Fog Creek Software
Discussion Board


Occasionally I think of suggesting that the performance of music, art, literature and so on should not be copyrightable but considered a work for hire, and all recordings of such.

And then I think, but then when all the programs I write in their performance would also be works for hire.

And I think Yes, and it would be a Good Thing.

Simon Lucy
Monday, July 28, 2003

They finally got wireless in your pub, didn't they? Have another.


Mark Hoffman
Monday, July 28, 2003

Maybe I don't get it, but WTF is the difference between "copyrightable" and "work for hire"?

You "work for hire" for your employer, who, because of this, owns the copyright to your work.

If you mean that software should not be "copyrightable" - so, if I write something, by default everybody could copy it and used it for free, then I would stop writing software.


I like writing software, but I also have to eat. Because of this, if I couldn't earn money by writing software, I would simply move to another field which allowed me to earn money.

It's that simple.

John K.
Monday, July 28, 2003

Oh, well it was almost entirely throwaway.

But in this context I was implying that those that pay for the performance should 'own' that performance and of course in the end a great many people may pay for a performance that ends up being recorded and they own their own particular intact fragment of it.

This does not mean people do not get paid, this does not mean that copyright of original works as copyright was originally intended doesn't exist.

What can people usually do if they buy an object?
They can make legitimate use of it, legitimate doesn't necessarily mean sensible, they can nail it to the wall if they feel like it.

They can dispose of it, they can sell it someone else, note this does not mean make copies and sell those copies as if they were the original item, but they might be able to legitimately broadcast it for their own use.

In truth there's little difference in treating  a copy of a performance as breach of copyright or theft  but it is more honest and might make it clearer and more understandable exactly what is being done.

In software terms it could clear up the issue between performance and non-performance since if the performance of software was not acceptable and didn't meet the description the purchaser could get reasonable redress.

And if someone thinks that that is what we have now with licencing then no it is not.  A licence to use a rock you bought from Fred is not the same as buying that same rock and being able to sell it to Jim at some point in the future.

Simon Lucy
Monday, July 28, 2003

I believe there exists something like this... it's called open source  ;)

Monday, July 28, 2003

Hmmm, no, that's different.

Simon Lucy
Monday, July 28, 2003

Probably... but not much.

Monday, July 28, 2003


With very few exceptions, making an open-source project effectively eliminates the possibility of you getting paid.

There are some people or companies who get paid from open source, but there is a very, very small number.

From the open source companies, very few are profitable - almost all of them are losing money.

Is this a future we are all looking forward to see?

I would rather see a future where programmers have decent pay for their work.

John K.
Monday, July 28, 2003

Simon said, "But in this context I was implying that those that pay for the performance should 'own' that performance and of course in the end a great many people may pay for a performance that ends up being recorded and they own their own particular intact fragment of it."

Well, in that case you might well end up paying more to see that performance.  For example, take a symphony that's preparing to do a live performance that is also being recorded for publication as a CD.  If the symphony is denied the right to distribute its performance's IP and make money off the CD, then they will likely wnat to be compensated for that loss through increased admission fees from the audience.

I think fees for software development often work the same way.  I generally do work as "work for hire", thus the client receives merely a license.  But if a client wanted to pay me more money I would sign over ownership of the IP I create.

Legal ownership rights in a particular thing have often been described as like "a bundle of sticks".  You can dole out the sticks anyway you want, despite the fact that certain sticks have often been sold together.  There are a multitude of different ways ownership can be split up between different parties. 

A savvy consumer will pay only for the portion they need.  For example, most consumers have no need or desire for the right to sell and/or distribute copies of software they have custom developed.  But the developer often has a need to retain that right, if not to distribute the program as a whole, then to retain the right to use particular modules of the whole in subsequent projects.  The most efficient use of money is to have client pay only for what they need (i.e., a license) and have developer retain rights it needs to develop more software using the same modules. 

This allows each part of the ownership bundle to be allocated to its "best and highest use".  The license is more valuable to the client.  (In fact, licensing to the client cost the developer virtually nothing other than the lost right to grant an 'exclusive license' to another party.)  The right to distribute underlying IP is more valuable (usually) to the individual developer.  Of course there will be exceptions.  And if the client wants to purchase the right to distribute (exclusively or jointly held with developer)  the developer is certain to sell it if the price the client is willing to pay is higher than the value the developer places on his or her right to distribute.

The legal "bundle of sticks" that constitutes ownership of buildings and financial instruments has been separated out almost as far as imaginable during the last twenty years.  (Most common example is probably 'financial derivatives'.)  Each separate part is sold to the party that values it highest.  Economically this places the property in the hands of the party that values it highest and is most likely to use it as a capital asset, thus producing more and more value for more and more people.  The free economy at work.  Perhaps we will see similar breaking apart of the bundle of software IP in the future.

Herbert Sitz
Monday, July 28, 2003

I think I largely agree, and there's no reason a symphony orchestra shouldn't sell the performance its their performance, that's their work for hire.

The members of the audience paid to see the performance they got, I wasn't trying to imply each member of the audience should record their own copy.

Simon Lucy
Monday, July 28, 2003

Open Source is entirely different because there is no work for hire component, the original creator, or the performer doesn't get paid.

What I'm saying is that the performer should get paid.

Simon Lucy
Monday, July 28, 2003

"I generally do work as "work for hire", thus the client receives merely a license. "

That's backwards. If it's work for hire, the client owns the copyright in full.  Note this is US law, other countries may vary.

Monday, July 28, 2003

mb -- Quite right, I got it back asswards.  Should have just said my work is NOT generally work for hire.

Simon -- So you're saying that with regard to open source the person who creates something and makes it available under an open source license should be compensated? 

One question I would ask is, "What makes you think they're not getting compensated for relinquishing exclusive rights?"  Unless someone is an idiot or a philanthropist, they don't give something away for nothing.  (And you could make a good argument that neither the idiot nor the philanthopist do, either.)

So if they aren't relinquishing exclusive rights for nothing, what are they getting in return?

What follows are some ramblings on that subject:

First, sometimes that person is compensated.  They may be paid for their services in developing the software, for example, if an employer (e.g., IBM) directs employees to work on software that will be open source.  Or even if a developer charges for services in developing custom software for a client, retains the right to distribute, and then instead of charging for the software distributes it by making it available under an OS license.

So I take it you're saying that the developers of open source software should be paid not just for services in developing the software, but for foregoing their exclusive right to distribute and create derivative works off the code.

So the question for me becomes, "Why would anyone place their IP under an open source license, rather than retain exclusive rights?"  What benefit do they get from doing this?  And why would they do it if they don't benefit?

(1)  Lots more people are likely to use software if it's open source than commercial, at least for the typical small sorts of projects that get open sourced.  Since lots of programmers get an emotional kick out of knowing that their software is being used, this functions as some sort of 'non-monetary' benefit that encourages them to open source their software.

(2)  Lots of programmers get an emotional kick out of having their work recognized by other programmers.  This is another benefit that functions as a sort of payment for open source.

(3)  A programmer may not have time to devote to a project anymore, but wants to make it available via open source with the hope that others will improve on it, then he or she can incorporate their improvements into the original work.  Thus, the possibility of "free improvements" functions as a sort of "payment" for open sourcing.

(4)  A programmer may think that his or her software product isn't likely to take the world by storm as a commercial product, but that it will be widely adopted if offered as open source.  The programmer may adopt the idea of publishing under open source in order to build a services business around the product.  Who better for third parties to go to for service regarding the product than the author him or herself.

(5) Individuals or businesses could join forces in an open source effort in order to combine their efforts.  They gain the benefit of having more people than themselves working on the software (reason 2).  Also, a business frequently doesn't want to make money off the software (it's not in the "software business"), but it nonetheless needs to have the software to run its business.  It's cheaper to pool resources with other parties, plus you develop a superior product.  The business doing this realizes that the software developed will be available to any of its competitors.  It just makes a decision that it will focus its main efforts on the core of its business, not on differentiating itself by developing a superior software infrastructure.  (Probably because the company figures it would be too expensive for the company to do this on its own.)

Five or six years ago I think you frequently saw reasons (1) and (2) as reasons for open sourcing software.  I think that's how most of the biggies got their start, including Linux itself.  Linus wanted to have his creation used, and was happy to get recognition from others as a master programmer.

More recently I think the reasons reasons for open sourcing are closer to something like (3) or, more likely, (4) or (5).  Unlike (1) and (2), there is a perceived economic benefit from open sourcing, it's just not in the form of an outright payment for the rights.  So there is economic benefit, but it's indirect.

It would be nice if an open source author could just be compensated directly for publishing under open source.  The question is, "Who would pay?".  Open source licenses arguably benefit almost all individuals in one way or another (although certainly not all companies:, think Microsoft). 

But why would somebody pay for a benefit that accrues to all?  The government is in a position where this could make sense.  But for political reasons it's not likely to happen. 

Businesses or corporations might pay the independent open source developer, if it fits in somehow into something like reasons (3) or (4) above for that business.  And enlightened individuals or organizations of a philanthropic nature might pay a developer to make open source software.

I think there is actually plenty of stuff happening of the kind I describe in the previous paragraph, but that it's hard to separate what part of payment is payment for services in developing the software, and what part is payment for the promise to release under an open souce license. 

As one example, take a company like Apache.  If I recall correctly, Apache receives plenty of "donations" from businesses that serve to help Apache extend and improve upon their software.  Much of this is probably paid by business with a rationale related to reason (5) above:  the business figures why develop it ourselves?  It will be incredibly expensive and we'll end up with an inferior product.  Instead, let's contribute to Apache, like lots of other businesses are doing.  Sure other business will gain access to the same software we do, but we'll gain a superior product at a far smaller expense.  We'll concentrate our efforts on the core part of our business, not on software.  We're not competing with other businesses on software, we're competing in our main busines area:  xxxxxxx. 

So I think many open source creators are being paid to release the "work for hire".  It's just never spelled out what portion is paid services and what portion is paid for relinquishing exclusive ownership rights.  In some cases there is no monetary payment, but programmers get emotional benefits that prompt them to release software as open source nonetheless.

Herbert Sitz
Monday, July 28, 2003

Probably lots more errors and typos, but one I see immediately is that the second sentence in paragraph numbered (5) should be:

They gain the benefit of having more people than themselves working on the software (reason 3).  [not reason (2)]


Herbert Sitz
Monday, July 28, 2003

Here's a link to a brief paper suggesting that people are their are economic incentives (i.e., indirect payment) to publish as open source.

Herbert Sitz
Monday, July 28, 2003

Turns out there's lots of stuff on the net relevant to this thread.  Here's a good one titled "Code, Culture, and Cash: The Fading Altruism of Open Source Development".  It suggests that while in Europe the reasons for releasing open source may still be the "hacker ethic" reasons (reasons (1) and (2) in my post above), in US the main reasons for software being released as open source are now commercial.

But in both cases it acknowledges there is a tangible benefit for relinquishing exclusive ownership.  To the hackers: fame and satisfaction.  To others who release under open source: commercial benefit.

Herbert Sitz
Monday, July 28, 2003

Yes there's a lot of good points there, Herbert.

I accept there are all sorts of non-monetary and monetary ways of people receiving value for the contribution made.  However, it isn't directly proportional or even associated with specific performance.  One of the myths of Open Source is that it _necessarily_ produces better products simply because there are more Eyes and the revolving improvements always tend to improve.

There are some examples of that, Apache is certainly one, the Linux kernel another both of them succeed not because they are Open Source but because they are rigidly controlled.  The source is always open, the development not.

The benefits of the people involved in those developments is pretty much directly connected to their performance, either in the work they subsequently get or the organisation is directly funded by industry.

For the most part though most Open Software could be lost tomorrow and it would not be missed.

I'm not even specifically confining this to Open Source, it isn't about whether the source is available, for consumers this has always been of dubious merit, for code musicians its necessary just as the corpus of music written and remembered is important for instrumental musicians.

Music was the starting point because of the pressure to own and control by the music industry (who produce no music at all).  If you really do own the music you buy and you really do own the software you buy the value (as you pointed out), also rises.

The mechanics of how someone gets paid for their performance is something else, it almost necessitates some kind of levy spread over all the means of enjoying that performance.  A kind of needle time. 

When I wear a consultancy hat I get paid specifically for performance; when I busk sometimes in public either with code or curmudgeonly advice it would be nice for pennies to drop into the hat.  So long as people enjoy the performance.

At some time in the past I mused about a currency of appreciation,  a currency that could be self generated and spent on the appreciation of others and how that might affect the society it was used in. The Ego Bank.

Ramble over

Simon Lucy
Tuesday, July 29, 2003

The whole argument that a programmer get his kicks out of contributing to an open source arises from the fact that most doing this are inexperienced, and not very good.

Tuesday, July 29, 2003

Quite possibly. It is, after all, always better that people remain inexperienced and not very good. It's so terrible to work to improve yourself, especially when noone's paying you for it.

Tuesday, July 29, 2003

Simon -- I guess I have no idea what you're talking about at all.

Originally you said, "But in this context I was implying that those that pay for the performance should 'own' that performance and of course in the end a great many people may pay for a performance that ends up being recorded and they own their own particular intact fragment of it."

That was what prompted me to respond that the performers do in fact own the rights to their performance, unless they sign those rights away.  The audience pays to be present at the live performance.  If the audience wants the right to record and distribute that performance, then they can pay extra to get that right from the performers.

What exactly is the issue here?

Whether the courts eventually declare file-trading systems like Kazaa and Morpheus to be legal or illegal is irrelevant to whether copyright exists.  Even if filesharing services are legal, there's no question that it is illegal to use them to "trade" recordings.  It's just that it will be impossible to enforce copyright violations when trades take place in this way.  It's not that copyright would be abolished.  It's just that it would, for all practical purposes, be an unenforceable right.  That's a different thing. 

But I don't even know if that speaks to what you're saying because I'm massively confused.  Oh well.

Herbert Sitz
Tuesday, July 29, 2003

*  Recent Topics

*  Fog Creek Home