Funding GPL projects or funding the GPL?

In a recent discussion about developers running out of money (arch, perl) and what can we do for helping them I can up with an idea. I'd like to share it with anyone in the list even though it'll probably be disregarded or flamed. The Kernel may well be nicely funded, because of companies supported. But that's not always the case, and the schema just fails in a lot of key areas of OSS.

Why post it here then? Because for it to work it must be supported by at least some of the grand developements of OSS.

Here's the actual idea, which actually is a slashdot repost

Re:Isn't dual-licensing with the GPL perfect for t (Score:2) by fferreres on Friday July 26, @11:01AM (#3958396) (User #525414 Info | http://www.arrancar.com/)

Yes, but that underfunds the projects. You can see this clearly when Microsoft can sell lots of buggy software and of the best OSS developers can't earn a decent salary.

I'd love to see a new license, that could be called the fGPL. That would be the "Funded GPL". To be able to use fGPLd programs you'll HAVE to contribute some small amount of money to the fGPL foundation. You'll not be required to pay for any individual fGPL software, just a plain simple yearly $10 or $20 charge. And you will be able to distribute exactly where that money goes, among all the different projects. If you can't pay $20 a year it will be no problem, just a bit penalty: all fGPL software would be free as in beer once the year passes (old releases).

The money paid to the developers would only cover salaries and some expenses that are needing to continue developement. So if any proyect gets over-funded, you'll be noticed that you must reasign some of your credits.

It'd always be free as in freedom. We only need to bring some beer for that to happen. It'll also kill the anti OSS argument that the system is for comunists or anti-american. I know that is FUD, but do your representatives know that? It will also kill most of the other FUD targeted at OSS and will also bust developement to unknown levels.

What do we need for this to happen?

To have the Linux Kernel, the Red Hat distro, mplayer, X and gcc (for example, could be others as well) adopting the fGPL for the next releases. After that, we'll see most every GPLd program adopting the fGPL. After that, you'll start to see how much sense it made to pay $20 a year. And even the ones that can't pay (if any) will be able to use the software (though 1 year old, but their hardware si severla years old for sure).

This is my opinion. I'd gladly pay the $20, as long as EVERYONE ELSE pays their $20. That's why we don't see many donations now: because you have this felling everyone else is just waiting for a fool like you to contribute to project X in order to save it.

------------- end -------------

Thanks for everything and to everyone here!

Federico
fferreres (@) ojf com
pd: Please CC if you need my reply as I am not on the list

So, if I have a cheap/buggy motherboard and its faulty behavior is fixed with a new kernel release last week, I don't have those $20... I have to wait a year for a fix? I don't think so...

Regards

Robinson Maureira Castillo
Asesor DAI
INACAP

You can chose to assign your money _whenever_ you want (you have 1 year flexibility).

Project could open a "feature requests" if they like. Some of these could work in a "reward" fashion. A reward could look like this:
"whoever finishes a working driver for <insert card> gets <sum of funds offered by members>".

Federico

You didn't get it... I _do not_ have the money, but I _do_ need the fix, why do I have to wait a whole year?

Robinson Maureira Castillo
Asesor DAI
INACAP

The problem of not having enough resources to achieve a given goal has plagued mankind for years. Open-source, GPL or magic dust are not going to solve that problem.

} You didn't get it... I _do not_ have the money, but I _do_ need the fix, } why do I have to wait a whole year?

I Agree.

But I don't think that the kernel should adopt this new license, what stop some developers to fork another tree under the GPL?

IMHO, kernel should stay free...

Best regards

Robinson Maureira Castillo
Asesor DAI
INACAP

Nothing is preventing them, and that's a good thing. As at is now, everyone can fork a kernel. And everybody is inteligent enough to know which one to pick.

If they want to switch the license to GPL, they have to start with a 1 year old source or stick to the fGPL.

Now, It may be true that some projects should not be fGPLd or have exceptions, for one reason or another. It's up to the developers to decide what's best for them.

Free as in no money to the developers? It isn't really free. You need Internet access to download the kernel (or to buy a CD), a computer and electricity among other things. And the developers need to eat, to have bandwith donate, hardware donated, etc.

Federico

Because you do not have the money.

Idea 1: You do have $10 and you are trying to free-ride (honest answer).

Idea 2: You don't have $10 or $20 a year to "spare"? Then probably you are using (old and) already supported hardware. And because past years developements are free as in beer. If you ARE a developer, you get a free permament membership so you shouldn't care about all this.

Idea 3: Make bugfixes and hardware support are always GPL and not fGPL. The fGPL will force you to distribute under GPL (or fGPL, at your choise).

Idea 4: Write your own drivers if nobody else would.

Idea 5: You are a student and as such are granted a free membership until you finish your studies.

Federico

On 26 Jul 2002, Federico Ferreres wrote:

[sigh... I shouldn't be doing that, but...]

What you and the rest of armchair generals do not get is that "adding features" is _not_ the hard part of work. Doing that in a way that wouldn't be a permanent source of bugs afterwards and cleaning up the existing sources of bugs _IS_. So is doing infrastructure work. So is auditing code. So is removing crap code.

None of that is covered by your "model". 99:1 that your "working driver for card" is going to contain a bunch of root holes. _And_ be unmaintainable.

I had seen quite a few vendor drivers. Every time I'm looking at one of them, I'm reminded of MST3K. Yup, Mistery Science Theater 3000. With guy being forced to watch crap selected to drive him mad.

And it's not just vendor drivers. Example: recently a new version of cpio(1) had been released (after 6 years of inactivity). It still contains idiotic holes reported (with fixes) years ago. Many times. Some of these holes going back to 1993 (first report I'd been able to find, followed by ~5 rediscoveries of the same bug). Who should be paying to whom in cases like that?

The thing being, absolute majority of software is crap. That has nothing to getting paid for it and everything with average quality of programmers. For how many years did the (well-paid) rogering tosspots in SGI ship IRIX with sendmail choke-full of known root holes? (not to mention that it was configured as an open relay effectively hiding the IP of submitting host - spammers' dream come true). For how many years did Sun ship systems with mind-boggling default configuration? (NIS holes galore) For how many years does Microsoft ship what they are shipping?

As long as crap software is considered acceptable and people who write crap - employable, the things will be bad and job market - overcrowded. It's that simple.

[cut]

http://www.tuxedo.org/~esr/jargon/html/entry/Sturgeon's-Law.html

Nothing special about programming, this applies to television programs, acts of congress, the skill of most drivers on the highways, grocery shopping...

Fighting off Sturgeon's law is what maintainers DO. It's why 90% of all patches are dropped...

It's also the core business of Red Hat, Google, Slashdot, Slashdot's moderation system, KernelTrap and Kernel Traffic... (This isn't to say that the end result is less than 50% crud, just that they try very hard to change the percentages. With varying degrees of success...)

Rob

Al, your point is 100% valid. That's why is said "Projects could open "feature requests" if they like." I think the important thing is that developers are the ones that should run the show.

The idea here is to offer a monetary compensation for the hard work of developers directly from the users in such a way that the costs will be spread among millions of people.

How to pump money into projects and at the same time allow developers to do The Right Thing, is something that can be discussed and shaped according to each developer needs. The kernel will not want to accept code that came from a feature requests (maintainability or security or for whatever the reason). And that's good! Only good code should make it in.

An idea for funding the core developements would be to preassign a fixed 15% of all the funds at this developements (gcc, glibc, kernel, OpenSSH, etc.)

The rest will go to userland as the members see fit, as most (home) users think everything is about apps. We might even be able to fund games and (uneeded but wanted) things like that, boosting Linux adoption. We may be even able to buy some patents (like the SGI ones that MS bought, before somebody elses limits us). Or we may be able to open the code to certain apps to turn them into fGPL, if need be.

So in brief, the power will remain at the developers, because they are the ones that know what they are doing. If some folks want an insecure piece of code to make it into (for example) the kernel, they'll be so out of luck as they are today.

The idea may be crap in many people minds. It's not perfect, it's not without dangers. Maybe OSS doesn't have a need for those funds. But they money is there. It just needs a way to flow from users to developers. Right now it's flowing from the users to Microsoft. That's fine, but we could do better, faster and bring freedom to users.

All we'll need is discussion of pro/cons and whoever likes the pros more can fGPL. After 4 or 5 core apps fGPL, everyone will want to fGPL and get some money to pay the expenses and hard work they are doing.

The choice is all yours ... you own this market, you control it and you deserve to manage things as you see fit. My $20 will be waiting for a way to contribute in an inteligent manner which also forces everyone else to do the same.

Federico

This isn't even possible for the kernel nor is it possible for any GPL app. If you want this you will need to start from scratch since "fGPL" would not be compatable with GPL you would need to get every last person who submitted a patch to relicence.

I just don't see the point of trying to argue for what is essentially an imposibillity and that's even if everyone here agrees to it.

Gerhard

PS There is something to be had here that is better than money.. it's a chance to read and learn from the best. I'm not kernel level yet but I do contribute to other projects. I learn a lot just by reading and one day I hope to have code good enough to be slammed by Alan Cox, Al Viro or even Linus himself ;)

--
Gerhard Mack

[email protected]

<>< As a computer I find your faith in technology amusing.

That's a real problem (and that's probably why Sun tries to be the copyright holder in OO)

Maybe the kernel would be wrong place to start, besides the kernel is already well funded. (It could provoke a demostration or cascading effect).

Yes, that's clear. It's great. You learn, then you contribute back. But that doesn't allow you to pay your bills, does it? The thing is massive amounts of people _don't_ code, never will, are willing to contribute but don't do it, because they are not beign asked for support in the right way. You need them to know they'll put $10 and get $100.000.000 worth every year.

But if OSS finds a better way to fund developement, it certainly won't hurt your learning experience. It may well improve it (more code, more research, more accesibility). The point is to allow more people full dedication to OSS.

Federico
PD: Maybe some day we'll find out the perfect solution.

Thank you for taking Economics 101. Your final score is F-. Don't let it discourage you from taking the test next year. Pay particular attention to the reasons why pyramid schemes are not sustainable. We wish you a nice summer and hope to see you next semester.

Al,

Do you know the conditions for a pyramid scheme? A piramid scheme starts when a guy can convince some (more than 1) other guys to pay him a certain amount, with the promise that they will be able to do just the same sale to other people. Of course, at some point the system stops working as everyone knows (easy math, no 101 needed).

The fGPL scheme means you pay $10, and you get code in exchange. The developer receiving the funds either buys food, a computer, a book, or maybe even uses it for a nice trip with his/her family after their hard work is done. Let me know when you find the pyramid!

You can be agressive all that you want, but it doesn't look good when you are making a fool of youself. Millions of people like me are free riding your hard work. Dump my idea, laugh at it, hate me for being so stupid...we couldn't care less :)

Federico
PD: I do know you are one of the most talented people around, and I couldn't thank you enough for the hard work you do everyday. But that doesn't mean you are always right. Sorry.

Sigh... That stops working at the same point where pyramid does - when number of recepients becomes a sufficiently large fraction of all potential participants. It is not sustainable.

Simple math: Debian got several thousand developers. Most of the packages contain both contributions from said developers and stuff from upstream. By very conservative estimates it's tens of thousands. If your $10 is per package update - it's impossible to pay. _Really_ impossible - on a reasonable system it will easily amount to $50 _daily_. If it's $10 per year - count the number of installations and look how much it will give for one developer.

It won't work for the same reason why pyramids are unsustainable - when you have too many recepients, both go to hell. And there _is_ too many for that to work.

It's $10 or $20 per user per year to be able to use fGPL code _alone_. In addition to that, you'll have to pay for any extras (services or non fGPL code) you may like purchasing: the nice Red Hat Network services, your Debian CD and my Slackware CD (but I would preffer to just fund Patrick though fGPL instead). The membership just allows you into "the club".

Also, publishing something under the fGPL doesn't mean the developer will get the funds, unless people choose to fund it or if it's a core app (this last decided by a board of directors for example, lead by Linus or whoever is most respected and wants to get involved and up to a fixed % of the yearly fund raising).

This doesn't kill any other funding source you already have established. IBM may not want to fGPL webalizer, Covalent or MySQL AB as well. And that's fine. Copyright owners can offer them other licenses suitable for them (X style, BSD, whatever). Also if IBM (or any other company) wants to fund a different project because it makes sense for them, and users and small companies do not want to pay for it, then they will fund it themselves, as it is the case today.

Now some figures: if you have 50.000.000 (end users and companies) there paying $20 (as an example), that's about $1B a year average as Larry suggested. If you have 5.000.000 it's only $100M. You at least have some extra funds available that go DIRECTLY to developers. And as all projects progress faster, you will have widespread adoption of OSS: I will just be the natural step money-wise.

If 1B a year is not enough, then some projects will have to wait, or the users base must grow (and it will grow surelly).

The Linux Kernel and the core of OSS will have a fixed percentaje pumped to them, so the ones that know what must be done are not bothered. Say 10% (or a different %). That's between $100M and $10M extra a every year, and you can always count on it. If you don't need that much, then you could resign the right. And if it's not enough, then it's much better than nothing / "charity".

It works for Microsoft and they make crappy software which is also propietary, closed, and expensive. Why wouldn't it work for OSS? You'll be making everyone else in the world a favor. They collect most of their revenues due to the Windows plataform. The kernel may provide that base for all the GNU deck.

can't pay $20 year, doesn't know how to code, is not a student or a charity organization may well thank everybody else's for they will be able to use last year developement for free as in freedom and in beer. If anyone is concerned about the impact $20 will have, as long as you allow piracy (a la Microsoft), there's no problem. It may even help OSS.

Now the real concerns would be:

- Will the money be used wisely?
- Will money change the spirit of OSS and bring bad code with it?

I couldn't answer those. If the answer is negative, then OSS is better off underfunded (as it is today). All for the love.

But what if it works out nice? You are indeed underestimating the willing of non-coders to contribute. They will probably even be proud members of the fGPL foundation.

Federico

You want us to make crappy software with silly restrictions, too ?

Rik

Bravely reimplemented by the knights who say "NIH".

http://www.surriel.com/		http://distro.conectiva.com/

Cute. How about we ease off on this guy a bit and ask ourselves if he doesn't have a point? It's not clear that he has the right answer but it is blindingly clear that pure open source business models don't work. Every single one of them which has any success has found a way to derive money from some other source to fund the open source development.

You could argue that everything is fine and you'd be wrong. Not all applications can find a way to piggyback on some other sale. Consider a Word replacement, how do you generate money for that? You can't tie it to hardware sales, you can try the support route but that isn't going to work, so what do you do? And there are a million other apps which are the same.

It's easy to beat up on this guy but he's trying to find a reasonable compromise which allows open source to be self sustaining for all projects. That's a great goal. Right now, open source funding is a joke. Microsoft wipes their butt with more money than all of the open source revenue put together. Yeah, it's very macho to say that you'll program circles around them, but it's also wrong. You may in one or two areas but they have a zillion apps that lawyers and doctors and other non-geeks use every day. Where is the money for those apps? It's sure as hell not coming from "donations".

So how about we all cut the guy some slack, maybe he's misdirected, maybe he has the wrong answer, but he's not working on the wrong problem in my opinion. Try helping him, it might be in our best interest to do so.

---
Larry McVoy            	 lm at bitmover.com           http://www.bitmover.com/lm 

Agreed. What we need is funding for the development of those things which aren't already taken care of by the community.

In the areas where open source software is doing well, there already is funding by interested parties (companies, government, universities, ...).

Some other areas won't ever get the funding through donations, simply because people will freeload whenever they can and try funding development as much as they can. We've seen that with BitKeeper and you had to "tighten up" the license a bit in order to make sure development stayed funded.

In short, I believe the voluntary donations aren't needed in most areas people would donate to and won't make enough of an impact in the areas where they are needed.

regards,

Rik

Bravely reimplemented by the knights who say "NIH".

http://www.surriel.com/		http://distro.conectiva.com/

Yeah, and I wasn't thrilled about it, I can tell you that. I had the same starry eyed idealistic idea that things would just work out and they don't. So we're essentially in a proprietary model with a funky dual free/pay licensing model. That's OK for us because the money people value their privacy.

Actually, that's an interesting topic. Other applications could use the BK model of "free if you're out in the open" and pay otherwise. It's pretty effective. However, it doesn't work very well when the community beats you to hell for not being GPLed. I had a thick enough skin to deal with it, I doubt others would, they'd give up. It also doesn't work when people refuse to obey the license because they don't agree with it (we had plenty of that).

That leads to the real question: what is an acceptable model for the community and the vendor for applications which don't work under the standard GPL model?

Agreed, but his thought was to make it non-voluntary. I have my doubts about that working too, but there ought to be something which would work.

---
Larry McVoy            	 lm at bitmover.com           http://www.bitmover.com/lm 

Trolltech does the same. And have taken the same amount of heat. (Probably A LOT more... since their Qt is the base for KDE)

/RogerL

Roger Larsson
Skellefteå
Sweden

The problem with the present type of donation systems is knowing your donation is not going to be abused.

I though a model that may work is a United Way type of model ... have a central non-profit act as a clearing house for donations to specific open source projects. The non-profit would do the proper account reporting of receipts, overhead, and funding to the various open source projects (an effort to keep donation abuse down and allow open non-bias account reporting). Allowing either a general donation or target donation from the public to a particular project. Then a specific project would register with the non-profit to receive funding help. In order to qualify you would need to provide certain technical/financial reporting of your project to the non-profit to demonstrate you need the financial help and you are working on a needed project.

The real problem is getting something like this started.

Hi to u all!

It's funny how some things seem to pop in to peoples minds almost simultainously. I believe some people call this phenomenon "A collective consciousness" I've had some open source funding Ideas spinning in my head since the beginning of May. My Ideas differ from Fredricos in some crucial ways. But I believe that they might be connectable. in a way that will form a nice symbiosis.

I'm willing to put considerable amount of time in trying to resolve, design and develop both of them together with others, at least in part into a reasonable functionality.

This will take coordinated discussions, transcripting, more coordinated discussions A lot of book reading ( law practise ), advocacy, contacting people at nearby Universities, forum creation, advertizing on the internet, Writing forms, making inquiries, contacting authorities, etc, etc, etc

It's not something that can be done just by one or two dudes.

It will probably not be achievable unless at least 50 persons get deeply committed around the world for starters. and perhaps 200 guys for additional support.

First I want to know If more people are interested in doing the same. I have an Email adress that is not yet in use. , so if you send a mail to

[email protected]

with "some cool subject line"
and a body saying something like:
======================

I'm In for now. I have some spare time
to do some tedious and boring stuff.
Let's see where this "fantasy" is heading.

======================

I will send the result as a reply to everyone that showed interest in about three days from the time this letter got posted. It will NOT go on LKML.

Oh yes, almost forgot about the math part:

$1000 000 000 / 100 000 hackers = $10 000 per hacker.

I would be quite happy if I got $10 000 a year just for getting a little patch that I wrote in a day or two into some GPL:ed software. Especially if I knew that someone sent me this money just because this dude appreciates the software that I work with specifically.

( even just a $1000 which would finance two new boxes / year )

The basis of my own idea is built in part on creating an

Open Patent Funding service organization.

Finance is partly solved by administration of several small charity foundations with identical purpose, created by the same amount of SW-companies in order to avoid tax, as well as being able to use new SW-methods freely. I've googled for this without any success.

kind regards
/Daniel Mose

... and leverage the synergy. Or something.

[more spew snipped]

Larry, do you really like the company? Slashdorks and pretentious self-important wankers speaking PHBese... YMMV, but IMO "prose" above is every bit as disgusting as usual GPL advocacy. Same depth of thought and amount of clue behind it, same lack of taste...

I'm not a big fan of fanatics and I see your point. My perspective is probably skewed from the whole BK experience. These guys may be a waste of time but I don't see why they should be shut down. I happen to agree that there is a problem which needs solving. If they want to solve it, or better yet, if they simply act as a catalyst and someone else solves it, that's cool by me.

---
Larry McVoy            	 lm at bitmover.com           http://www.bitmover.com/lm 

I stated a simple idea aimed at solving a real world issue. And you haven't proved it wrong. It may not be what you or the kernel hackers need/want (which is FINE). But it would solve ALL the funding problems at least.

The name calling is ok, but only if it makes you feel smarter or superior.

Regards,

Federico

You don't get it. So far the only guy who had been charitable was Larry, who felt that problem was real but had serious doubts about viability of your idea. I don't feel charitable and I've no reason to hesitate telling that you guys _are_ waste of time. No maybes about it. It's that simple...

Actually, this is what Larry said:

| One problem I see is that you'd be talking a huge amount of money,
| potentially money on Microsoft scale.  Managing that money, making
| it go to the right places, without it sticking to the fingers of
| management, isn't likely to happen.

This is not a problem of fundraising, but a problem of fund management and policy. You get the funds, how you use them is another concern.

| Another problem is that GPLed software is essentially software in the | public domain.

But we don't want less people using (for example), the Linux kernel. We don't even want people who can't pay to pay. As long as a large enough group pays, you're really better off.

And here is what Rik said:

| In the areas where open source software is doing well, there
| already is funding by interested parties (companies, government,
| universities, ...).

Meaning the kernel and some other OSS areas are already well funded.

| Some other areas won't ever get the funding through donations,
| simply because people will freeload whenever they can and try
| funding development as much as they can.  We've seen that with
| BitKeeper and you had to "tighten up" the license a bit in order
| to make sure development stayed funded.

Here he seems to be suggesting that certain areas of OSS that can't get enough funds should either relicense, drop the towel or do it for the love as a part time job. I am suggesting they could all relicense to a single license that leverages each others work and centrally manages fundraising (but not fundspendings).

| In short, I believe the voluntary donations aren't needed in
| most areas people would donate to and won't make enough of an
| impact in the areas where they are needed.

Now he is poining out something that I had stated in my first message:

On Fri, Jul 26, 2002 at 01:09:32PM -0300, Federico Ferreres wrote:
+ The Kernel may well be nicely funded ...
+ Why post it here then? Because for it to work it must be supported
+ by at least some of the grand developements of OSS

... and has nothing to do with the ability to raise the funds. The Kernel may not need the money but that it could help in a lot of areas of OSS where a kickstart is needed (after which I am sure they will be able to contribute funds back to the pool).

Regarding your opinion on the pyramid nature of the system well that can be forgiven :)

The real facts that I am aware are:

- the kernel doesn't need the money
- the idea of _enforcing_ payment may not be an option or what is wanted
- there are doubt's about how many people would end up paying
- it will not be easy to manage the funds and could have side effects
- it may be nearly imposible to relicense the kernel due to the large number of people involved even if they wanted

But the last fact invalidates all others:

"The idea is discarded, nobody likes it"

Which is fine! And I certainly am at the wrong place, poluting a developement list and wasting your time. So I'll stop replying to your flames (even though I find them funny and creative :-)

Regards!

Federico

Viro is abusive to everyone (by email, he is likable in person oddly enough), usually without understanding what he is talking about at a level of depth any deeper than it is new therefor wrong (see devfs thread where he rejects devfs on the basis of endless details without understanding that the basic idea had any merit).

Your idea has some merit in my opinion. I think that my version of it that I presented at linuxworld some years ago, which is called an "Open Sale", has some advantages.

"Open Sales" are the best method for reducing the economic
distortion caused by non-zero marginal pricing of
products with zero marginal cost without losing the incentive to
produce. An open sale is when a user group agrees to
pay X% of their hardware expenditures to all of those who provide
them with the right to use their software, with the
users performing the allocation to the providers based on the
users' usage and perception of quality. The users
determine the value of X, setting it in accordance with their
interest in attracting providers. It is open in the
sense that any software producer may add their software to the list
whose usage and quality will be
evaluated. Automation of usage monitoring and sampling techniques
are expected refinements of the basic idea. It is
based on a belief that economic distortion is proportional to the
ratio between marginal price and marginal cost, not
just the dollar total of the gap, and that increasing the price of
hardware by X% is less of a distortion than
increasing the cost of software by an infinite ratio over its
production cost. An open sale retains the consumer/user
driven decentralized economic allocation of traditional sales, and
in the case of government agencies and other large
purchasers, it enhances it. Payment of the X% is required to
legally use the software.

My approach has the advantage that the fee scales with hardware costs, and that it is set by users.

However, you should understand that an idea is not enough, you must have sufficient sociological mass to pull it off. Neither you nor I are in that position at this time. Most people will be hostile to you if you propose an idea that you lack the sociological position to effectuate.....

Hans

Your description of the problem is perfect. The solution, though, is difficult to enforce as there's not a single incentive for OEMs to fund anyone else that MS (reinstalling Windows).

That's really an improvement over my idea (in the corporate area mostly. In the home user area it doesn't matter, as he will probably use 1 computer at a time).

I can clearly see that a combination of both schemes would be needed, because they are opposite sides of the same coin (your idea could be though as an x% tax on hardware to fund developement, mine as a direct payment from the users).

So one natural way of getting the best of both worlds would be:

- Hardware EOM could pay an x% per machine to the fGPL foundation, and that would grant the buyer of that hardware a permanent license to all fGPLd software. Users could request the OEM (IBM, etc) to offer them that non-expiring (per machine) license instead of a preinstaled MS Windows. This would mean OSS remains totaly free as in freedom and _beer_.

- Hardware manufacturers that do NOT want to offer their customers that option will be penalized because the user will have to pay the regular 1 year fGPL I have already described or get their hardware somewhere else. This would create a great incentive for OEM to offer the fGPL license instead of a preinstaled MS Windows.

- If desired, all past hardware sales could be granted a free non-expiring "per-machine" fGPL license (so you'll see funds starting low and growing every year in size) so that there can be NO complains from anyone. Maybe the grant (for old hardware) could be limited to individuals, non-profit organizations, goverments and educational institutions.

- Manufacturers will be granted the right to claim to be GNU, Linux or OSS ready ONLY if they actively offer the fGPL license as choice to customers (instead of the preinstaled Windows). They could also be required to include a "GNU/Linux ready" sticker in their hardware.

I believe this can be made to work in reasonable time, with minimun effort and minimun hassle. There are problems, but they could be worked out if there's support.

Everything else remains the same (as Hans and I have said, the users will have a limited ability to chose what they need. I say limited because for an Office application to work the core must remain funded even though the user may not notice it).

Makes sense. But if the idea is good, the ones that can make a difference will support it, assuming they analized what is being proposed.

I will (silently) put some resources on polishing the hard edges of idea. It believe it may be worth it.

Federico

PS: I probably never reach the needed sociological position in OSS, so at some point someone will have to help (if they believe it's a good thing).

Mine is a direct fee payment from users with allocation of fee instructions accompanying the fee. The fee is less for buyers of cheap computers since the hardware cost is low, so folks in Brazil on average pay less than folks in Silicon Valley. The more you spend in hardware the more you can afford to pay/contribute for software.

Limited? What limit? If you pay the X% of hardware cost fee you can use all the software in the pool.

You are assuming they think. They don't. They sense herd movement. This is wise of them, because unless a large portion of the herd moves to it, it has no value. Thus it has no value. Thus the herd does not move.

Hans

Does it matter who collects the money? It should only be based on convenience. And it's much easier/safer/efficient to have the OEM collect the funds than any other party.

Imagine going to a store and asking "Two Sony Vaio laptops please, and make them OSS, NOT Windows" and beign charged the computer's quoted price (or maybe even less, but never more). Imagine not having to research on how to get a refund for that Windows license that came bundled and that you don't need. Imagine not having to _require_ the OEM to preinstall any OSS on the machine for it to "OSS ready" (unless you ask for redhat/suse/etc), yet beign able to use any OSS without having to pay any extra dime.

There's this ilusion in normal people's minds that Windows is free because it comes with the computer. We should take advantage of that fact. And anyone that claims OSS is free is having an ilusion for OSS costs money to develop and to be able to use OSS you need to spend a lot of money (even if you don't pay anything back to the OSS developers).

Small note: funding OSS is not about the money. The money here is a medium (funds) for an end (free software). Only businesses invert the relation and put money as the objective objective and turn work into a medium.

Here's a clear version of what I meant: Limited ability to "allocate the money paid" based on what they'd like to be improved.

moves

move.

Whatever they decide is fine for me as it's their show and not mine. You and Larry can think differently, as you are involved in developement.

Federico

Hi,

Not to mention all the people who think that Windows is the computer, indivisibly, and never heard of the term "Operating System". Oh, well...

Thunder

.-../../-./..-/-..- .-./..-/.-.././.../.-.-.-

There are a couple of interesting points raised in this thread.
1) That a substantial part of the work of software is not building it, but is maintaining it.
2) That several people feel that it is hard to make a business plan with open source software.
3) Many successful software companies, have made money with a hardware tax. And open source could feasibly get part of that action, all you have to do is to demand that Linux be pre-installed. And the distributions should get involved to help the hardware manufacturers.
4) Raising money is conceptually very simple. But practically hard.
5) How society perceives the solution is a very important part of any solution.

A lot of economics is modeling and finding a system where the greedy algorithm, can be used to optimize the system. As all decisions are local with the greedy algorithm it scales very well. Most forms of central planning, distribution whatever fall down because they are not built on algorithms that scale.

The basic economic model of supply and demand with competition works fairly well. But it has problems when you have either an external cost (Pollution) or an external benefit (Free Software). With some small amount of government regulation it is theoretically possible to introduce these external costs, into the direct costs normally dealt with. The best scheme I have seen proposed is to sell permits to pollute, with the maximum number of permits limited by the maximum amount of pollution you wish to allow. It might be possible adapt this to free-software by allowing having permits to sell non-free software. But that requires a large shift in how this are accomplished.

But the assertion made by Al Viro that software maintenance is where the bulk of the work is, is interesting. Long term this is trivially true because the all of the code has been written, and there is no new development to do. Services like distributions and device driver writers, and kernel maintainers appear to be in the area where maintenance is important. Maintenance can be handled by maintenance/support contracts, making the economic model with closed source and open source the same, except with open source it is easier for multiple maintainers to cooperate. And in the areas where the work is primarily maintenance is where open source has been observed to be well funded so this appears to work in practice.

Given that software maintenance is the primary problem, it is only the creators of innovative open source programs whose costs are external to the economic model, that making business plans harder to deal with. So the question becomes how in the open source community do we encourage true innovation, while not encouraging it so much we fail to weed out the dumb ideas. Innovation always has a large share of external benefit so the problem of how to encourage and compensate innovators is not new, but the open source landscape is.

The only idea that spring to my mind to encourage true innovation are obligating the distributors to pay the innovators something for the programs they distribute. Or having universities and other research institutions pay the salaries of the researchers. Off the top of my head I cannot think of anything that takes advantage of the potential to bypass an old guard of maintainers, and universities and go directly to the consumer.

Eric

Another large part of the work is in pursuing innovations in software that fail. Note that failed software is generally not maintained.

What I don't understand about Viro's argument is why he thinks that maintainers won't also get paid under the schemes discussed. There is nothing that prevents paying for maintenance in either scheme.

Hans

OK. How about this?

Let us observe that the fundemtal 'problem' here is that closed source software companies get funding by levraging it's control on the inelectual property it generates to get money - 'pay us or live with out it'.

The challange is therefore to find a way, a 'business plan' if you will, to get money *without* execersizing control on generated intellectual property. To do this, let's note that there are already equivelent fields of business with some very close charateristics to software (that is they involve brain work, not digging coal) where for various reasons, control of intelectual property is not how you get paid.

Let's take a specific example: lawyers cannot control the intelectual property they generate because it's part of the law system that once a lawyer thought up a new idea anyone can use it. In fact, AFAIK precedence is one of the basic ideas of the modern law systems.

Are lawyers out of a job? Are there less and less lawyers around because they can't make money? Are they perhaps not innovative? I don't think so. As a matter of fact some would even argue that they maybe too much lawyers around but let's not go there...

Ok. So how do they do this? simple, the way they organise their business is suitable for the job: AFAIK, a typical law firm has a couple of senior lawyers who are the 'partners'. The rest of the staff, including the less senior lawyers, are hired help.

A couple of interesting to note:

1. The top brass are lawyers, a lot of the time practicing lawyers, not 'managment'.
2. No sales, marketing etc. etc..

Where these are not true is usually in the really big (and therefore we assume succfull) companies. That is it begins to maybe become slightly different only AFTER you succeed.

To paraphrase Robert Heinlin, in a typical law firm senior managment 'everyone drops'.

And that's it. (I think :-) the problem with current Open Source operations which don't do so well is that they are trying to build a company structured around how closed source software companies are built
- senior manamenbt that don't program. CEOs from that other (closed source!) company that worked. Peopole who do PR. Sales people. People who do not 'drop'. Can you say 'Overhead'?

In short and plain words, what I think the lawyers know and we need to learn is that to make a business built on something else then intelectual property control, you need low overhead. You need a firm where (almost) everyone drops. Building bloated companies only works if you grab someone in the balls (the customer) and make him maintain the overhead. When you can't grab someone in the balls you've got to be mean and lean. When you're big enough then you can efford some of the bloat and by then maybe you even need it. Maybe.

Imagine for a second a software company built according to 'law firm principles' - we have the 'Senior Hackers'. They all write code in addition to taking care of the other sides of business. Sure, they might hire an accountant and other employees to take some of the work etc. but there is no bloated 'managment' that can't hack. Intresting concept if nothing else...

I'm sure by know you are all smiling and thinking "OK, I want to have some of what he's been smoking...". Fine, but the lawyers are doing something like this for a long time. It works somehow. Maybe we can learn.

Just my 2EUROs,
Gilad.

Gilad Ben-Yossef <[email protected]>

"You got an EMP device in the server room? That is so cool."
      -- from a hackers-il thread on paranoia

Gilad Ben-Yossef <[email protected]>:

Ummm. nope. Wrong (or at least very incomplete).

Lawyers CAN own the new idea - they can copyright their presentations preventing others from using them without first paying for the right to use all/part of the presentation in other cases.

Lawyers are closer to high profile consultants.

The same can be said for consultants :-)

Actually, they are the management. These practicing lawyers only handle the most exclusive clients (and for the most $$$$ possible). All others contribute a cut to the firm (and charge less $$$$ to the customer).

Thats the same way a consulting company works.

...

And who is charged for the time?
Who gets the direct benifit of the work?

You must have a customer/client that is identified for all billing. Personally, I think Red Hat is doing the best job of this:

1. provide consulting to clients paying for help.
2. providing custom software to clients paying for it.
3. providing short debug/fix/patches cycles to clients/customers paying for it.
4. advertising their capabilities to attract more (paying) customers/clients.
5. funding developement out of their charges to customers.
6. funding debug/fix/patch cycles for applications that do not have paying clients/customers. Probably charging this to "overhead" or "marketing/advertisment" to expand their list of capabilities.
7. Being good enought to pass those developements/fix/patches to the rest of the world via GPL/LGPL licenses.

IBM is doing a similar thing, though I suspect they count the expense of the open source development/debug people into the charge passed to the customer purchases of their hardware (ie. bundled into hardware cost) or overhead/marketing overhead, or in consulting fees.

Not being an employee of either RH or IBM, the above statements are, of, course, my opinion of how they are looking at things.

The other side of coin is "what do the customers get"
1. someone to call when they have problems
2. customized software aimed at their environment
3. a lower expense than the M$ tax
4. a better (ie. more reliable/secure) product than M$ can produce

What does the world get out of it:

1. an alternative to the M$ tax
2. a better (ie. more reliable/secure) product than M$ can produce
3. the ability to modify the software, learn how things work, limitations of the software
4. a better personal knowlege environment allowing the potential to be better employed by other busineses.
5. an ever expanding knowlege of how/why/what software works

-------------------------------------------------------------------------
Jesse I Pollard, II
Email: [email protected]

Any opinions expressed are solely my own.

That's _not_ the assertion I've made. It's not about innovation vs. maintenance and frankly, I'm insulted by that interpretation (which is quite an achievement - I've got really thick skin).

Actual split is not into innovation and maintenance - it's into new visible features and everything else. Which happens to include
* writing maintainable code.
* designing sane interfaces.
* writing code in a way that allows audits.
* _doing_ audits of your code.
* getting already existing code into form that allows aforementioned
new features - and if they are really new it's 99% of the work; the only way to avoid it is by creating, er, private copies of existing subsystems and kludging them up into something you can use. Effects of such strategy are left as an exercise to readers. Or, if your new features happen to be similar enough to something not new you get to reuse the work already done back when those features were implemented.
* etc., etc., including the things traditionally considered as
maintenance.

All mechanisms I've seen mentioned in that thread encourage only one thing - new features as fast as possible at whatever cost for future work. Calling that innovation and relegating everything else to support contracts... I'd beg to differ.

THAT is the problem - mechanism that gives incentive to being quick, dirty and sloppy (and encourages featuritis, while we are at it - inventing features that are not new by any stretch of imagination and selling them as great innovations is the oldest trick in the book). Folks, it's not a good thing. There's already more than enough pressure in that direction and it had already given us dungpiles of unsupportable, exploit-ridden software. Commercial and free alike. And guess what? After a while you _can't_ do anything really new on a platform, since it's already choke-full of kludges from "quick innovators" - to the point when touching it in any non-trivial way breaks something and nobody actually understands the thing enough to fix the breakage, so more and more kludges are layered. Sheesh...

Viro, do I understand you right that you feel that what is visible to users will get more funding from users, and that because there will be far more suppliers the users will be more inclined to not pay for the unseen benefits than when buying cars?

My answer to this is that users should be given a sample of, say, five providers, each of whom gets to make their case for what they have done for the user.

We seek not perfection, but improvement.

Hans

PS

Some of you may be inclined to suggest that such a sampling scheme would tend to benefit the small contributors..... and to this I say that more sophisticated mathematical models which reduce the variance between the importance of those in each sample so as to increase user precision of assessment can come later after the basic idea is started up.

Does the average user know about code clarity and maintainability? I think I see his point. I've seen what happens when non coders make decisions as to what features need to go in. Don't think it happens in the open source world? Go look at what most irc nets use for code.

It's why I actively avoid programming for a living.

Gerhard

--
Gerhard Mack

[email protected]

<>< As a computer I find your faith in technology amusing.

That's part of the problem, indeed, but nowhere near the entire picture.

Or, easier yet, go to prep.ai.mit.edu and try to RTFS. Use of safety barfbags is mandatory.

There is nothing in the funding mechanism that causes features to go in, merely to cause features to get paid for.

Do you understand that capitalism has this problem in general, that when consumers/users decide who gets money they make a lot of mistakes. The only thing we know is that when the experts of production decide things instead the results are usually worse (communist systems have things determined by producers not consumers, though it must be said that there are some things that communists designed better than the west).

Not seeking perfection, just improvement....

Hans

Hans

A little research will show that the governments have little to do with design. Governments just tend to determine what will be designed. Most designers, Engineers, Architects, etc., design in spite of governments influence, not because of it. Both governments and consumers generate markets for products. When consumers make mistakes, people lose their jobs. When governments make mistakes people lose their lives. Government is both necessary and dangerous, like oxygen.

Cheers,
Dick Johnson
Penguin : Linux version 2.4.18 on an i686 machine (797.90 BogoMips). The US military has given us many words, FUBAR, SNAFU, now ENRON. Yes, top management were graduates of West Point and Annapolis.

He doesn't need to prove it wrong, new ideas need to be proven right.

regards,

Rik

Bravely reimplemented by the knights who say "NIH".

http://www.surriel.com/		http://distro.conectiva.com/

If you made your source code available so that it could be customized then I would stop beating you for your license. As it is, any new features have to get written by you, and I have requested a trivial to author feature from you that I need and you haven't written despite your promising to do so before I made the switch to bitkeeper. I don't at all blame you for being busy, but if we had the source it would be 5 minutes for us to add the quite valuable to us feature.

Make your license charge for privacy but have the software come with source code, and the license is brilliant. Otherwise, it is a source of frustration, and yet another example of how software copyright is being abused and twisted away from its original intent so as to permit information to be hidden rather than published. In the long term, it is the accumulation and compounding of knowledge that matters, not limited time monopolization for 14 years (original length of copyrights), and secret source code frustrates that.

Hans

I'd just like to add that I respect Larry's problems regarding:

* small number of seats means large number of dollars per seat is necessary to justify the business investment

* people modifying the source to turn off open logging and thus defeating his original business model (this would really piss me off also)

He is also an all around nice guy who has given us some nice tools, and tries hard to be generous while keeping his business alive.

I would like to see him try making his license such that if open logging is not used, and he is not paid a license fee, then the user has agreed that the software is GPL'd and any potential user may sue to claim that software. I think this would scare managers into paying their fees, and I think it would work, but I also understand and respect that it is not my business that would go under if I was wrong, and it is not I who would have to explain to his workers why payroll couldn't be met.

Hans

The problem is not the community, but how the comunity is structured in certain areas. If big companies, educational institutions or goverments are not interested in your project, the OSS model fails as you stated.

100% true.

Good point. I started at the wrong place. Anyway, if your funding is based on final users payments you need to make payment a non-option. So if vital parts of GNU don't help them a bit, it may just not work. If Windows/DOS didn't help fund Office, would we be having the Office compatibility problems now? Extend the list to all MS userland apps.

As long as payment is optional and fragmented, OSS will have a hard time in userland core apps. Maybe an fGPL could work for them. Bad luck the projects that could help are already well funded.

I'll get back and rethink the problem ...

Federico
PS: Larry, it's ok really. Harsh comments that make sense are much better than silence. Though without your posts it'd have been mostly flames. Thanks!

I forgot to answer that one. It doesn't matter how many people want to join a project. It's up to the copyright owners to decide how much stuff they need, who to "work with", "how much folks are needed" and how is the line of comand structured. At some point, contributors will have to send free patches as they are doing now. That may or may not earn them into the "core team".

Users will decide how much money to put in such projects. If not enough funds are pointed in your direction (based on how much the developer requested), you'll just have to do with less (which is better than nothing). And the asked funds will probably make sense, because if they don't there will be an incentive to branchthe project (except for the core apps as already pointed out).

In the Debian scheme, people can't choose where the money goes, and there's no strict limit to the number of developers, and it's not a big enough user base (though big). As you point out, the ships sinks, as it's a classical example of the Fishermen dylema: the more you capture, the better of you are off, but you can't prevents others from doing the same. After some time there's so much fishermen that nobody capture enough to make a living from it.

And that's why fGPL should allow people to choose their funded projects (except the core systems) and why lead developers should be able to manage the funds as they see fit.

Federico

I have already donated $30 last year (but to the EFF as I think freedom is also important). But how many individuals donate though? I'd say 99% of the people just free ride. That means resources are very scarce and not enough OSS is produced.

But something is wrong with this. If I donate then it's no longer free. And only if you never donate then you can call it free. Efectively, that means that to get to a certain level of funding donators have pay 99 times the amount that would otherwise be needed if everyone contributed.

This reminds me of my countrie's idea of "social justice": the rich go to the tax funded, free universities while the poor grow corn and work 18 hours, 7 days a week and also pay the taxes (40%). Of course these poor can't pay for housing and food while studing. But every time someone suggests charging a small fee to univ. students (offering state loans for the ones that can't really pay that would cover food, housing and the fee) to partially fund those universities, and in order to lower the consumer tax (poors tax), they are dismised because "free access to the university is there so that the poor can study".

OSS is not much different either, but it's very difficult for most people to understand it for some unknown reason.

Federico

Hi!

There's no such thing as a free ride. Programs are buggy, all of them, and each "free rider" helps testing. Programs are not feature-complete, and each "free rider" is pretty likely to add feature they need. This means no more money once all bugs are killed and once all features are done, but thats okay.
Pavel

Casualities in World Trade Center: ~3k dead inside the building,
cryptography in U.S.A. and free speech in Czech Republic.

The GPL explains your misunderstanding:

When we speak of free software, we are referring to freedom, not price. Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for this service if you wish), that you receive source code or can get it if you want it, that you can change the software or use pieces of it in new free programs; and that you know you can do these things.

Free software is about freedom, not about pricing.

DS

That is fantastic, unless what you need is not beign developed and supported. In these cases, the GPL effectively prevents the software from beign developed and funded.

Suppose Adobe wants to GPL Photoshop. They need funds to keep improving it, patenting stuff so that other don't abuse their ideas (imagine those patents under a General Public Patent License). The designers want to have a Photoshop. But Adobe will have no way to charge for it, except by redistributing it. But who want's to pay for a Photoshop CD or download from Adobe when you'll have it bundled in your RedHat (or favorite) distro? So Adobe doesn't want to GPL it, GIMP doesn't get enough funds to improve it faster and the designed end up having to use a lower quality product (GIMP) or having to pay for Windows and Photoshop.

The GPL effectively help the distribution of software but not the developement of software. The fGPL helps developement of software and doesn't harm redistribution.

But I do not expect any developer that's getting paid by a distribution, the goverments or a big company to agree with me. Any other developer that's _trully_ independant (from distributors, universities, tax and big corps money), did a great contribution for OSS and can't get the funds to keep on working, should agree.

Federico

Replying to Alexander Viro:

One remark is - Microsoft actually "conquered" the market by adding features _only_.

Paul P 'Stingray' Komkoff 'Greatest' Jr /// (icq)23200764 /// (http)stingr.net
  When you're invisible, the only one really watching you is you (my keychain)

This is totally unrelated to the ecconomic model, we have many proofs that code quality is unrelated to financial compensation. People write crap code for both fun and profit. So what you say is totally true, but has zero to do with why the author wrote the code. You have to QA any code before using it, why the developer wrote it is irrelevant.

That said, let me suggest another possible model for funding free software. It's in two parts. It would be helpful perhaps to discuss "is this a good thing to do" first, before "how could you do that," and I do know some similar things have been proposed and tried in the past.

First the user driven part:

1. User wants functionality X. Defines a functional spec. Also defines the open source license to be used for the finished work. GPL, LGPL, BSD or similar.

2. Concept approval. If the functionality is a library change, driver, kernel hack, or similar, the entity in charge commits to accept the functionality *if and only if* it meets some standard of fitness, reliability, and maintainability. Otherwise some neutral party (user group, FSF, individual) is selected as the acceptor.

3. Bidding. At that point the proposal goes up {somewhere} with a price (I will pay $Y for functionality X). Developers may offer to do the work, and most importantly other users may add to the offer. Not a fixed $20, but whatever it's worth to them. After some time either an acceptable developer is found or the offer expires.

4. Acceptance. Developer tests the software, publishes the result. The acceptor tests the software and either accepts it or rejects it for a objective reason related to not meeting the functional specs.

5. Deployment. Developer gets paid, software released under open source license, if it's part of a package it's queued for the next release.

Developer driven part (only changes from user shown):

1. Definition. Developer states s/he wants to develop X, states functional spec, license, and price of the work.

3. Bidding. If the stated desired price is not met after some time, other developers may accept the current pledge (with user approval, of course).

bill davidsen <[email protected]>
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.

No matter how many proofs you have or how good they are, I won't believe it
because this fails the giggle test. Here's a simple counter-proof. I want to write an SQL server from scratch. I create two teams, one with $50,000 and one with $5,000,000. You can honestly tell me that it's equally like that either team will produce a higher quality SQL server?

This reminds me of the proofs that supposedly showed that locking up
convicted criminals for longer didn't lower the crime rate. Are we honestly supposed to believe that otherwise honest people commit more crimes to make up the difference?

DS

Intuition is misleading. It mostly depends which of the teams has the good engineers and smart management. The .com bust is just one example of that.

In some cases yes. There are a whole variety of well understood reasons why this occurs - probability of capture not length of capture is the deterrent. Most criminals won't reoffend after a short or long sentence (and there is little evidence that length of sentence decreases the probability of reoffence. In addition there is a category of crime (that involving a dispute between two parties) where one person being sent to jail causes their entire family and relations to become 'criminals'.

I don't however see the relationship between the two, other than both being demonstrations that you must do the actual science and statistics before you rely on intuition.

I don't get it.. he should know full well there are programmres out there who will not provide good code no mater how much you pay them.

In fact, I can name at least one example from the past employees of the company providing his email address.

Also see "Brook's Law" from here:
http://info.astrian.net/jargon/terms/b/Brooks_s_Law.html

Gerhard

--
Gerhard Mack

[email protected]

<>< As a computer I find your faith in technology amusing.

Sure, and there are people who can't write a symphony no matter how much you
pay them. But, believe it or not, the people who can write symphonies are more likely to get paid to write them than people who aren't. Or, in other words, there will tend to be a strong correlation between how much a person is paid to write symphonies and how well he or she writes them. This is because in general the people who control the flow of money tend to pay the people who write the symphonies that meet their needs.

Now if you do a study where you randomly choose 100 people and pay them an
extra $500 to write symphonies, you are very unlikely to get improved symphonies. But this hardly shows that more money won't make better symphonies.

For an individual given programmer, in most cases code quality won't
correspond to financial compensation. You can't pay a randomly selected person more and more money and find them writing better and better symphonies. However, if I had to get a symphony written by any means at my disposal, the more money I had, the better the symphony would be.

DS

[email protected] (David Schwartz) wrote on 01.08.02 in <[email protected]@whenever>:

Unfortunately, the computing field is full of counter examples to that.

They tend to follow a fairly simple logic: If I hire an unskilled worker, I can pay him less; the problems with that approach only surface in one or two years. Then I simply replace one monkey with another monkey - assuming it's even still my problem and I haven't moved to a different job meantime.

Thus, as Alan said, the .com bubble.

Thus, Microsoft.

Unfortunately, reality does not seem to bear that out.

In general, that's exactly what they don't do. In fact, this not happening in reality is exactly why the Berne convention was written to reserve rights for half a century after the death of the creator.

I really, really doubt that, *unless* you *yourself* already understand (and I mean really *understand*) symphonies to start with.

Unless you can judge the quality of composers yourself, the money is rather unlikely to buy you a good one.

MfG Kai

First you say that nobody knows what a good programmer is and it's virtually
impossible to tell the difference. Then you assert that I won't find one. Well, if your first argument is correct, then there's no way to know whether I've found one or not, and it becomes impossible for you to assert that I won't find one.

To put it another way, if nobody can tell the difference between good and
bad programmers, then it doesn't matter whether you have a good or bad programmer. If it matters, it must be possible to tell the difference.

Even if the only way you can tell is to wait until you're done and then see
whether you have a good or bad program, with enough money, you can keep trying over and over.

Suppose one in a hundred programmers are really good and they all charge the
same but nobody can tell the difference. The more money I have, the more independent programming teams I can hire, and then I can choose the best result. Unless you want to argue that I can't even tell the results apart, but then you're back to their being literally no difference between good and bad programmers.

Unless you want to make an argument that more money actively causes harm,
you can't do anything about the obvious fact that more money means more options.

The original argument that money doesn't produce better code quality was
based upon programmers getting more money, not projects getting more money. With this I agree, in general, paying a person more won't make them produce better code. But it doesn't follow that this applies to projects or even to groups of programmers.

DS

They're both example of how the properties of individuals differ from the
properties of groups of those same individuals, and I believe the same error is being made in both cases.

In the programming case, a programmer working on a project will not be
likely to produce better code simply because his salary goes up. Generally, people will produce the best code they are capable of producing. Similarly, longer lengths of incarceration won't decrease the probability that a particular person will re-offend. In both cases, there just isn't a cause and effect relationship with the individual.

However, this does not mean that a group will behave the same way. For
example, if we lock everybody who commits a violent crime up for twice as long, crime in the group as a whole will go down simply because repeat offenders will spend less time out of prison.

Similarly for programmers. With more money, you can employ more and/or
better programmers. You can still employ cheaper programmers, it doesn't stop you.

How many people currently working on the Linux kernel would devote more time
to it if they received regular anonymous donations? How many new talented programmers who can't find employment (or don't want/need to) would take more interest in working on Linux?

I'm not saying this guarantees better code. For example, it's possible that
the greater volume of code produced might overload choke points in the development that can't be expanded due to key individuals already working at their limits. However, money brings new options.

As you (I think?) pointed out, most free software is crap and most
commercial software is crap. The hard part is to find the good software and the good people and then give them an incentive to produce the code that you really want. Money is a tool for doing that.

DS

First, we were talking about written for free vs. written to make money. Second, the quality of the output depends on the quality of the process, not how much you pay for it. Equally likely isn't what I said, either.

One last time: commercial software is not a guaranty of quality nor is being free an indication of being shoddy. Clearly if you underpay people for any work you are likely to get poor work, but that doesn't apply to someone who is being paid in satisfaction and recognition, and who has a real motivation to do it to the best of her/his ability.

Glad it reminds you, I sure as hell don't see the point... and I never saw any such thing. Studies show that locking people up longer doesn't make *that person* less likely to commit a crime, which is not at all the same thing as the crime rate in crimes per unit time by all persons.

bill davidsen <[email protected]>
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.

I give one person $2,000 and one person $50,000 to buy a car. Would you
argue that they are equally likely to come back with quality cars?

Now you're changing the argument on me. First we were talking about code
quailty and financial compensation. Now we're talking about code quality and cost. The two are really not related. A person can produce a product that is free either with or without being paid.

Yes, but who will that person be and how many of them will there be? That
will depend upon how much money is available. Some people will do wonderful work for free, and if you have no money, those are the only people you can use and you get as much time as they can spare or afford to give at best. If you have money, you can still use those people, but you can also use people who need money.

This is exactly the point. A given programmer may not create better code
with greater compensation. However, a programming project can create better code with greater compensation. (Assuming we don't randomly pick projects and dump money on them, of course. Assuming the money is at least placed by the people who went to the trouble of earning it choosing where they think it will do the most good.)

DS

If you send both to area famous for seedy dealers? Yeah - both will come back with painted pieces of shit, unless they'll have uncommon luck to find something better. And that's a *BIG* luck.

Face it - in a lot of categories the market of software is market
of snake oil. When all alternatives on the market are utter crap and that situation hadn't been changed in decades... And I mean all - be they commercial, free, for Linux, Solaris, Windows, VMS, HP-UX, yaddda, yadda; it doesn't matter.

It's nice to pretend that this is not the case. Just don't forget
to pray when you view a PDF or PostScript document from some site. Or listen to mp3. Or view a picture in one of $BIGNUM formats.

And don't give me that crap about Sturgeon's Law - for many categories
we are talking about 100%, not 90%. Sheesh...

First:
I have received a rather harsh complaint email that told me that it was considered to be RUDE to post to Subject:
"Funding GPL projects..(etc)" in context of LKML.

I agree in that Subject Sentence is rather [OT], but it seems like SUBJECT is of interest to air for many LK developers (and others). Therefore I propose the above change in subject line to narrow the subject down to specifically mean kernel related development projects. I hope that no one is offended by this remark of mine. Most postings so far in subject does in practise touch funding of kernel related development in one way or the other.

[ snipped ]

I believe that a lot of kernel development funding is already ruled somewhat (unofficially of course ) by similar structures as above. Only: The User in point 1. is a larger company such as f ex IBM. By making the scheeme above available to anybody would probably be a nice improvement. Some question that remains to be solved is however:

A. Who will want to make the nescessary work that is
involved to achieve this? I do believe that most of the Linux kernel development team are a bit too buzy making linux patches, while working on ordinary jobs on the side.

B. How should the work be organized in order to achieve
project credibility and fair progress ? There are numerous court examples of funding abuse.

C. The scheeme above might actually turn out to "need" funding
for it self. There is even a big "riscue" that it will it
self eat up all the funding that it receives. ( Old bad own experiences. )

Kind regards
/Daniel Mose

P.S.

If any body actually find my posting in subject above to be rude I would of course be happy to know this by mail. Being rude is certainly NOT my intention. Of course any kind of feedback besides this is always welcome

D.S.

egads! the price of cheese is of interest to many LKML readers, but that DOES NOT MAKE IT a legit topic. please, please direct any followups to alt.linux.kernel.cheese.

Hi,

...will certainly kick most of the payers.

Thunder

If this model could be enforced, i.e., everyone had to pay, then this would indeed be a revolutionary change to how software is developed. It would bring in more than enough money.

One problem I see is that you'd be talking a huge amount of money, potentially money on Microsoft scale. Managing that money, making it go to the right places, without it sticking to the fingers of management, isn't likely to happen. You'd need a real corporate structure to do this and I suspect it would fail because noone would trust them to do the right thing. There are plenty of people who don't trust the FSF now. Imagine what the feeling would be if $2B/year were headed their way.

One possible answer is to make each program its own profit and loss center or corporation. But now you have to send $20 to the kernel.com people and $20 to the apache.com people and $20 to ...

Another problem is that GPLed software is essentially software in the public domain. Many people in many parts of the world will not obey the license and will just stop shipping the source. Yes, you can catch and pressure some of them, but you'll not catch the majority of them, just the dumb ones. We found this out with BitKeeper, people downloaded the source and promptly removed the openlogging feature, even checking in a changeset with comments like "Disable that stupid openlogging feature". You couldn't find a more blatent violation of our license if you tried, but that doesn't stop people from doing it.

The problem looks pretty intractable to me. I'm glad you are thinking about it, I'd much prefer a world that was closer to open source than to Microsoft. I've thought about it a lot and my attempts have pretty much failed, so it's encouraging to see someone else thinking hard about this.

---
Larry McVoy            	 lm at bitmover.com           http://www.bitmover.com/lm 

Larry, what the hell are you smoking? It's a repost from tabloid, for fsck sake - clearly says so in the beginning. Thinking is, indeed, hard for these guys, but what's encouraging about that?

Even tabloids can have good content, though I agree it's rare. After watching all the dot com and open source companies take a dive when it became apparent that no business model == no business, it's somewhat nice to see someone besides myself trying to figure out an answer which is actually sustainable. It's not an enjoyable thing, everyone hates you if you don't work all day long on their problems for free, but it's a good topic for further thought.

I think that the guy is on the right track, I've frequently described software sales as similar to insurance, noone pays what it actually costs to handle the problem, everyone pays a little and the cost is spread out over everyone. The more people who pay, the less each has to pay, and that seems to be his message. I agree with that. He's basically right in theory, the problem is putting it into practice looks hard or impossible. But maybe someone will figure out a way, so I'm trying to be encouraging.

Consider this my obligatory, once a year, "that's a good idea" post. Intended to balance out the zillion other posts saying "that's braindead". :)

---
Larry McVoy            	 lm at bitmover.com           http://www.bitmover.com/lm 

"Bat-Boy creates open-source project"

After defeating the evil-doers of the world Bat-Boy has taken up another cause - titanic software companies. He vows to tirelessly create sourceforge.net projects one after another until the sheer weight of them collapses onto, and crushes, Microsoft.

"What's our business model?", asked Bat-Boy through a translator. "I eat vermine and insects so I don't need much money. When I do need money, I sell photos of myself to tabloids".

Bat-Boy, you do us all proud!

} Even tabloids can have good content, though I agree it's rare.
} After watching all the dot com and open source companies take a
} dive when it became apparent that no business model == no business, } it's somewhat nice to see someone besides myself trying to figure out } an answer which is actually sustainable. It's not an enjoyable thing, } everyone hates you if you don't work all day long on their problems } for free, but it's a good topic for further thought.
}
} I think that the guy is on the right track, I've frequently described } software sales as similar to insurance, noone pays what it actually costs } to handle the problem, everyone pays a little and the cost is spread } out over everyone. The more people who pay, the less each has to pay, } and that seems to be his message. I agree with that. He's basically } right in theory, the problem is putting it into practice looks hard } or impossible. But maybe someone will figure out a way, so I'm
} trying to be encouraging.
}
} Consider this my obligatory, once a year, "that's a good idea" post. } Intended to balance out the zillion other posts saying "that's braindead". :) } --
} ---
} Larry McVoy lm at bitmover.com http://www.bitmover.com/lm
} -
} To unsubscribe from this list: send the line "unsubscribe linux-kernel" in } the body of a message to [email protected]
} More majordomo info at http://vger.kernel.org/majordomo-info.html
} Please read the FAQ at http://www.tux.org/lkml/

Wasn't there a South Park episode making fun of the dot coms where they were starting a business and someone asked about the plan and they said

Step 1: Idea
Step 2: Err...
Step 3: Make money

Or something like that. Where ever it was, it was pretty funny, it was quite obvious that the vast majority of the failures could be traced to a lack of a step 2. Seems so simple, doesn't it?

---
Larry McVoy            	 lm at bitmover.com           http://www.bitmover.com/lm 

http://www.fool.com/news/foth/2001/foth011108.htm

} Wasn't there a South Park episode making fun of the dot coms where they } were starting a business and someone asked about the plan and they said }
} Step 1: Idea
} Step 2: Err...
} Step 3: Make money
}
} Or something like that. Where ever it was, it was pretty funny, it was } quite obvious that the vast majority of the failures could be traced } to a lack of a step 2. Seems so simple, doesn't it?
}
} On Fri, Jul 26, 2002 at 11:08:26AM -0600, Cort Dougan wrote:
} > "Bat-Boy creates open-source project"
}
} --
} ---
} Larry McVoy lm at bitmover.com http://www.bitmover.com/lm

Would it have made any difference if I posted it here first? Or if I was a respected developer? Shame on me, I happened to have studied economics.

Anyway, the OSS model is just fine, so you shouldn't care. If at any point you change your mind and want/need my $20 bucks, I will be ready to support you (as well as the hundred millions x $20 waiting out there arround the globe).

Federico

If you want to donate money to a project(s), I'm sure that a simple email to the maintainer(s) of that project would get you an address to send such a donation to. Myself, I feel that donation should be optional, besides the obvious practical difficulties with something like this (e.g., the task of keeping track of who has paid, not to mention the necessary restrictions on downloading at various sites, not to mention the fact that people will end up just getting source and posting it for everyone, not to mention...)

Some people don't want to be only ones donating, some others just will just wait & see if somebody elses money saves the project. Also, your money may not be enough to fund that project and you are left with no improvements. Or you could be overfunding the project (never seen this happen though). You have no warranties where and how the money would be usued.

A fundation focusing on this issue alone would be the best solution for solving this "fragmentation" problem.

No problem here really. No restricted downloads, not anti-piracy code. No added burdens. No BSA. Just the legal obligation to pay $10 a year. If you don't pay, nobody will come looking for you, but at least you'll know you owe somebody real money.

Federico