Proposal to Fund Debian Sparks Debate

lisah writes "The announcement earlier this week of 'experimental' group Dunc-Tank's plans to bankroll the work of certain Debian developers has sparked some controversy across the open source community. The leaders of Dunc-Tank say their primary motivation is to see that Debian version 4.0, also known as etch, is released on time this December. Debian developer Lucas Nussbaum, however, says that research shows that 'sometimes, paying volunteers decreases the overall participation.' Dunc-Tank member Raphaël Hertzog countered that the opposite is true and 'many Debian developers are motivated to work when things evolve,' a veiled reference to Debian's notoriously slow release cycle. Dunc-Tank member and kernel developer Ted Ts'o took the idea a step further and said, 'If money were among anybody's primary motivators...they probably wouldn't be accepting a grant from Dunc-Tank; they could probably make more money by applying for a job with Google — or Microsoft.'"

No no no

[1310nm]

Debian is one of a very few of the major staple distros that hasn't been taken over by greed (see RH (RH), Novell (SUSE)). I really like the fact that the Debian I use is the same Debian everyone else is using, not a development playground or redheaded stepchild money pit.

Money isn't Everything...

[denmarkw00t]

but is something, and that something is, well, money.

I've rarely seen a better motivator for getting something done - especially in a timely manner - than money. If I'm volunteering with children or for a good cause (no, I know - Debian is a good cause too, but you know what I mean) then I'm going to do my best regardless because I feel like I'm helping benefit people who are less fortunate than me. However, if I'm working a job to maintain myself (and possibly my family) and I'm volunteering to develop a large open-source project and not getting payed for that extra work I do when I get home or when I'm up late at night, then a little money can go a long way.

I don't think money would cause those being payed to work less at all, instead I think we'd see an increase in both the timeliness of development and the quality of code in the next Debian release.

What happened?

[Frogbert]

What has happened to Debian of late? I'm the first to admit that I don't follow the politics of the linux scene with anything more then a passing glance but the current Debian team appears to be disolving into a clusterfuck of massive egos clashing about trivial changes. Wost still it seems that they end up bitching and debating more then they actually spend doing something. The whole situation reminds me of the People's Front of Judea, fighting with the Judean Peoples Front. They just debate endlessly and end up doing nothing.

Nonsense

[EmbeddedJanitor]

You speak of things you don't understand.

Highly motivated people can often not devote as much time as they would like to OSS because they have to go to a regular job to pay for food etc.

There are a lot of key Linux developers who provide huge benefit to the community, but would like to make it pay so that they can make a fulltime job of it. Go look at what some people like Hans Reiser have to say http://kerneltrap.org/node/5654 [kerneltrap.org] "Doing GPL work is doing charity work in our current legal and economic framework. That should be and could be changed, but for now it is so. I have done my share of charity, and I would not have a problem doing proprietary work.", and http://www.namesys.com/ [namesys.com] "For free software based on support revenues to be viable, people have to be more inclined to use our support service than they are to use the support services of persons who bundle our software with what they sell. Frankly, they are not, and this is why providing service on free software is failing as a business model for producing free software."

For my own part, I write OSS that saves people literally millions of dollars per year, yet I can only treat it as a hobby because it can't pay my bills.

Hopefully at some stage people start **paying** for stuff that is valuable to them. Unfortunately people grab what they can get for free.

Having good roads is very valuable, and you would not have those if they were not paid for. They are typically paid for by taxes because most people would not voluntarily dip into their pockets to pay for roads etc.

I think any methods that help get money into the hands of **key** OSS developers is a good thing.

Re:What happened?

[glwtta]

There does seem to be a lot of actual development activity as well. I wonder if the people bitching and the people doing work are, as usual, different people?

Pay for the boring stuff

[Qwavel]

Paying selected developers could cause problems.

Instead, use the money to ensure that any developer who wants to contribute has a good experience, and to get the stuff done that no developers want to do. For example, you could pay people to do testing.

Re:How is this any different

[Saxophonist]

I work a lot with Drupal and see this on the message boards often. "I'd like to see this feature built and I'm willing to pay XXX for it" Someone builds the feature and cashes in. Innovation and capitalism at work.

Some open-source projects have seemed to operate almost entirely on this principle. Take, for instance, LilyPond [lilypond.org]. Development for some time seemed to be done almost entirely by core developers who seemed to be getting paid for custom features. Spending time on these custom features, though, meant that other, more basic features were sometimes missing or lacking. (I wish I could give examples, but it has been a few months since I have used LilyPond.)

Now, the LilyPond site seems to emphasize the involvement of other developers, documentation writers, etc. [lilypond.org] There is a FAQ item that leads to a page about sponsoring features [lilypond.org], but I wonder if the focus has shifted more toward getting other volunteer contributors. The "call-for-help" page cites these reasons for wanting help:

Hopefully, together we can address problems in the LilyPond development process, among others

* Stable releases don't happen often enough.
* Development is too much centralized.
* The learning curve is too steep.

I would say only the first reason really applies to Debian, but it is interesting that LilyPond seems to be taking the opposite approach to solving the problem.

Re:Nonsense

[Monkelectric]

You are right, but I would like to add a management perspective (although I am not a manager), I have worked on at least one very large OSS project. When you bring money into the equation -- you have to make a judgement call about who is most important, and who deserves money the most, and that inevitably pisses people off. There really is no fair way to distribute money in an OS project. So you loose people, and the project goes slower.

Be careful with bounties

[EmbeddedJanitor]

They can lead to competition which can get unhealthy. Instead of collaboration, you see people hiding info because the other bounty hunters might use it to get ahead.

Re:Nonsense

[EmbeddedJanitor]

Or people don't have the time to devote to the project, so the project goes slower anyway....

I don't think payment is a solution to all developer problems, but it would allow some **key** developers to be able to do their OSS stuff fulltime instead of just part time. If you have a full-time job + family, then you can only spend x hours per week on OSS before you get fired or the wife kicks you out or whatever.

Why pay?

[dragonquest]

You could achieve the same level of motivation by a Karma system, works here doesn't it. :-), but seriously, people would be happy just to see some credit given even if its in terms of silly awards or fun titles.

No problem

[atokata]

You've gotta pay people to work on deadlines sometimes. I've done volunteer work which has led to paying work, driven by nessecity. Most volunteers have day jobs, and aren't extremely wealthy. If they're being asked to put in a lot of hours, it's only fair to compensate them for the time they can't spend working at their normal job, be it freelancing, a normal desk job, or whatever.

Re:the office

[ex-geek]

I think you misunderstood Nussbaum. I believe he was trying to say that once some volunteers are paid, the other volunteers lose interest. I've witnessed this first hand in an originally volunteer based NPO.

Re:Nonsense

[QuantumG]

Hopefully at some stage people start **paying** for stuff that is valuable to them. Unfortunately people grab what they can get for free.

You're the one handing it out for free, what do you expect?

Try eating karma

[EmbeddedJanitor]

Maybe RMS knows how to make a Karma sandwich, I sure don't.

Karma is overrated. Sure you can get a buzz to know your software is being used all over the world by hundreds of thousands of people, but it's far easier to get a buzz out of knowing that while you're driving around in a nifty new car paid for by your earnings.

Primary motivators

[Anonymous Coward]

If the primary motivators were money, than there would be no Debian. Businesses who spend money will eventually (at least on the long run) try to get more and more influence untill the point they explicitly want bang for their buck and slowly forget the power of the open source community. I think that open source is more about that people create what they think is missing in some piece of software. Most of the people who want to spent their free time don't like the idea of a business telling them something to do.

For the developers who spend so much time they have little time left to make a decent living for themselves: those precious people have forgot to set the priorities in life. Money is not the answer, personal basic priorities are. I think the success of a open source product is more related to the community than dependence on individuals or companies.

What would be realistic expectations when you spend some developtime in some product? Not money, only some functionalit and if lucky also a developer community. Expecting money is unrealistic for the open source developer, nor does it contribute to a better community. If people tease with cash ans start complaining about the slow development of Debian, they should start using another product.

Re:Agile Vs Debian

[marcello_dl]

I want the apps in debian stable to be bug fixed not updated to the latest versions. And debian does that already. I don't want a newer version to ship with modified configuration files and option and having all the debian servers require to edit the configuration files just because the new config has a different default value. There's nothing more agile than debian stable updates when it comes to upgrading. Security updates often run as a cron job if one is not concerned with testing them before deploying. If you want agility as in frequent updates go the debian unstable route, it's not like the box will crash often. I use unstable for a xen hosted server and it never went down unless the whole host did, so it's either very stable or is capable of taking down the host: both things seem very impressive to me :D. You need to know a lil about apt upgrade vs. dist-upgrade to do so. I also don't get how other distro which ship big updates like fedora and IIRC ubuntu, are more agile than an unstable distro which has 100 mb of updates a week to be kept current, and updating is painless 999 out of 1000 times.

Re:Nonsense

[Kjella]

You're the one handing it out for free, what do you expect?

Imagine for a moment, that you were working on an extremely specialized OSS project that only one company could profit from (absurd, I know). Since the code is free, they needn't pay you anything. However, you tell them that you need to put food on the table, and that to do any future work he will need pay. What can you expect of pay? If the expected value of having that developer work on that project is worth $X to the company, they should rationally be willing to donate up to $X voluntarily (ignoring some details like risk premiums). Why? Because the return on investment is good. Obviously in this case it'd be easier to hire him as an employee or contractor and make it an internal instead of OSS project, but that's not the point.

Now instead imagine that there's a million people who each would get $1 of value if that developer kept developing. For a modest $50k salary, that means a ROI on 1900%. Sure you could not pay, but it'd be stupid. However, here's where it breaks down: Imagine one person doesn't want to pay. You now have 999,999 people to share the costs, which means it's still profitable (expected value > investment), but it is far more profitable to the one not paying at all. Repeat that 950,000 times and it's no longer profitable. And the last 50,000 will go "Why should we be paying for everyone else?" and not pay either.

Basicly, it's the mass version of the prisoner's dilemma. They could have gotten a very good value for their money, but because everyone is acting egoistically, the result is that they don't.

Re:Nonsense

[Tim C]

I'm not sure you get the OP's point. Take me, for example - I have a 9-5:30 job, an hour long commute either way, and a family. Between that and the need to eat, sleep and do chores, etc, I get maybe 2 hours per night to myself. That I have to split between everything I want to do - code, watch TV, socialise, etc. I imagine that most working family people are in the same position.

If I were coding on an OSS project, they'd get maybe a handful of hours per week out of me. Perhaps that's enough, perhaps it isn't. If I were being paid to work on it, and paid enough to do it full time, suddenly that goes from maybe 10 hours/week to 40 or 50.

It's not a question of interest, it's a question of time - there are only a certain number of hours available, and when you have a fulltime job and a family, you "lose" almost all of them to those commitments.

He's not saying that any given person should be paid because they deserve it, just that if people were to be paid, they could devote much more time to it.

But everyone is doing it

[Uteck]

Even Linus accepts money in return for his contributions to OSS and I don't see kernel devlopment slowing down because he gets paid for it other kernel hackers don't. Debian sees their problem and have come up with a good solution for it, perhaps this is a sign that some of the other things at Debian that move at glacial speeds will be reworked and made more dynamic.

Change can be good people, and it's not like this will be a perminate paying job. It's just for the next 2 months.

The difference between Work and Play

[Respect_my_Authority]

Linus Torvalds started to build a Unix-like kernel "just for fun" and his fun project soon attracted contibutions even though Linus never offered any bounty or payment. So what's the difference between Work and Play? The former often sucks all the fun out of doing things while the latter usually encourages people to contribute simply because it's fun.

Raising funds to employ one or two release managers for a short period of time just before the "etch" release may actually be a very good idea but I hope that the people behind this "Dunc-Tank" idea keep in their mind that fun and play will always be much more powerful motivators than money in a volunteer project like Debian. A crash course into understanding why this should be so can be found in the second chapter of "Adventures of Tom Sawyer" by Mark Twain:

http://www.gutenberg.org/files/74/74-h/p1.htm#c2 [gutenberg.org]

Re:Money isn't Everything...

[Rogerborg]

The code I write at home for my own enjoyment is far higher quality than what I write to satisfy the terms of my employment contract. End of anecdote.

You don't get it

[xeno-cat]

"For one, I can hire someone and claim 100% rights over their work."

And not gain any of the benefits of open source. The reason to use open source on a project is to gain the benefits of that approach. If your gaining benefits than it should not be such a stretch for you to pay to maintain those benefits as long as the cost/benefit ratio is in your favor.

You could hire an in house tech to work on some secret version of Debian for you alone or you could just pay the foundation to get things done quicker in the trunk. It should be readily apparent why the latter option would be preferable.

Re:Volunteering

[eraserewind]

That really destroyed my motivation. Why give away your time for free when others that are less motivated and less qualified are getting paid?

Same reason you volunteered in the first place?

Or free OS supported by unfree games!!

[Peter Trepan]

Have paid developers work on high-quality closed-source Linux-only games, or maybe only *partially* open like Quake, then use proceeds to help fund the OS on which they run. Games are non-essential, and therefore, I think, do not break the spirit of the GPL ideology when they're sold closed-source.

Re:Nonsense

[gmack]

This has already been proven to be completely untrue. If I hire my own programmer that programmer has to start from scratch (can't use OSS code) and duplicate thousands of hours of work already done or that programmer needs to spend a large amount of time synching with the lastest release of whatever software is being used as a base so it doesn't get left behind. It's actually much easier to hire a single developer to work officially on the project and then everyone benefits.

Many companies hire OSS deveopers to improve performance and add features critical to their buisness. A large number of Linux kernel developers are actually payed to work full time on the kernel. IBM is a notable example, so are adaptech, SGI, namesys, SuSE and Redhat. For companies that can't afford a full dev they often donate to OSDL instead.

Re:How is this any different

[QuantumG]

$100 is exactly the problem. I can earn $100 in an hour, what makes you think I can write a vpn frontend in an hour? People are not willing to pay market price for their bounties, so no-one bothers collecting them. Now, if you actually started writing a vpn frontend (even if that means just firing up the GIMP and putting together some mockups) and told people what you would doing, you would find you receive a lot of offers for help from people who also want that.