 years before, what happens 
when you ask people to share software with you that don't want to 
share it with you. So I just hung up-- neatly hung up the phone 
and gave up. And it was really, I think, a surreal experience 
because this was programmer tech-support. They're supposed to help 
other programmers. But the first thing that probably happened the 
first day this guy went in to work was he was ordered to sign 
something that made him promise not to do his job. He probably had 
to sign a non-disclosure agreement about the software he was 
viewing. He probably was required, as part of his job, to not help 
the programmers who called him, in the most natural way two 
programmers help each other. I'm sure those of you that are 
computer science students have had the experience of trying to get 
a friend to help you debug a program.  And eventually it comes 
down to showing the friend your source code so they can see what 
you're doing. Because even anything beyond the most trivial of 
questions, requires actually looking and sharing the source code 
to work together to solve the problem. But in this particular 
case, the tech-support for programmers was set up in a way that 
required that they not help any programmers whatsoever, in the 
natural way that two programmers help each other. I think that 
this is a bizarre thing to have, be true. But, it was standard and 
often still is standard. When you get tech-support for programming 
from somebody who produces middleware, they are not allowed to 
share the software with you and help you in the most natural way 
that makes sense. So I just gave up on that particular problem and 
I found that as my career continued, this happened over and over 
again. I was always in the situation where I was helpless to solve 
some particular problem because the upstream software that we had, 
had some bug or another in it. And it's not just that I and the 
programmers I worked with lost productivity, the users lost a 
great deal of productivity as well. The company I was working for 
did blood typing on lab results. They didn't care where the bugs 
were that we had to find. They wouldn't even have cared if we had 
made the change and donated the software back to Informix. All 
they wanted was for the software to work so they could get their 
job done. That had nothing to do with what the underlying database 
did or some bug in the software. They would have happily payed us 
to find Informix's bug, so that we could move on and get to the 
next thing that we needed to do for them. And the problem was, 
though, that we didn't have the freedom to do that. We didn't have 
the right to do that because the software was licensed under 
proprietary terms.

[32:04]
And as I continued my software development career it continued to 
be pretty much the same. So I decided at one point that I wouldn't 
do software development any more. That there was another career I 
could have, another career track in computer science, that I could 
do. And I heard how easy it was and you didn't have to do any 
work, you just sat around all day. It was called system 
administration. [crowd laughs] So I was pretty excited until I 
actually started doing the job and realized that it was a lot more 
work most of the time than software development. And a lot more 
"under the gun" kind of stuff. But, at the time, I was excited 
'cause it sounded real easy -- in the land of dreams where you 
just sit there and hide from the users and never do any work.

[32:42]
So, my first sysadmin job that I got was for Westinghouse 
(by-and-large the divisions I worked for don't exist anymore -- 
they were swallowed up by Northrop Grumman). And I was with the 
mobile satellite telephone division of Westinghouse. These were 
satellite telephones that talk directly up to the satellites and 
don't go through any of the typical cellular mobile 
infrastructure. And they're used in-- Forest Rangers use them and, 
and other places where you don't have infrastructure. They're used 
in a lot of rescue environments where, if you have a hurricane 
going on in Florida, the mobile towers might be down or not 
functioning. So you have a way to communicate with regular 
telephones with regular telephone numbers across a satellite 
link. And they were pretty expensive devices and they developed 
the embedded software that went inside the telephones.  And I was 
just supporting their use of the Solaris systems, that were there, 
that they used to actually do that software development. So the 
details of what I was doing really didn't matter to what the 
company was doing. They just needed these machines to work, so 
that they could get their job done. To write this software, so 
that forest rangers could talk to each other and that they could 
do rescue situations when there were hurricanes and that sort of 
thing. So I was in charge of these Solaris machines. And they had 
a whole list of bugs when I showed up. They had been without a 
sysadmin for a while and there was these annoying things about the 
system as it was set up that were causing them trouble.

[34:01]
The number one thing on my list was, the-- They were using this 
NIS+ authentication system, a proprietary Solaris login 
authentication system, and they discovered this bug that when xdm, 
which is the, which is the X-windows login program that presents 
you with you login password prompt on an X-windows screen, that 
when they typed in their username and password the first time in 
the morning, it didn't work correctly. Even if they typed 
everything without any mistakes. And they discovered, though, that 
those users who had logged in the night before, through a modem 
connection via remote access, were able to login just fine. So it 
was some sort of timeout bug with the length of time since you'd 
last logged in. So what they would do every morning is, they would 
find somebody who had logged in the night before.  Which means 
anyone who got in that morning and had an active login, they'd go 
over to their machine, remotely log in to the machine back on 
their desk, which made an active login. They'd log back out; 
they'd come back to their machine and log in, and everything would 
work fine, because they'd recently had a remote login close enough 
so the timeout bug wasn't causing a problem. This took about 45 
minutes every morning. Everybody would wander around to each 
others desks -- trying to find somebody who got an active login 
that morning -- use it to log back into their own machine and 
everybody would be able to log on. It worked "just fine". Uhm, and 
there was a Westinghouse rule that you couldn't leave a soft login 
overnight, so they couldn't do that either. So they, they had this 
problem, and they wanted me to solve it and find a way that they 
didn't have to wander around. Management was particularly worried 
about this because it was wasting valuable time. Every morning all 
these people wandered around and chatted looking for active 
logins.

[35:31]
So I had looked into the problem and I found actually on Sun's 
website a list of bugs related to NIS+ with their interaction with 
X-windows and this was actually one of them. "Xdm login not 
functioning but remote OK" or something like that. So I called 
them [Sun] up and said: "Great, you know about this bug 
already". I don't have to like convince them it's a real bug. I 
called up Sun tech-support and said: "Hey look, we have this 
bug. We don't know what to do. How do I fix it?  It's not working 
and you've got patches for Solaris 2.5 but we're running 2.4. You 
know, can we get the backported patch for Solaris 2.4." And they 
said: "Well, no, you have to upgrade to Solaris 2.5." And I saidL 
"Well, how much is that?" It was like $800 a seat, or something 
like that. I said: "Well, we can't do that. (1) because I don't 
have the budget for it.  (2) because we've actually done an ISO 
9000 regression test on the whole environment. And it has listed 
in it all the versions of software we're using and the only thing 
we can get away with doing are critical bug fixes. We can't do any 
major change to upgrade the operating system; we just have to 
stick with bug fixes. And after all, Solaris 2.4 is only about a 
year and a half old at this point. So it's still pretty actively 
used." And I said: "By the way, I noticed that you fixed a number 
of other X-windows/NIS+ interaction bugs on the website for 2.4 
and there is patches for those, which of course I've installed, 
but they don't fix this particular bug. Why did you fix some of 
them and not all of them?" They said: "Oh, we did those because 
NASA requested them and they are a big Sun customer. And they can 
tell us to do things and we have to do them, but you're just not a 
big enough Sun customer to demand that kind of service." Now, of 
course, I was surprised to learn that Westinghouse was a small Sun 
customer, [crowd laughs] because Westinghouse was pretty big in 
these days, but we were a small division of Westinghouse so they 
treated us like a tiny customer.  So, I wasn't able to fix the 
problem.

[37:17]
And I found that I found myself going back to the users and 
delivering that same message, that I had heard a couple of years 
before from the Informix tech-support guy: "Yeah, this is a bug in 
NIS+. No, I don't know when it's going to get fixed. We're just 
gonna have to hope somebody logged in the night before, so we can 
all get logins in the morning." So that kind of apathy that gets 
created by proprietary software is rampant.  People don't feel 
like they can get their jobs done, and they are not empowered to 
get their jobs done, because they don't have access to the 
software. I'd actually run strace against the binary; I knew the 
exact place where it was making the call back to X-windows and 
exactly  -- more or less exactly as I didn't have the source code 
-- what needed to be fixed. Or at least where the bug would be and 
where I'd have to go in the code. But I didn't have access to the 
source code for the version of X-windows I was running, I didn't 
have access to the source code for NIS+.  Sun didn't want to help 
me because I wasn't NASA. So I had nothing I could do tho fix the 
problem. Just had to hope that the users would be happy with the 
shoddy system that they had.

[38:20]
Now all of this isn't to say Free Software isn't without bugs. We 
have our share. A lot of people argue we have more than our 
share. It's not even to say that the original developer of the 
particular Free Software program you find a bug in will care about 
your bug. They may be just like Sun and Microsoft in those 
situations, or just like Informix. They don't care about this bug, 
it doesn't matter to them. The difference though, with Free 
Software, is that you have the means yourself to improve and fix 
the software. Or to hire anyone you want to improve and fix the 
software for you. And this I think is the fundamental reason that 
I moved to Free Software and become a professional advocate for 
Free Software. But at the time I was kind of convinced that I 
could never do what I wanted to do in computer science. I came 
into this field because I wanted to make computers easier to use 
for people. I wanted peoples lives to be better because of 
technology. But I was pretty sure at that point that I couldn't do 
that. And this was most evident when I remember an experience 
right between my time at RH lab and my time at Westinghouse.

[39:34]
I worked for three months on a contract for Lucent Technologies to 
re-engineer a legacy system. It was used to configure telephone 
switches.  And the telephone switches in those days were actually 
sort of sophisticated Unix boxes. And -- ah -- in fact that 4ESS, 
which is a standard telephone switch, is actually used today in a 
lot of places still. And the gurus who understand these boxes can 
actually log into the thing via SSH these days and actually fix 
things with vi on the command line and do all kind of weird stuff 
to the telephone switch. But the average person you call up, when 
you ask for Lucent to implement a Centrex or some other system for 
you business phone system for you, don't know how to do this. So 
Lucent had developed a product for use by the Baby Bells, who use 
these switches, to configure them through a nice little user 
interface and so forth that they -- well, it wasn't a nice user 
interface, it was a rather clunky interface but certainly better 
than what the gurus had to know to be able to get the job 
done. And in fact, as they were beginning this re-engineering 
process they had this big user group meeting. They invited all the 
Baby Bells to come (back when they were Baby Bells, right?)  They 
all came -- Nynex was there, Southwestern Bell was there, Pacific 
Bell was there -- all talking about how they wanted to use the 
program. I thought: "This is great!  We're actually listening to 
users! We're gonna do what they want us to do and make the 
software better."  I was actually pretty optimistic as I went into 
this user group meeting. And we listened to them describe all 
these various bugs that they had in the software they were 
using. And there is this one bug that is wasting like 5 or 7 
minutes every call, because they had to go in deep into one menu 
structure, and then escape all the way out, and there was lag in 
the system so they hit escape one too many times they were too far 
out, and they had to dig back down into a separate menu structure, 
and there was no easy way to connect up these many systems in the 
way the software was originally designed. And I was looking at our 
new design. I went back to my desk after that days user group 
meeting, and started looking at what we were doing. And I went to 
my boss and said: "This is just great! I can solve that problem 
for free because it just comes out of the wash with our new 
redesign. And the menu structures won't have to be that way and 
will be perfect, to solve this problem." He looks at me and says 
"OK, do you know how to turn it off?" I said: "What do you mean, 
turn it off? Nobody wanted the menu structures laid out in the way 
they are. We don't want to turn this off, it should be the 
default." He goes: "No, we have to be able to turn it off. We have 
to make the old interface the default and make this an add-on 
option." "So why is that?"  He says: "Because we can get more 
money out of them by requiring them to pay extra for that." This 
is interesting because two days before I had seen a budget sheet 
for the division I was working in. Division of fifty people, was 
making a profit of $500,000 a month of licensing fees and support 
fees and upgrade fees from all these Baby Bells, who were just 
paying every month, for large amounts of money for the new 
upgrades. But they had a way in the contract that they could add a 
little extra thing, a bell and whistle, that they could charge 
separately for. So we were gonna exploit this little hole in the 
contract, even tough their profits were completely fine for the 
time. Because they could get away with it.

[42:34]
And what that experience showed me, more than all the others, was 
that in most of the software industry, turning a profit and 
getting the profit as high as possible, is substantially more 
important, in a lot of proprietary software companies, than making 
the lives of people better through technology. Now I don't have 
anything against profit. I like to get paid.  I like to make 
money. I enjoy that I don't have to not work and that I can get 
money for what I do. But I also don't think that the biggest goal 
in working with computers is to make money. And if you happen to 
be one of the people who thinks that it is, and feels that the 
only reason you're doing this computer stuff is to make a lot of 
money, then we probably don't have much we can connect on. And you 
probably ought to apply for a job at Microsoft and try to move up 
in there corporate structure, because that's the way they think 
there and Oracle is much the same. But if you perhaps think that 
making money doesn't have to be number one, even if it's just as 
number two and that the first thing is to make the technology 
better for people to be happier and for peoples lives to be better 
than I think we have something to talk about. Because, what I 
discovered, in the IT industry, is that with the focus solely on 
profit, the whole mission of computer science is forgotten. It let 
me to be pretty much apathetic to the whole industry and how it 
worked. I pretty much wanted to give up on computer science all 
together. But I was discovering, because I was going to college at 
the same time I was working, that there was actually an other 
alternative, that there was an other way of doing things.

[44:09]
And this was because in 1992 I became an early adopter of what 
today is the GNU+Linux system that you may have heard under the 
name just Linux. I didn't understand any issues about software 
freedom or why sharing was important other than what I vaguely 
understood from my experiences. I just needed a low cost 
(i.e. free beer)  Unix-like operating system that I could use on 
my own machine to get my computer science homework done. And I 
discovered that there was this software out there. And I installed 
it on my machine and started showing it to my professors and 
showing them how it worked and that it was available and it was 
under this license called the GPL and you could copy it as much as 
you wanted. It wasn't illegal and you were allowed to share 
it. And they said: "That's really interesting because we just got 
a price quote from Sun to buy a bunch of new machines for our Lab 
but we can't afford them: (a) because Sun's hardware is too 
expensive and (b) because the software licenses are too 
expensive."  Meanwhile commodity PC hardware we could get three 
times the machines at the same price. And if you could install all 
this stuff on the systems we'd have Unix-like workstations for all 
the students. And we wouldn't have this problem where everybody is 
trying to share [the hardware] because we have just a very few 
number of machines. We have enough for everybody so that nobody 
really has to wait for a machine. So they hired me that summer to 
actually install GNU/Linux on this new batch of Dell computers 
that we were getting. So actually as early as 1992 I went into the 
business of supporting Free Software. I was paid a stipend for 
that summer to make these systems work with Free Software. And I 
quickly discovered that it wasn't as easy as it all looked. There 
were bugs and problems.

The biggest problem we had in those day's was that the 
Ctrl-Alt-Del sequence on the keyboards -- were hot [rather] -- was 
hardwired in the kernel named Linux. The activity that would 
happened if you hit the Ctrl, Alt and Del all at once:  it 
actually automatically did a system reboot.  These days, that's 
controlled by the init system and you can tell it what -- you'd 
[rather] -- what operation on the system you want to happen when 
somebody hits those keys. But in those days it was hardwired, so 
it appeared. Now, the problem was, that a lot of the students were 
coming from regular computer labs, into this new lab that we were 
creating that had Unix-like systems, and in the regular labs they 
apparently had to hit this key sequence a lot. Apparently they 
were running some operating system that requires you to hit this 
over and over again, for no apparent reason. So, we didn't want 
them to come in and do this particularly because we wanted to 
allow remote access to these machines, and somebody could be 
logged in from there dorm room using the machine remotely, and if 
somebody hit this key sequence the rebooted the machine on them, 
in an instant. So they said: "Well, you got to fix that, that's a 
problem."

So, what was I going to do?  Of course, I was a sophomore in 
computer science.  [noise from crowd] (Oh, there's some sophomores 
in the room!).  But I had this belief that, well, *any* software 
problem: I know how to fix, because I am a programmer.  So I 
started grep'ing the source code for the kernel named Linux for 
the words "control",  "alt", and "delete".  Lo and behold, I found 
a function called "ctrl_alt_del". And I saw that it did a bunch of 
things I didn't really understand, then it called the halt system 
call. So, being a sophomore in computer science, I said: "Well, 
anytime something doesn't do what you want it to do. You just 
comment it out, recompile and everything will be fine."   Now in 
the general case, this is a really bad thing to do. In this 
specific case, it was exactly what was needed. I commented out the 
halt system call, recompiled the kernel named Linux, installed it 
across the lab and walked around pressing Ctrl-Alt-Del on every 
system. And absolutely nothing happened. Which was precisely the 
behavior we wanted for this lab. So, I found it amazingly 
rewarding and exhilarating to be able to fix this problem 
myself. I think it's ultimately liberating when you give people 
the tools to solve their own problems, rather than requiring them 
to call up some tech support and beg for help and beg to get their 
problem paid attention to or fixed.

[48:07]
And because of this experience and many others that I had with 
Free Software, by the mid 1990's I understood what this ethical 
thing was about. I had seen real problems in the real world 
created by proprietary software. Specifically because it was 
licensed under these proprietary terms. And I saw two things in 
the Free Software world. One, I saw that you could solve these 
problems yourself and two, I saw that if you knew how to solve 
these problems yourself, and could convince others that you were 
good at it, you could get paid to solve these kinds of 
problems. You could get paid to actually rewarding and 
exhilarating work in the software industry rather than calling up 
tech support and begging them to fix your problem.

[48:49]
Now as time progressed this Free Software stuff showed up on 
everybody else's radar screen, sometime around the late 1990's 
along with the .com boom. And in fact it unfortunately got 
associated with that which is really unfortunate because it 
existed long before that and it's going to exist long after. And 
the mainstream media picked up on this and started focusing on a 
very specific aspect of what our community had done: that we had 
made reliable, useful, dependable software that had (by and large) 
fewer bugs than the proprietary equivalents in very specific 
spaces such as web servers and operating system server 
software. And because they got focused on this aspect, the ideas 
of freedom and sharing -- throughout the late 1990's and early 
into this decade -- people didn't realize that there was a 
political and philosophical basis that generated this 
community. And it goes back really far, way back in 1984.   When I 
was busy in my Commodore 64 club, being called a pirate and 
getting told I shouldn't share software,  Richard Stallman (who is 
my boss these days) was thinking about these same issues. He was 
considering this because he had watched his own community of 
software sharing disappear. He is from the era, and many people of 
his generation, remember the 1970's and what computing was like in 
the 1970's. In those days, you had big computer manufacturers who 
sold you giant pieces of iron, and the software was on there 
almost as an afterthought. It was there to make that big piece of 
iron useful to you.  And they didn't really care if universities 
got access to it and improved it -- that sold more hardware. They 
didn't care if you got access to improve it and sold it with 
another one of there clients. It might sell another hardware unit 
if you do that. It didn't really matter that people had the source 
code to the software and were sharing it. If anything, it helped 
the sale of the hardware.

[50:54]
But as the mid 1980s wore on, the whole thing changed. The PC 
revolution happened. And a few people discovered that they could 
invent a new way of doing software. They could "license" software, 
rather than include it as part of the package. Bill Gates is often 
credited for inventing software licensing although I've never seen 
any direct evidence that it's true. But certainly he is one of the 
few who benefited most from it, at least financially. They 
discovered that you could divide the whole user base.  Conquer 
each person one by one, by forbidding them, under license, from 
helping the  next person, forbidding the next person from helping 
the next person. And if you could get each of them to pay some 
amount of money, for stuff that actually has zero marginal 
cost. It was a pretty innovative idea from a business point of 
view, but it was not all that innovative from a freedom point of 
view. It took what was a community of software sharing and 
academic research and endeavor to improve computers and improve 
how they did their jobs for people, into effectively a way to stop 
people from helping each other.

[52:06]
And in looking at this, Stallman decided to create the Free 
Software movement. And it could have been done in a straight 
forward way, like a lot of movements that don't have the means to 
change things themselves.  This is why you see a lot of 
protests. And I suppose the Free Software movement could have 
designed itself like that: eh, you know, a movement that stood 
outside of offices of proprietary software companies like 
Microsoft and held signs that said "Make your software free!" or 
we're gonna write letters and we're gonna chain ourselves to your 
door and all these sort of things. And I feel bad for movements 
who have no other choice but to do that. But the innovative thing 
that Stallman discovered in the Free Software movement -- in 
creating the Free Software movement -- was that there was a really 
productive way of political action that could be done. People were 
upset about this proprietary software, upset that software didn't 
come with source code with the right to share it or modify it or 
improve it. And they said: "Well what do you need to create an 
alternative?" You need access to a bunch of computers, and people 
with good minds for programming to create software. And in fact, 
the community that was falling apart because of the 
proprietarization of software was there to do this job.

[53:21]
So what they did was start the GNU project: a project to create an 
operating system that works like Unix but was, in fact, not 
Unix. Created from scratch, to work just like the proprietary 
equivalent but be available Free as in Freedom, with the right to 
copy share modify and redistribute software. And this is where it 
all started, this is where we began to create something 
different. While it wasn't focused as profit as a primary motive, 
free software has made a lot of money for a lot of people. I know 
even after the .com boom personally three millionaires who made 
there million from Free Software. Now, it hasn't generated any 
billionaires: you've probably got to do proprietary software to be 
billionaire, but maybe you can settle for being a millionaire 
instead and do Free Software, rather than being a billionaire and 
doing proprietary software.

Because they can generate that level of unbelievable financial 
wealth by treating users badly and putting them in effectively an 
unfair position.  See, when I think back to my first experience 
with Eveyln and asking for a copy of "Little Computer People", I 
realized that she was actually put in an unfair position. It 
wasn't her fault. She had to make a decision between obeying a 
software license that she had (in good faith) agreed to, and doing 
the right thing for a child who was asking for a favor that was 
easy to do. I don't like myself to be put into that situation. If 
something is easy to do for somebody to help them out, I do it. I 
think we all do because we want to be good people and help the 
world. But proprietary software puts us in this situation where we 
have to choose between doing the nice thing and obeying the 
license we agreed to. And Free Software frees you from that 
ethical dilemma.

I believe that we should have laws in a society. I believe that 
copyright law probably had a very good goal behind it. Indeed our 
constitution said it was there to create and promote progress in 
science and the useful arts. A laudable goal, no question!  But 
when you implement laws, they should make peoples lives 
better. They should increase good will and cooperation; that's why 
we should have laws in our society. And the copyright system -- as 
it's typically applied to computer software -- does exactly the 
opposite. It uses this system to stop people from being allowed to 
help each other. Stop people from improving the software they use 
and understanding the software they use.

[55:58]
You can make a lot of profit the other way. But you can make a 
reasonable profit this way too. And, in fact, in the Free Software 
world you make profit in the typical way that you do in most 
proprietary software companies. Only the largest, most powerful 
companies like Oracle and Microsoft actually can make a profit on 
licensing alone. Most software companies make the bulk of their 
revenue from the typical models that we also use in the Free 
Software world. They make their revenue from service contracts, 
from support contracts, from custom modification for  software 
commissioned by a company who needs an improved feature. These 
methods of developing software -- where you're commissioned to 
improve a software program that already exists or you're selling a 
service contract for software that's already available -- these 
are the models that make most of the money at the "smaller 
company" level in the software industry.

[56:55]
I worked myself for a proprietary software company that did, in 
fact, charge licensing fees, but if you looked at the budget 
sheets the licensing fees were this tiny little piece that barely 
even covered the cost of one of us or two of us and there were 
like five or six programmers working for the company. The rest of 
the revenue  paid to everybody else and to make the profit the 
company sent back to its shareholders (or private owners, in this 
case) was from this service contracts and support and the hiring 
to improve the software for new changes and new features that was 
done.

[57:26]
So these models are exactly the same in the Free Software world. 
In fact they're more available because your ability to start off 
from scratch is much easier, because there is all this existing 
software infrastructure and the reference material, right in front 
of you, today. The source code to the entire operating system and 
all the applications available in Free Software is available to 
you right now out there on the Internet. You can learn not how the 
toy examples work, but how the production stuff that actually gets 
real work done in the real world every day works. And you can 
study it and you can learn it and you can improve it. You can 
become an expert on it. And in the Free Software space the experts 
make the most money, the people who know the most make the most 
money, get the most contracts, find the most clients.

[58:20]
This a purely capitalistic way of designing an industry. I'm 
amazed when capitalists come to us and call us un-American and 
that we created cancer on the software industry that will destroy 
it. We're substantially more capitalistic. We're opposing the 
government-determined monopoly in Seattle. We're saying: "Look, 
the software industry ought to be set up such that everybody has 
access to a competitive free market, where the  the amount you 
know about software and your ability to make it work for people is 
what determines how much money you can make and how many jobs you 
can get". That's the way I think a competitive capitalistic 
software industry ought to work. I'm not really deeply 
capitalistic, but I see how the capitalist's philosophy fits 
perfectly well with Free Software. It is, in effect, a meritocracy 
of people who know how the software works being able to get the 
best contracts and jobs.

[59:15]
Of course, as we started to distribute Free Software in the 
1980's, there was a real question about how to get it done. As I 
said, our goal was to create a system that would make sure the 
software we made was free.  We wanted everyone to get these 
freedoms, so we had to consider what the terms of distribution 
would be. We obviously weren't going to release it under a 
proprietary license, but we quickly discovered as a community that 
licensing would be necessary and here is why: The easiest way to 
avoid the proprietary way of doing things is to release software 
into the public domain. When you release software in the public 
domain you are asserting no copyright whatsoever over it. And in 
fact, public domain software gives you the right to copy, share, 
modify, redistribute and use the software in any way you see 
fit. The downside to that is that anyone else who would like to 
make a proprietary adaptation or modification to the software can 
do so and license it back out into the world under proprietary 
licensing terms. Pretty much everybody who worked early in the GNU 
project is sure, that if GNU had been licensed in this way from 
the start, it would likely have been a failure. Because anything 
that was particularly useful would have immediately been picked up 
by a proprietary software company, made into proprietary software, 
and the Free version would constantly play catch-up with the 
proprietary version, since there was no incentive for them to 
cooperate with the free developers.

[1:00:50]
So, what was discovered was a new way of using copyright law, to 
take the world back to the old way that software existed before 
proprietary licensing. This new idea was called copyleft, because 
it took what copyright law typically did and flipped it over -- 
turned it on its head -- to something different. What it said was, 
rather than prohibit the things that copyright law governs, which 
most software licenses do (i.e., copying, modifying and 
distributing the software),  this copyleft license, from the 
start, grants you permission to copy, share, modify and 
redistribute the software. The only restriction placed upon you is 
that you can't -- if you engage in these activities -- take those 
rights from anyone else. In essence, it grants you all the 
freedoms that I talked about at the beginning of this talk, but 
forbids you to forbid those freedoms of anyone else.

[1:01:48]
This is the general idea of how to do software sharing, and it 
needed a specific implementation. So Stallman sat down with a 
couple of lawyers and wrote the GNU General Public License or the 
GNU GPL. GPL is effectively the constitution of the Free Software 
Movement. It explains how legally, you get all these freedoms to 
copy, share, modify and distribute, but that no one who 
subsequently exercises those freedoms is able to take away your 
rights nor anyone else's rights in the community. It does an 
interesting thing: it avoids the tragedy of the commons that comes 
about when software is in the public domain. Companies are 
unwilling to service the commons because there is no incentive to 
do so.   If there is no requirement to do so, they'll take but not 
give back. Bill Gates, when he spoke here a couple of months ago, 
talked very highly of this kind of software because it allowed 
companies to innovate around software, but he was really talking 
about [the fact that] it allowed companies to take without giving 
back. It allows companies to chop down all the trees in your 
nature conservancy but not plant any new ones. What GPL does is it 
say: "Great, we want companies to be involved,  if you want to 
be. If you don't [want] to be, then write your software from 
scratch." If Bill Gates wants to never touch GPL's code, he is 
permitted to. Unlike what he and his other executives say, it is 
not a cancer or virus. You can't catch it just by sneezing in the 
direction of proprietary software. It has to actually be brought 
into your software, and once you choose to do that activity, you 
have some obligations to the community that gave that 
benefit. It's a club that nobody is required to join, but once you 
join, you have rules that you have to follow. You have to make 
sure that you give other people the same rights that you got when 
you got the software.

[1:03:37]
I -- [and] We don't have anything against Free Software under 
non-copylefted licenses. It has this drawback. In fact, I have 
experienced it myself when I was working at Westinghouse with the 
Solaris machines.  X-windows, in fact, was Free Software under a 
modified-BSD license which is effectively the same as public 
domain, in its details. So I got a copy of xdm that was Free 
Software when Sun got it, but they made proprietary modifications 
to integrate it with NIS+ and by the time I got it, it was 
proprietary software. So I've actually been a victim myself of 
this problem that comes up when software is not copylefted. But at 
the same time, as we were trying to build a Free Software 
operating system from scratch we had to look for shortcuts. There 
was no reason to reject existing Free Software that granted 
freedoms if it was there, even if it wasn't copylefted. So, the 
GNU project started adopting other various Free Software programs 
that were around, the X-windows system is a great example, it's 
the default windowing system for GNU+Linux systems, the TeX 
document formatting system was adopted as the official formatting 
engine for the GNU system. And by 1991, pretty much all the 
different components had been assembled except for one: the 
kernel. Meanwhile, someone was interested in operating system 
development over in Finland and decided to write a kernel from 
scratch called Linux. That kernel integrated with the existing GNU 
software and created what is the modern GNU+Linux system that most 
people shorten to just Linux.

[1:05:12]
And the reason that we continue to talk about this history of the 
GNU system in relation to Linux is that GNU was started because of 
a philosophy for freedom. Linux was started because Linus was 
interested in technology and wanted to start doing a project. He 
picked our license, the GPL, for his project, because he liked the 
stuff that we had already done.  But it isn't a project that's 
focused on this political issue of software freedom -- the rights 
you ought to get with software. But this whole community came into 
being specifically because people who were very focused on rights 
for users and programmers were working to create this operating 
system that would be freely available and freely shared. So we 
continue in our community to focus on these ethical issues and 
talk about the importance of community and cooperation and sharing 
of software to help other people. A lot of times, you hear it 
talked about in terms of "Open Source".   That term has primarily 
become the corporate way of thinking about it. We think it's good 
that corporations want to get involved with what we do and we 
understand that companies aren't particularly interested in 
talking about political issues. So that's a good term to think 
about, in those terms,  when you are thinking about the corporate 
side of this: why it is usable software and why a company would 
want to use it. I continue to say "Free Software" rather that 
"Open Source" to get across the idea that this is about freedom -- 
that we did all this work to give people certain rights and to 
effectively make those rights inalienable. Because I believe, very 
strongly, that computer scientists need to consider the ethical 
and sociological impact to the work they do.

[66:56]
Too often, we get focused as computer scientists on technology. We 
focus on the exciting, great wizz-bang new technology that we've 
developed, the new software we have written that does all kinds of 
amazing things, and forget that: (a) there is a bunch of users out 
there who need to be given the benefit of this technology and the 
technology should be shared with them, and, (b) we have to think 
about how that creates a community, and how software can be used 
to foster community and foster people helping each other. Too 
often we get focused just on the technological issues. So I 
encourage you to think about this issue when you're developing 
software. 

A lot of people say they don't have a choice: that they have to 
write proprietary software, particularly the way the economy is 
now. People are happy to get any job they can get. And, I am not 
going to say that you've done a terrible thing if you take a 
proprietary software job. But, think about this issue. Try, if you 
do take a proprietary software job. Those of you who are students, 
when you graduate, talk to your employers about Free Software. Try 
to get them to do a little bit of it, if you can. If you're able, 
volunteer some time for various Free Software projects when you 
are not working. Try help create a free software infrastructure. I 
guarantee you if learn a piece of free software really well, 
you'll find it helps you greatly in your career. It will change 
the way you are able to understand software and collaborate on 
projects. And that is a skill any employer needs, no matter what 
kind of software development they do. And, I encourage you to 
think about these issues and see if you can find a way to make 
money doing Free Software.  There are a lot of business models out 
there -- support, service, modifications and so forth -- that are 
really usable and they work.

Now, the other side is going to argue that we are here to stop 
innovation.  You are going to hear a lot of that. I'm sure you 
heard it when Bill Gates was here. I heard some pieces of excerpts 
of the speech when he was saying that we are out here to stop 
innovation and stop things. And I actually think that the truth is 
we make a competitive free market like I was talking about. Free 
Software ends up causing the software industry to function a lot 
like other professional endeavors that we are used to.

[69:10]
One great example is the law. Lawyers still make a lot of money. 
In pretty much any economy good lawyers make a lot of money. But 
the law in the United States is free in the senses that I've been 
talking about. All the laws are available to study. You can go to 
a library and read all the laws of our society and learn them. If 
you don't want to learn them on your own, you can go to law school 
and pay them to teach you about the law. But most of the time we 
don't do that as individual citizens. We hire an expert to help us 
understand things that we don't understand rather than trying to 
learn it all by ourselves. This is why people hire lawyers.  And 
people who are better lawyers and understand the law better are 
able to get more money for their services. Now we could design a 
society that has proprietary law in the same that we have 
proprietary software. In my science fiction fandom, I have seen 
this kind of culture in the Star Trek universe there is the 
Cardassians who have a legal system whereby you are told you 
sentence before you go to trial. And the trial is just a big show 
thing and they tell you: "Ok, you are going to be executed and you 
don't get to see the evidence and examine it. And, you don't get 
to know which law you violated that is causing you to be 
executed."  We could design that kind of legal system; that has 
existed in our history. But nobody would call that legal system 
"free". It certainly functions more efficiently. It certainly 
works to make the society move along nice and smooth. But it's not 
free, in the senses of freedom that we think about.  And the same 
thing is true in the software world. Sometimes Free Software 
requires a little more work to get going. Sometimes, to get a 
project done, it takes more effort. But the end outcome is free in 
that sense.  It's available to the public in that sense. And the 
experts still are an important part of the economic ecosystem of 
software, just as they are in a proprietary world. Even more so in 
the Free Software world.

[1:11:10]
I want to end by saying (and [then] take as many questions you'll 
like to ask), by pointing out that it is up to the next generation 
of computer scientists to decide where this goes. I was told 
today, in one of your classes (in one of the ethics classes here 
on campus) that we are failing, because we haven't succeeded in 
beating Microsoft. [But] I'm not looking what happens in the next 
few years. I'm looking further out, in the next 10, the next 20, 
the next 50 [years]. What is computer technology going to be like? 
Is it going to be available to everyone freely,  So that they can 
learn and grow and actually have access to the technology that 
they are using? Or, is it going to be closed and proprietary, and 
forbid people from sharing and forbid people from learning? Are 
the computer science education programs of the next generation 
going to be reading manuals of proprietary software and being 
told: "Well, when you get to work for Oracle or Microsoft someday, 
you'll actually get to learn how this stuff works in the real 
world, but we can't tell you [now] because it's proprietary". It's 
a strange way to run an education system, but we are headed in 
that direction. If you are using Microsoft computers to do your 
computer science homework, think about the fact that you are 
learning computer science on a system that is an implementation of 
ideas you are learning, yet you can't study that 
implementation. It's a strange system to be under.

I hope that you'll choose to use the GPL for the software you 
write, to help society and help people learn, and help the 
community get more access to software. And, I hope that if you end 
up working in the proprietary software world, you'll at least 
become a member of the Free Software Foundation and send some of 
that big fat check to us so that we can continue to work, even if 
you are not able to contribute, for whatever reason. I hope you 
will fight for a free and cooperative culture. That profits not 
just you individually or companies, but the whole of society.

[1:13:10]
I thank you for your time and am happy to talk in detail about 
anything that you guys want to ask about.

[applause]