UTOSC 2009 - Stormy Peters - Would you do it again for free?

Let's introduce our keynote.

For those that have not heard of Stormy Peters, Executive Director of the GNOME Foundation and overall a very cool person, if you follow her blog.

There's a great context that a lot of us some times miss about the interaction between the people

and what open source really has become to a lot of us personally and that's really what she's going to talk to us about

and it's going to build on what we heard about yesterday

about what open source can really do for the community and how that affects us and what we can get out of it.

So with that I'd like to welcome Stormy Peters. (applause)

Thanks.

So last night we heard that you would always give a talk about something that you are passionate about.

And while I'm passionate about free and open source software, I'm even more passionate about the people that work on it.

To me it's possible because of all the people.

It's an amazing community because of all the people.

At the GNOME Foundation I get to work with people that are doing, they're working on GNOME because they are passionate about working on it

and that makes my job so much more fun than going to work with people that are getting paid to do it.

So a couple of years ago I was at a conference and I was sitting in the audience and I don't remember what the talk was on.

It was at OSCON though. I was sitting there and the woman that was speaking said if you start working on something because you love it and then you get paid to work on it, if that paycheck goes away, you'll stop working on it.

I thought whoa, whoa, whoa. Because at the time I was working at HP and then OpenLogic

And my whole job was about getting companies and communities aligned and helping bring corporate funding to open source.

I still do that at the GNOME Foundation.

So I thought I better look into this.

I better figure out if we are really killing the way it works.

As a show of hands, how many people got into free software as not part of their job? 75-80% of the room or more.

How many of you now get paid to work on free software? So about a 1/3 to 1/2.

So maybe this question is relevant.

I am going to share some of the studies that I found with you and some thoughts about how companies get involved in free and open source software

and how that might be affecting projects.

And I don't have the urls for all the studies on here. They are all some where on my blog. I've been trying to translate.

I spent the last week trying to type out my talk so I could put it in an ebook on my blog.

I tried to use Google Voice to do it. I tried to call Google Voice and talk my talk and see if it would transcribe it.

Didn't work.

So one of the first studies that stuck me was one that was done with preschool kids.

The preschool kids, they took two classrooms and they brought in these special markers.

Just regular markers. They told the kids they were special markers.

In the first classroom they had prizes so after the kids finished drawing with the special markers, they handed out prizes for the pictures.

I think like all US sports, everyone got a prize. It wasn't for the best.

In the other class they took out the special markers, they called them special markers, the kids colored,

and when they were done coloring, they did whatever they do whatever they normally do with pictures.

They put them in the recycle bin, or in their cubbies to take home or in the trashcan and no big deal was made of it.

A couple of days later they brought out the special markers again and they put them out in the rooms.

In the first room, the one that had gotten prizes for playing with them the first time, nobody picked them up to play with them.

In the second room the kids picked up the markers and played with them like they did the first time.

So they asked the kids in the first room, how come you didn't color with the special markers?

And they said, but there's no prizes, why would I want to do that?

In the second room they asked the kids so why are you playing with the markers and the kids said

Because it's fun! It seemed pretty relevant to open source software. The next study didn't mean to be a study.

I think I first read about this one in Freakonomics. It was in Switzerland and Switzerland uses a lot nuclear power and they end up with nuclear waste and nobody want to take the nuclear waste.

No towns want to have it so they went around to little towns and they went door to door and they said would

you take this nuclear waste as your duty as a Swiss citizen. We have cheap efficient power and somebody

needs to take the by product. About half the people said ok, all right, I don't want it, but it's my duty as a Swiss citizen, you can put it in my town.

And they thought ok, that's awesome, we got half the people to say yes.

What if we threw in a little incentive?

So they decided the incentive would be something like 6 weeks worth of pay.

Average pay for the area. A month's and a half worth of pay. So they went back door to door.

And said we have all this nuclear waste, you have good cheap power.

As your duty, as a Swiss citizen would you take this nuclear waste and in addition we'll give you like 6 weeks of pay every year

because your town has to put up with this and in that case only a quarter of the people said yes.

Now that they were offered money, it was ok to say no. It wasn't ok to say no I won't be a good Swiss citizen

but it was ok to say no I don't want 6 weeks wroth of pay. So they had kind of given people an out.

Another study. Again with kids. A little closer to home. They offered. This started in 2007, it's still an ongoing experiment in the New York area,

they decided that perhaps they could pay kids to do well in school.

The kids weren't attending, they weren't doing well on tests so they thought maybe we can pay, offer an incentive.

They actually offered the incentive to the parents.

I think total per kid was like $500. So it's not a huge amount of money but I guess depending on your income level it makes a significant difference.

They offered these parents of these kids $500. It has a bunch of things, you get so much for attendance, so they get so much good grades,

so much for doing well on the standardized tests.

They've been running it for 2 years.

They're still claiming they don't have enough data to know if it's working.

But from the data that you can tell on the website, it's made no difference in elementary and middle school

and it's made some difference in high school. And at the time the study came out, it got a lot of push back from people especially from professors, college psychology professors.

ike Barry Swarthmore wrote an editorial in the New York Times saying the problem isn't paying kids to go to school.

You need to figure out what's wrong with school and figure out why it's not fun and make it a fun, interesting place for them to come to.

As opposed to paying them and setting up a whole different standard for awards.

That one has me a little worried because my 3 year old can get through the night without wetting his bed if you give him a piece of candy in the morning.

So I'm kind of wondering when can we take away the piece of candy? Will he pee his bed forever?

The next study, also with kids, I guess kids are easy to study or the studies attract them, I don't know.

This one was an Israeli daycare but I think it's a problem that daycares all over the world face.

They had parents showing up late. They're supposed to come pick up their kids by 6:00.

Parents come in at 6:05, 6:10, 6:15 and the daycare center said, come on guys, we have a life.

We have to go pick up our kids at school. You need to show up on time.

So they finally decided well maybe we can just charge them a big fine and they'll come pick up on time.

So they instituted a fine and the opposite thing happened. It was like the fine made it ok.

If you were paying someone to stay late then you didn't have to feel guilty about making them stay late.

It was like babysitting and you didn't even have to find the babysitter.

So they said, ok, guys, forget that, forget that. Come pick up on your kids on time.

We'll get rid of the fine, feel guilty again. We want to get home to our families.

And it didn't work. People did not go back to showing up on time.

Once they had paid the fine and they no longer felt guilty, the guilt was gone.

They now had this precedence for showing up late.

At can tell you at my daycare it was a $1/minute and the daycare owner got really fed up. She made it $5/minute.

And she says that worked for her, so maybe if you put the price high enough.

Although my boyfriend was working late one time, scrapped a multimillion dollar antenna at work,

was in a very bad mood, was running 5 minutes late, and he walked in and they told him he owned them a fine.

They called me, I was at a conference some where else in the world and said I hope everything's ok, I'm really really sorry!

So they didn't charge us $25 for that one.

So the question is, those are kind of the studies that say if someone does something because they feel an internal reason to do it whether it's guilt or love

or love for markers or duty as a Swiss citizen or guilt for making the day care provider stay late.

If they do it for those internal reasons and then you give them an external reason like money or a fine,

it kind of replaces the internal reason. So my question was, if theese open source software projects

have been really successful, a lot of the really large projects got started as free software projects.

If you look at Linux, it was a college project. If you look at, the Apache Foundation projects, GNOME,

a lot of the really large projects got started as free software projects. And then people got jobs, they got dream jobs.

Suddenly your spouse is happy, because you are getting a paycheck for it.

Supposedly you now won't have to spend so many weekend and evening hours on it because you get to spend 40 hours a week on it and the world is good.

You get paid to do what you enjoy doing.

So what happens if you don't like your boss and you quit or the dot com you took a job with goes bust or your company looses a large client and you don't get to that job any more, what happens?

Will all these people go away? And so, how many people does that affect? I think the numbers in the room are actually representative of this.

For all the studies that I've found show that for most large projects about 40% of the contributors are paid to work on that project.

So like 40% of GNOME contributors are paid to work on GNOME by a company like Novell or Red Hat.

I think when we did the show of hands that was kind of representative here.

There has been a little more pushback that perhaps the people that are paid end up being bigger contributors.

Or more important contributors. Or get to work on more important parts. And usually that's because once you get paid to work on it, you get 40 hours a week or more to work on it.

You're not trying to squish it in after your day job. I haven't looked into that area.

So this affects at the moment, the fact does pay affect people's contributions, affects about 40% of the poeple. And then I had ... show of hands.

show of hands. Almost everyone worked in free software, like 80 or 90%.

How many of you think ... how many people get paid to work on free software. Show of hands?

So how many of you think that if your paid job went away, you'd some how continue to work on free software?

How many people think they wouldn't? So everyone pretty much thinks they'd continue to work on free software?

And so I thought this is a really hard survey question to ask, because I can ask you now. You're getting a nice paycheck.

You're working on something you really love. And if I ask you if you'd continue to work on it, you'll say yes.

Maybe it's kind of like changing a tire. You say of course I'd stop and help someone along side the road if they had a flat tire.

I would stop and help them change the tire.

But when you see someone with a flat tire along side the road, you see about half a million cars drive by before someone stops to help.

So I thought, how do you really find out if people would?

I'm going to show you three things I went through the thought process to try to figure that out.

The first thing is, how does the company change the project. So if you're working on a free software project, and you get paid,

does that company's involvement through you or as a donation to the project, so in GNOME we have people that are paid to work on the project, that changes what they work on.

We have companies just give us money to help fund things that we think are important to the project.

How does that change the project? And then I'm going to talk about .. .3 things.

I'm going to talk about how companies change the project, how they pay, how they pay the project whether it's a salary or a donation,

why developers work on free software in the first place, how that pay change their original motivations and then

how free software, free software's development model is fundamentally different than traditional software.

So I think those 3 factors have to be weighed in. So I'm going to start with like what motivates the developers.

Then you guys can chime in. You're welcome to raise your hand and ask questions or just jump in.

But why do poeple work on free software? What motivated people?

Especially when open source software first came out , this is the question I got most often.

I would go to companies and talk about how great free and open source software was.

And they would go, so why are these people working on it?

You're telling me some guys working on it in his free time at home? What's in it for him?

It's really strange to people not involved in the culture to understand why people do it.

So most of the studies. There's been 2 or 3 studies that worked on this. And they all started out with the premise that people worked on it for external reasons.

For extrinsic reasons.

They assumed that because most people got paid to work on software that the reason people would work on free software had something to do with their external environment.

What they found was that most of the reasons were actually internal.

Most of the reasons that people chose to work on free software were not from an outside influence

but something inside them.

For example, they're intrinsic.

They call them extrinsic and intrinsic if you want to go read the sociology research.

One of the things is that people find it really interesting.

People work on open source software because it's fun or it's challenging.

The joke I like to use here is how many of you are trying to solve the maze.

How many people participated in the logo contest last night?

It's just fun and interesting and that's why people work on it.

It's interesting to see if you can call Google Voice with your whole presentation and have it transcribe it.

Then there's something to fee software about learning.

You have your day job and it's the same old thing you've been doing all the time.

And in free software you get to go out and learn, if not a new programming language, a whole new environment.

You can figure out mobile programming.

It's easy to get started with Android as we were talking about earlier.

You can learn things. The studies have found that for most people, learning is worth about a 20% raise.

So if you were offered the same job you are doing now at a 20% raise or you were offered a new job doing some thing new and exciting, that's about the trade off point.

So if you were offered a 30% raise, you might stick with the old job.

If you were offered a 10% raise, you'd probably pick the new job at the same pay.

And then there's being creative.

One of the things I really like about coming to conferences is seeing the talks and seeing what people are working on.

We saw the ignite talks last night. We saw about robots, we saw about music.

It's really fun to see what creative applications people have done with things.

My favorite still has to be the talk I saw at Jonathon Oxer give.

He'd actually set up his mailbox at home. So when the mailbox at home is opened, it sends him an email.

I really need that.

And then there's the traditional reason that people have given for working on open source software.

Especially when it first came out.

Everyone thought you worked on open source software to scratch an itch.

You worked on open source software because you had a need.

You worked on it because you had a DVD.

You had a desktop and your DVD wouldn't play on your Linux desktop.

So you wrote the software.

Or your kids can't speak so you wrote a speech synthesizer for your computer so your kid could use it to speak with the world.

People thought the main reason that people worked on free and open source software was to scratch an itch.

And I still think it's a very large portion of it.

But it's missing the other ones.

And then people work on free software to learn new skills.

I'm not talking about learning new skills that you can put on your resume. So you can get another job.

I've tried to sell free software a lot that way and it doesn't seem to appeal to people at that level.

I'm talking about learning new skills like you’d take an art class or you'd take a carpentry class.

To learn something new because it's fun.

Kind of like that 20% raise.

It adds a little excitement to your life and you learn something new.

All the studies put these next two in intrinsic and I really think they are extrinsic. To some level.

They said a lot of people work on free software to show that they are clever.

So they work on ... they write the little mailbox thing so that when someone gives you mail, you get an email.

And then you get to go out and give a talk on it or put it on the mailing list and say look how clever I am, look how funny this is.

And to me that's looking an external reward. Looking for applause.

But they valued it as internal. And psychologists must know more than I do on the topic.

One of the things that I think is really key about free software is that it's clever and it's fame, it's recognition, from your peers from other people that have the same values as you do.

(Question form the audience.)

You're right. You know you're clever. You get to finish it and say wow I did it. Good point.

And then there's fame. People do it to be recognized.

It's a place where you can become famous not because of the way you look or the way you sing or how many friends you have.

You can become famous for what you can do. For what you can accomplish and the kind of code you can write.

You don't even get to see the other people until you come to a conference like this.

Any thoughts on which of those really appeal to people? Why you thought you got started?

How many, show of hands, how many people got ...

Whoops when back to the beginning ... you get a review.

This is how I practice my talks.

How many people work on free software because it's interesting? At least half of you.

How many people do it because they are learning new things? They could make more money doing something else?

Just about everyone. How many people do it because you get to be creative and come up with cool solutions? About a third.

I'm totally guessing at the number of hands. How many do it to scratch an itch?

You had a problem and you are trying to fix it? A third? Maybe half. I see someone waving like this at the back.

How many people do it to learn new skills?

Learn new programming language? Learn mobile environment? Maybe a 1/3 to 1/2.

How many pople do it becuase you feel clever? Or it's kind of fun to solve the problem at the end? About 1/3. 1/4 to a 1/3.

How many people do it because they want to be famous in the open source software community?

Nobody's going to admit to that. One guy. Two. (Laughter)

What's funny is that a lot of people know my name in the free and open source software community.

But my friends outside of it, don't know, they think I work with this strange thing where software is kind of free.

Which doesn't make sense. I explain it to them over and over again. I had a friend google me the other day. You're famous!

(laughter) It was kind of fun.

How many people here work on free software to do the right thing?

Because software should be free, should be open, should be available to more people.

I left this one out when I was first making my list and a lot of people came to me and said one of the reasons they work on it is to do the right thing.

One of the studies, I actually went and searched, and found a study that said

1/3 of open source software developers of free and open source software developers work on free software because they feel software should be free.

So obviously Richard Stallman is definitely in this camp. The whole idea that ... what?

We need a few people on both extremes.

I do think that by having free software we as a society are going to get further faster.

I used to manage the CDE team at HP and the thing that was so frustrating

Now I’m on the GNOME side, the open source side.

The thing that was so frustrating was that there was all these companies that were part of this standards body.

I don't remember how many of us there were, 5-10 companies, and we all wrote exactly the same thing to meet the standard.

We all had these teams of people. I had 30 people working on it. And so did IBM, so did Compaq.

They all had the same amount of people working to write the exact same software.

And nobody would have said CDE was a competitive advantage for HP.

So that was really frustrating. It felt like I was back in college.

And the whole room of a 100 people was writing the exact same program.

So those are the developer motivations. Very internal, very intrinsic, not external reasons.

Not about money for the most part.

How many people work on free software because they want to get rich with free software?

A few hands. Everyone wants to get rich.

But it's probably not the reason you got started. Could be.

The other thing to look at is the type of payment.

It turns out the way that companies invest money into free software projects makes a big difference.

And there's studies on this.

There's studies on how payment affects not only free software

but how payment affects motivations.

So the first thing is, is payment normal for this

So for example, it's not normal to get paid to go to church

It's not normal to pay kids to play with markers

so when you pay for that type of thing, you very much change the motivations that a person had

You might even be attracting people that never would have done it.

You know, if they paid me enough money maybe I would go to church every single Sunday.

Maybe if you paid kids enough money they would play with boring black markers instead of colored markers.

If you explained to them that money was going to buy them something else.

But on the other hand, it's a pretty normal thing to be paid to work on software.

So it's less it's less likely that paying someone to work on a free software project will change their motivations.

Because it's normal. You kind of expect to be paid on software.

So it's not strange.

Also it turns out that it changes people's motivations

if you pay for performance or completeness.

So for example if I pay you to write a new feature for Gimp,

that's different than if I just pay your salary to work on Gimp.

And the paying for feature actually changes your motivations more

than just getting a salary to work on the project.

(question from the audience)

Negative way. The salary is better than the feature or the performance or the job.

Then there's employment.

So lots of companies employ people to work on free software.

From what I can tell this is the most benevelant way to put funds into free software

of the last couple that I've said.

It's you pay someone to work on a project.

Companies vary from whether they pay someone to work on a project

or pay them to add company specific features to a project.

And then there's in-kind.

This one has been really popular but I think has not been really figured out yet.

So you see this like Google hands out developer android phones.

So this is handing out hardware or maybe even handing out swag so people will like your project

and work on your project.

So for example when I was at HP, Grant Grundler, who is now at Google,

we were working on getting Linux onto PA Risc boxes

and there wasn't a huge amount of community people

that were really interested in getting Linux working on PA Risc

so what we did is we offered to send free PA Risc boxes to anybody who wanted to help

We sent new ones. We scrambled around the lab and found old ones under the desk.

I lost all of my footwarmers.

We sent them around and when he first started.

When Grant first started sending them, he found the return rate

was like 10%.

So if he sent out 10 machines, he would get one developer

who actually delivered and worked on getting this stuff work.

So then he started getting a little more picky and interviewing the people

and getting their commitment before he sent them a machine.

And he raised the rate to 30%.

So that's still not huge.

It depends on how cheap hardware is for you.

But there's been people that are doing this with mixed results.

Then there's bounties.

Bounties also have kind of a mixed history.

So the idea behind a bounty

is that you pay for a feature that you want

and you tell on the web or on your website or on the project website,

here's the feature I want and here's how much money I'm willing to pay.

GNOME did this a long time ago.

We didn't have a lot of success with it.

We wrote up all the things that didn't work well for us

and we haven't tried to repeat it.

Funanbol, the company that creates mobile software for phones is using bounties and it's working really well for them

Elance.com works very well and you could argue that their system is really a bounty.

They say what they want written, you do it, you get paid.

The problem with bounties we found is that you have to be very clear with what you want up front.

So sometimes by the time you finish specifying exactly what you want up front, you might as well have

written it yourself.

Depending on how good you are at writing specs.

A lot of the bounty systems when they first started let as many people as they wanted, just do it.

And it was like the first person that came in, won.

And that's really frustrating to everyone.

On the other hand, we found in GNOME,

if you pick someone, if 5 people apply and you say

ok you can have it, and that guy never comes back,

never writes the software. At which point do you call it quits?

And let someone else work on it?

We had a couple of features that took much longer to get written because

we had assigned it to someone.

And then there's grants.

I think there's a lot more room for people to apply for grants in free software.

I actually interviewed ... oh, I spaced his name. He's on my blog.

I interviewed him. His whole career is applying for grants

to do projects. And he builds in his fees, it's a grant

for a specific reason and he builds in free software to do it.

So his whole income comes from writing software with free software based on grants.

I think there's a lot more room for other people to do this.

For example, Wikimedia just got a huge grant from the Ford Foundation

to help fund, work on their project.

GNOME has just applied for a couple. One to do with usability.

So there's funding out there to do real work. And the organizations funding it

like the Ford Foundation, they don't really care if it's free software or not.

You write what you're going to write and you can talk about how it's free software.

But you talk about the good that the project is going to do.

That's what they want to fund.

So those are a bunch of kind theoretical studies

and some real world kinds of ways of paying but

it would be easiest to answer the question of would you do it again for free

if it had actually happened.

Is there any project out there that we can look at where this has been true?

And, the only project that people pointed me at, that I found,

that is really truely representative I think of the issue is

Nautilus is the GNOME file manager and there's a company called Eazel that used to work on Nautilus.

And they had great plans for making lots of money off of Nautilus.

It didn't work.

The company is gone.

Nautilus is still there in GNOME.

It's a much smaller project than it was when Eazel employed a bunch of people.

And so I thought it should be easy to go find these people that worked at Eazel

and find out where they went.

So I don't have a list of the people and where they went

but anecdotally I've heard that everybody that was in the free software community

is still in the free software community.

Maybe not working on Nautilus.

But Eazel also employed a bunch of people that hadn't worked on free software before.

And those people have all gone off to other things. To non free software things.

Rumor is they all went to Apple.

Probably these days they all went to Google.

But they've gone on somewhere else.

So that really brought on the theory that I think answers the question

with free software.

I think I have a slide.

I think the thing is, if someone paid you to work on a project,

and then that project goes bust,

continuing to work on that project that someone believed in so much

that they were willing to pay for it

and now they are saying it's not work that paycheck.

I think it's hard to continue working on that project.

So I think most people that want to stay in the free software world

what they would actually do, they would continue to work on free software

but it would be a different project.

One that is still valid and valued by others.

How many people does that resonate with?

If the project you're working on, got paid to work on,

you no longer got paid to work on it, but would you continue to work on free software,

would it be that project or a different project?

How many people say that project?

How many people say they'd work on a different project?

You like the project you are working on?

Yeah, you're probably going to interview for jobs and look for another job.

Assuming you have to eat.

The other thing that really stood out is that, some of the research point out,

is the equality between who you pay and who you don't pay.

For example, someone left a comment on one of my blog posts on this topic.

They said they'd been at a company and they needed to work a lot of overtime.

And one night when they were working all their overtime

they found out that some of the people were getting paid over time

and some of them weren't.

So obviously at that point, everyone that wasn't getting paid, just quit.

Because you're not going to work when some people are getting paid.

I think that's very relevant to free software.

Because if you look at the GNOME community,

40% of people are getting paid and 60% aren't.

Those 60% aren't leaving, and why not?

It's a question to be asked.

I had talked to Mitchell Baker from Mozilla

and they said that, so Mozilla was pretty much all free, well it has a history,

but a lot of people were working on it for free

And then they started the Mozilla Corporation.

And she said the number of volunteers didn't go down.

That 30% of all contributions to Firefox are still by people that are not paid by the Mozilla corporation.

So this doesn't seem to affect the open source software community as much as it could.

So if people, this is my point, if people stopped working on that project

because you know, no one is paying them to work on it because the project isn't

worth as much, where do they go?

I really think they go on to other software projects.

I think part of this is ... how many people here work on just one free software project?

How many people here work on two software projects?

Keep your hands up as I go?

How many people work on three or more? Keep your hands up as I go.

So, yeah, you're working on a number of them any way.

So you're always kind of turning them over. One drops off.

You add another one to it.

So I think people tend to go to other projects.

As opposed to leaving the free software community.

But if they go to the other projects, is it really the same?

Is it really ... when companies come in and invest in free software,

do they change the project?

Whether or not they demotivate all these developers if they go bust,

while they are involved in that project, are they changing that project?

And bear with me here, I'm going to give you a little history

of the world in two minutes.

So at one point in time we had hunter gatherers

and supposedly everyone spent all of their time trying to find food

whether it was gathering up fruits or going and hunting animals

Now since then I've heard that the only hunter gatherer society left on this planet

that's truely hunter gatherer, they actually

have a ton of free time.

They actually spend a lot of time sitting around talking and playing with their friends and their family.

So I'm not sure about this myth that we've heard that they've

spent a 100% of their time looking for food.

They might have had it better off.

But then we introduce agriculture.

And what agriculture did was allow us to raise a lot of food at once.

And so it freed up ... so first of all

it allowed people to live in towns so they didn't have to live all spread out.

Where they could find the food.

Now all of a sudden you could have a field that could feed a whole bunch of people.

So you had people gather in cities.

So it freed up maybe 10-20% of the people to do something else.

That wasn't looking for food.

So when you had this 10-20%, you had

people, you could start trading.

You could trade your milk for your eggs.

You make trades. And you started having people like accountants.

You started having people like the right brain thinkers.

Then once you had enough people trading things, you had arguments.

So you had to have attorneys.

These lawyers would come in to settle the debates.

Then you had people that their families got rich, so they could afford to sit at home

and think about things. Or they could send their kids to school.

To learn about a broader array of things.

And we developed a lot of things.

Electricity and all that stuff.

And eventually we got to have computer programming jobs.

Told you it was going to be a quick history.

And so, just recently, and if you read literature,

like Thomas Friedman "The World is Flat",

Dan Pink's "The Whole New Brain",

there's a number of them that are coming out,

kind of the pop culture kind of books,

but they are saying that now is really the time of the creators.

So we had left brain thinkers before.

We had people that were accountants or attorneys

and those jobs were really valued. They paid a lot of money.

They still pay a lot of money.

And then we have people like artists and musicians.

And while we value them as a culture, we don't pay them a lot of money.

Now in the last 10-20 years, we're really seeing those two converge.

And I think you see it a lot in the free software world.

So instead of getting paid just to write code, you are actually creating a whole solution.

You don't have someone whose job it is to define the product,

and someone who's job it is to talk to customers,

and someone who's job it is to write code.

One person gets to be that whole thing.

And if you see a problem like you don't know when you are getting mail,

you can write a clever solution.

Someone didn't spec out the requirements of need to know when you are getting

mail so you need to know when the mail box opens

so you need to put in a sensor

when it closes, no one mapped all that out.

He did it all himself.

So we have people in the free software world who are not just programmers,

not just left brain thinkers, but they are also right brain thinkers

and they are designing and creating solutions.

So you have these people that are designers.

In the free software world, they got to decide,

you get to decide what GNOME looks like.

You might argue with a whole lot of people about whether the icons should

be in the menu or not.

But ultimately it's the coders that decide that.

And then they get their dream job working at a company

getting paid to do that.

And all of a sudden they have to do things the company way.

They get to go to meetings. They even get to go to teleconference meeting that lasts several hours.

How many people have attended a teleconference meeting for an open source software project?

Not many.

There's a few.

At a company all of a sudden you are attending a lot of meetings.

And more importantly, to me, is that

the process is no longer one that the person that is writing the code,

who is using the code, who is talking to the users,

figuring out what the product looks like.

You go back to the company method that we've been using for however long software has been around.

Where there is a person that talks to the customers,

decides what type of product they need,

the product manager specs it out,

writes up this whole design spec document that describes what it's going to look like,

and then they hand it over to the engineering team,

the engineering team developers it,

it gets handed off to the QA team,

the QA team tests it,

and then it gets handed off to customers,

and it's the customer support guys who answer the calls on teh product.

So now instead of having one person that kind of encompasses all of those roles,

you put people back into boxes.

And I think we lose something.

We lose the very thing that's made free software successful.

So we've got this, people that are used to doing it the free software way,

who've made these products,

that companies are very interested in, that have proven to be very successful.

But then when companies adopt it, they try to fit it into the company methods.

And there's a few companies now that are really ... that have learned.

That are really trying to work with the free software world.

But traditionally when they take the free software they try to peg it into the old ways.

And then people get really frustrated.

So how many of you here have had an argument with your manager or your product manager or your design

person

about the why the way they want to do it is not the right way.

Yeah.

And in the end, I've heard managers say, well you're going to do it this way because I said so.

To me that is kind of killing ...

the reason free software works is because those arguments were allowed to happen.

Those arguments happen.

They happen on the mailing lists, they happen on IRC.

They get very heated.

Sometimes they divide communities.

But it added something to the process when

you had those coders as part of the decision process.

Especially at large companies, that's not the way the process works.

And I actually had a friend who was working at a company and

he had (and I've heard a couple of stories like this)

He had a customer who contacted him and wanted a feature.

So he told his boss that he was going to work on the feature and the boss said

no, no, no, we have to put out a release next week.

We've promised. We've taken an order for $100,000

from this other customer to have this release out next week.

You can't fix that feature now. You can't do that.

So he came in and worked all weekend to work on the feature for the customer.

On his own time.

Monday morning his boss gets in. He says, look! I did the feature for the customer.

That customer is happy.

Now I can work on the release and everything will be good.

And his boss said, if you were going to come in here and

work all weekend, you needed to be working on what was important.

And I don't think the boss was a bad guy, I know he wasn't a bad guy.

But it's this ... it's a crash of the company is getting paid to do certain things

and it has a process to get those customer requirements in.

Where as in the free software world, customer requirements

come in through bugzilla or IRC chat channels or mailing lists,

so there's a different process.

So I definitely think that companies are changing the projects when they get involved.

You have these people that have gotten these projects to the point where it's at.

And now you have these companies come in and trying to participate in these projects

It's really exciting.

It's up to us to help corporations do the right thing.

It's up to us to explain that having discussions on mailing lists in public is a good thing.

It's really hard for a company that keeps all their secrets ...

all their information is proprietary information, especially customer information,

to have all their discussions on a mailing list with a bunch of random people.

To me the strangest thing about a mailing list is I don't know who's on it.

That really bothers me, still, when I write to a mailing list.

I don't know who all is reading it. I can't see the people I'm talking to.

For companies it's even that much more scary.

So we need to help them see how IRC chat rooms,

where you are available to people to ask questions at any time.

Or where having discussions on the mailing list where lots of people can give input,

even if you don't listen to it necessarily, how those things are good.

So my call to you guys is to show companies ...

not to go in and tell companies they are doing it wrong.

Not to go to your boss and say no, no, no, I'm not going to do it that way.

But to show them success stories.

So we're working on this in GNOME.

On showing here's how the decisions got made.

And we can write that up and say when we decided ...

One of the things we haven't done well with GNOME 3 is showing how we decided to make GNOME Shell and

Zeitgeist. But if we can write up that story

and show how we got there, we can show companies how that public decision process works

and how they can use it internally working with communities.

So this is my pitch at the end of the talk.

And I'm here for questions if anyone wants to ask.

And I really like to hear people's stories if people want to share them.

One of the things I'd like to see us all do as a community,

is support the free software projects ourselves.

So one of the ways that companies influence free software

is giving them money.

I work for the GNOME Foundation, my salary is paid by donations from large companies

so I know this for a fact.

But I'd really like to see us all support free software projects as individuals.

Everyone gives to charities so if everyone picked which projects

they were going to give a $100 to or $200 to every year,

we could be a completely independent organization.

We'd still take donations from the companies but we wouldn't

... our financial where withall would not be depend on them.

So my call is go out and pick a free software project that can accept money.

Not all of them can, not all of them know what to do with the money that they get.

I have no problem helping people figure out how to do hackfests or something.

But pick the Free Software Foundation, GNOME Foundation, Apache,

somebody and donate at an individual level, not just at a company level.

At the GNOME booth, I'm showing the tshirt that we give out to people that join

our Friends of GNOME project.

We're actually raffling or having a contest for that one.

Take a picture of yourself and GNOME in some way shape or form

upload it to Flickr to the tags that were mentioned before

and we'll be handing it out on Saturday.

So that's ... any questions about the talk or about supporting free software?

Audience: Can you give an example of a free software that's moved to a company situation

and still kept the same format?

A project that was bought by a company?

Free software that got donations and still kept that same type of liberty.

There's a lot of projects that now have a huge corporate influence.

Like the GNOME project has 40% of its contributors paid by companies.

GNOME and Linux have multiple companies that invest in them.

One that was probably like that was MySQL.

So MySQL was started by a company but they had a large community of people that worked on it.

And they eventually hired most of their community.

So anyone who had written anything significant for MySQL, got hired by MySQL.

I don't know long term ... it's a very successful project.

Being bought by Sun adds a whole new angle.

Being bought by Oracle adds a whole new angle.

So I don't know what will happen long term.

But that is definitely a project that was started a free software project

and had a huge corporate influence.

Yeah, was it started by the company or was it started as a free software project?

It was free, and then they started a company.

Oh, the other one, they're in Atlanta.

Drupal guys too.

Same thing, they started.

So Drupal is an open source software project, like WordPress,

and they both have open source software projects associated with them.

It's probably been a couple of years and they are both very successful.

(Audience) So something I noticed as I was visiting India,

companies that are bigger than just free software hackers

how do you keep the company ecos,

you start to hack for whatever reason,

you grow into a bigger company,

and you have all these people that work on it just for the money,

how do you keep those kind of conflicting view points

and everyone moving forward and all excited about the project?

(Stormy) I think personally if I had a company

I wouldn't hire anyone that was in it just for the money.

I think that's true of any company, any product, not just free software.

When you have a company, where you have some people that are just coming in for the paycheck,

your company won't be as successful.

Granted, everyone needs a paycheck to eat.

But you should be hiring people that they bought your prodcut at home because they thought it was cool.

When you work for a company that makes some kind of Unix product

and all your employees have Macs at home.

You need to stop and think, are you making the right product?

Are you letting them affect the product?

Why doesn't it do what they need it to do?

You need to always hire people that love the project.

But I think the key, your real question is, you need to make sure

the organization is set up right.

For example, with community mangers there's been a huge debate

of should community managers be in the engineering team?

the marketing team? the customer support team?

I think by picking where they belong, you make a huge difference in your company.

(Audience) Where do they belong?

(Stormy) I think ... it's a huge debate.

I think it depends on the life cycle of your project.

In the beginning they belong in engineering for sure.

Maybe they can move to marketing later.

But they have to live close to the project.

Especially in the beginning, a lot of the community things is problem people have with the product.

I also think engineering should have to do some kind of stint in customer support.

Just cause.

(Audience) In your research, when you found a company that lost, was open source

had open source, and they stopped,

did you ever find that people quit and then came back to the project?

(Stormy) I do think that the founders and stuff stick to it much longer.

I don't have any concrete data that shows that.

But I agree, the founders definitely usually care most.

(Audience) In your presentation, you said that companies that paid for features

often had a more negative impact than companies that paid salaries.

Can you explain that it a little more why that is happening?

(Stormy) Yeah, it didn't say it had a negative impact.

It's just not as benign as paying a salary.

I think the main thing is you work on free software because you want to,

you get to decide what you work on, you get to decide how to design it,

you have full control and you get to work within the framework that has

proven successful for free and open source software projects.

So you continue to work in that same framework.

So the paycheck doesn't change how you work or how you feel about it.

When you pay for a feature, now all of a sudden you're getting paid

not necessariy for something you believe is the right thing to add.

I'm going to give you $10,000 to add this feature, if it's $10,000,

and you need money, you probably aren't going to argue with me too much

that this feature isn't the right one to add to the project.

You're just going to write it. So I changed your motivation,

I changed what you worked on, I changed how much you would

argue back with me about whether it's a good feature because I'm just going to pay you for it

as opposed to asking you for it, so it changes what you work on and why.

(Audience) Have you or anyone looked at Google Summer of Code

and how effective it's been for the projects and participation?

(Stormy) Yes, so I've asked Leslie Hawthorne from Google for the numbers a number of times.

I'd love to see the number of students that participated, and obviously it wouldn't be 100%,

but I'd love to see how many stick with the project.

Also, the GNOME project is looking at our personal ones.

We do get people who go through Google Summer of Code and become GNOME contriburors and stay.

I don't know the success rate but we have successes.

That's very much ... it's almost a different question.

It's not were you working on it for free, you got paid and you no longer got paid.

In that case they attracted people that weren't working on free software

and they paid them to join the project.

And I actually think that's a harder sell.

And I actually think that's why we see less women at these conferences.

Because more men tend to get into free software on their own time

and then get paid to do it.

Where as I think if you take all, and I haven't done a survey, we are going to do a survey,

if you took all the women in free software, and asked them

how they got involved in free software, I think

a huge percentage of them got started as part of their paid job.

So they didn't start working on it first and then get paid,

they got paid to work on it.

(Audience) I think you might have just answered this.

Do you have any relational data to women in technology industry versus in specific open source.

(Stormy) That's interesting. I hadn't thought about it that way.

So the National Center for Women in Computing is going to be

doing a survey about women in free software.

And that's a good question to add.

So I don't know the answer to that.

My guess would be that 80% of women in free software are paid to do it

as opposed to 40%.

I'm just totally guessing.

(Audience) Professional expectations. My father was a lawyer.

When he started out there was the expectation that he would do probono work, a certain percentage.

Not a seperate job for people handling those cases.

I'm starting to see some of the non open source conferences,

that you ought to do some open source just to be professional.

What do you say to that?

(Stormy) That's a really cool movement if it works.

I haven't heard anything about it but I like the idea.

(Audience) It's seems to me that every project or every job has unpleasant or dreary tasks.

When you run up against those, you say that's why they're paying me,

so that's why you keep working.

Who does those tasks in open source and why?

What motivates them to go through the drudgery?

If they're not being paid.

(Different audience member) Some of them don't get done, like documentation!

(Laughter)

(Stormy) So some of them don't get done.

But at the same time, the GNOME documentation team is not paid to work on documentation.

And they just had a hackfest recently.

So I think some people just know that has to get done.

Or you believe in your project enough, you spend some amount of time pushing through it.

I also think luckily lots of people have different people have different interests.

So the documentation team in GNOME actually likes writing documentation.

So I think the myth that you have to pay someone ...

I think it's a myth that you have to pay someone to do the drudge work.

I think what actually does is that you have to sell the vision.

If you can sell the vision, and this project is really important,

and people in Africa are not going to be able to use our product,

unless we do all the localization bits so it can be translated,

I think you sell people in the vision.

As opposed to paying peope to get through it.

Kind of like the New York City school kids.

Instead of paying them to get through it, they should be motivating them to show them why it's important.

Way up there.

(Audience) Just to add on that,

When I was a kid, ....

you spread that load across everyone.

...

everyone does a little of the not so fun stuff to do then ...

(Stormy) And can you do it in fun ways.

Like the hackfests have proven successful for us.

I think openSUSE does like a hack week where everyone doesn't get together in the same place

but they all work on the same tasks that week.

Any other questions?

(Audience) A note on that.

Some of the unpleasant tasks have rewards.

I spent two days doing something that's ... but in the end the program ran five times faster.

(Stormy) Definitely.

(Audience) This goes back to the documentation thing.

I don't know how many people have heard of ... it's a PRP system, pretty complex.

There's no documentation out there.

So it was interesting.

A lot of developers consider documentation, well I'd rather be coding than writing manuals.

But there's people that are the opposite.

They'd rather be writing documentation than code.

So a lot of it's we in the community have this responsibility to not just talk to coders

about this open source software project but also talk to

people that can help in non coding ways.

For example, I went to University of Idaho near where I live and talked to professors

and they have it set up so students that are doing internships with ...

and writing manuals.

A lot of it's our responsibility to spread the word to noncoders.

(Stormy) And pulling on your user base a lot helps that.

Your power users are often most frustrated with the lack of documentation and willing to help out.

(Audience) 15 years ago a lot of the innovation in open source was

mostly unpaid. If you do a trend from 15 years to now

we're at 40%. If you extend that curve to 15 years from now

are we going to end up in the scenario where 90%

of the software is paid?

And not the same innovation or quality?

(Stormy) So I'm not saying companies decrease the quality.

I'm saying they're learning how to work with free software.

So we need to help them learn.

So I don't think they necessarily decrease the quality.

My key question was if they stop paying, will the people go away?

So will the number of the paid contributors go up? 0 to 40, will eventually everyone be paid?

I hope everyone gets their dream job some day.

So if what people really want to do is write free software, I hope they can find a job paying for that.

I don't think it will be all large corporations.

Right now I think the 40% are mostly large companies.

Like GNOME sponsors. Novell here employs a lot of people to work on free software.

So I don't think it will be 90% of people are employed by large companies.

I hope we can come up with some more models,

like small companies. Like small support staff

or local industries, that are free software jobs.

The grants, I'd really like to see more people writing grants to be able to work on free software projects

with money from foundations that support good works.

I don't know if we'll ever get to 100%, I don't think if we get to 100%

it will be bad for the free software community necessarily.

I don't know where we'll end up.

I do know that most of GNOME's new contributors come from universities.

And the students aren't paid.

So hopefully the trend of people starting to work on it

because they find it fun will continue.

(Audience) What is the value of development turnouts?

I work for Miguel de Icaza and he's a VP of big company.

He'll meet other VPs and they'll ask so what paradigms are you using?

We're trying to implement scrum over here.

(Stormy) He's using the open source paradigm.

(Audience) Is it a paradigm?

(Stormy) I think so.

(Audience) Or is it a lack of paradigm?

(Stormy) I think it's a paradigm.

And it's more than how release happen.

I think how the communication happens is like the really key points.

At least coming from proprietary software development to open soure software development,

to me one of the key pieces is how the communciation happens.

I think when people start working in free software,

the biggest shock is that there are no meetings,

there are kind of meetings, but they're not everyone gets in an office and talks.

All of the decisions and all of the discussions, including the disagreements

are very public.

There are mailing lists where you can search the archives

and find everything that has happened in that project in the past.

That's huge.

How many companies have a place within the company where you can go and search

and find out how the company made it to that place.

Why the product has that feature.

I know you said that you couldn't find the decision you were looking for the other day.

But generally in the free software world there's a whole archive.

I actually went back yesterday, I wanted to see how something had went in the advisory board at GNOME.

And we had email all the way back from 2001 when it started.

And I found things that I had participated in that I didn't even remember writing.

So those types of things are really powerful

and I think they make a paradigm.

It's not a formal process where if you do the following ten things you get

certified as a free software development process.

But those models do form a paradigm that's very powerful.

Maybe if we documented them more, people would understand them more and adopt them more.

I'm around for the morning, so if anyone wants to grab me afterwards, I'm here.

Thanks very much.

(Applause)