I mistakenly sent this off-list. I'm sending now on-list, and Alex's reply with permission:
----- My mail: From: "Alex Hudson" < home@alexhudson.com>
Well, you seem to have an unwritten implication of "unless it's cheap it's artificially expensive", which would be difficult to argue against :) Cups of coffee are very, very cheap to make - but you don't see the likes of Starbucks selling them for pennies/Euro cents, in fact, the price is actually going up not down.
OK, maybe my use of the word artificial is unclear. What I was referring to was, in essence, to legal (or other kinds) of restraints on trade and distribution, such as information monopolies like Copyright, or legal requirements of huge insurance (like in the case of digital certificates). These things are not inherent to the dynamics of producing, trading and distributing the goods, but instead are artificial man-made obstacles that help to monetize the good (often to the point of super-profits). For instance, consider what a copy of a song would cost, absent copyright.
Many consumer goods are now extremely cheap to make. Obviously, there are other costs involved - particularly fixed costs of staffing and rent if you have a shop or something - but the actual cost per unit is pretty much unimportant for large numbers of products.
Somewhat true, but I would argue that although a cup of coffee looks a lot like a good, for all effects it is a service. How much of a service it is can be clearly seen in how much people are willing to pay to have one at a café, versus having one for free in their office. Services have a very different cost structure.
And of course there are other types of product, like VOIP service, where the cost of having a customer is extremely low (sometimes nothing, depending on how your call routing works). Yet many people pay handsomely for the service.
Again, it seems clear to me that people aren't paying for a good here, but for a service. Essentially people are paying for not having to bother about it. Personally I think it would be a lot more rational to create some kind of consortium that would administrate the service for a set of businesses, so they can share infrastructure and costs and enjoy an economy of scale plus avoid paying the profit for the service providers, but for whatever reason this is not happening, and there might be something I am missing.
I think you're conflating two principles there. Obviously, having a monopoly gives you the ability to make mega-profits. The fact that some leading proprietary software makers are in that "happy" position makes it difficult to talk about, because I would argue that it's the monopoly which creates the massive profits, not the proprietary model.
But both things are intimately entwined. One point I made which you didn't touch on at all is that Copyright is a form of a monopoly ipso facto. There can be only one rights-holder which sets the conditions for copying and distributing the good. Given that software is so complex (if we imagine every line of code is a distinct piece for instance software is the most complex thing humans build) and so difficult to characterize, you are never (or hardly ever) going to be able to compete on the basis of perfect substitution, it's just not going to happen. So all proprietary software of a given complexity has a monopoly of a sort. That said, clearly not all monopolies are made equal, and some are going to be vastly more profitable than others, no question about that. In contrast, anyone can trade with and around Free Software, thus competition in the context of perfect substitution becomes possible, and one gets non-monopoly pricing but competition pricing. This is a bit less true than it is likely to be in the future at this point, because sufficient expertise to deliver goods and services in the free software world is not perhaps developed to a sufficient extent, and so there are not that many people competing in the space of, say, Red Hat, but I have little doubt this will change.
That's really an argument about business model. Getting users to sign up for software and amortise the costs that way is a great idea; it's thoroughly impractical though, because you'd need many users up-front willing to pay the price. The lifespan of a piece of software is usually a few years, which is much easier to write costs off over. It's the same for free software, too.
I somehow don't see this argument at all. Say a user A needs or will need a piece of software to do X. Why does it make more sense for X to buy it at monopoly pricing, having close to zero input in the software's feature set, having no freedom to get someone else to improve the software at competitive prices, than pay in advance, perhaps in installments, pooling with other users, and potentially backed with an insurance if the software ends up not being produced, after which A gets an unencumbered copy of the software which has at least more of a chance to being taylored to their needs?
I think discussing this in terms of Microsoft, Adobe, Oracle, SAP, and Apple, IBM, etc. is really missing the point. These guys are global multi-national corporations. They're not representative of proprietary software houses, and they're pretty much bound to make huge profits no matter what they do, because that's the nature of the beast: they serve a huge market and have fearsome economies of scale.
Somewhat agreed, I just suspect that profit margins of proprietary software houses are generally high.
And it's not like software licensing is the highest margin product necessarily: in Oracle's last 10K filing, they say that their support business has margins of 84%. Practically every one of their software customers buys support as well, which is an ongoing revenue.
Fair enough, services can be extremely profitable too. But consider that this is an ancillary service related to a good that is held in monopoly. Noone but Oracle can give that level of service for Oracle products. It is precluded in the vary fabric of proprietary distribution. Oracle is not competing to service Oracle products in equality of conditions to other companies. So this is just another front on which proprietary software houses can extract super-profits, by charging non-competitive rates for services they alone can provide.
At the end of the day, prices are set pretty much what a market will bear, give or take a business's goals wrt. volume. I don't think software, free or not, would work much differently: although I do think there are well-known proprietary vendors who have a very comfortable monopoly in the market, I don't think that it's actually that common.
It must be kept in mind that software has artificially (through copyright) been made into a rivalrous good. So although the market has something to say, there are important market distortions in place1. Distortions which would not take place in a Free Software context. Market pricing is notoriously distorted by situations of monopoly or even sufficient asymmetry in power between agents.
I also don't think free software is any kind of economic tool: while the ability to share notionally does change things, in practise people copy proprietary software so much I don't think it changes much at all.
This is quite true of individuals and SMEs. It is much less true of bigger companies and governments. In addition, proprietary software which is sufficiently rare or specialized can be very difficult to find for people to get copies at no cost. Also, there is an important tendency towards utilizing technical measures in order to place obstacles to the free copying of proprietary software, which makes this argument more compelling. While I doubt that there will be perfect copy protection, sufficiently good copy protection acts as a substantial barrier that enforces proprietary pricing.
1: I am not saying that a free, unrestrained, unregulated market is necessarily what we should be after, just making observations on what differences there are, and what consequences derive from them. --David. End -----
----- Alex's reply:
David,
(Did you mean to post this offlist? Feel free to forward my reply onwards if not).
On Wed, 2007-11-07 at 16:40 +0100, David Picón Álvarez wrote:
From: "Alex Hudson" <
home@alexhudson.com>
Many consumer goods are now extremely cheap to make. Obviously, there are other costs involved - particularly fixed costs of staffing and rent if you have a shop or something - but the actual cost per unit is pretty much unimportant for large numbers of products.
Somewhat true, but I would argue that although a cup of coffee looks a lot like a good, for all effects it is a service. How much of a service it is can be clearly seen in how much people are willing to pay to have one at a café, versus having one for free in their office. Services have a very different cost structure.
I don't think you can argue a cup of coffee is a service. Being able to drink it in a café is a bonus, but not required - there are plenty of street vendors (e.g., AMT in London) who sell them on the street for the same price. The product is definitely the drink.
For VOIP, I agree it's a service, but I was pointing out that people will pay a lot of money for something which cost-wise "isn't much".
But both things are intimately entwined. One point I made which you didn't touch on at all is that Copyright is a form of a monopoly ipso facto.
Yeah, I did ignore it, mostly because I don't think it matters.
For example, Red Hat sell their Enterprise OS and Cent OS give it away. Their desktop costs ~£200, and that's before you add in the various non-basic support options. By your logic, they're making huge profits on each copy, but the price hasn't changed since the competitor giving it away for no cost (CentOS) came into the market. In fact, I think they put prices up after that - and it's effectively the same product with a different name on it.
So, when you say:
In contrast, anyone can trade with and around Free Software, thus competition in the context of perfect substitution becomes possible, and one gets non-monopoly pricing but competition pricing.
..... I don't think that's actually borne out in practise, and my point is that you're not correctly distinguishing between monopoly and non-monopoly pricing. The monopoly effect of copyright is actually pretty limited.
[on amortising fixed-costs up front]
I somehow don't see this argument at all. Say a user A needs or will need a piece of software to do X. Why does it make more sense for X to buy it at monopoly pricing, having close to zero input in the software's feature set, having no freedom to get someone else to improve the software at competitive prices, than pay in advance, perhaps in installments, pooling with other users, and potentially backed with an insurance if the software ends up not being produced, after which A gets an unencumbered copy of the software which has at least more of a chance to being taylored to their needs?
Simply because you'd never get the numbers of users to sign up to make the costs bearable.
Software sales is almost classically entrepreneurial: "I will take the risk of this up-front cost in order to try to make profit later". Your pitch above is almost the reverse of that, "I will do this thing if there is no risk to me".
If you look at the numbers, in the first scenario you say "I will sell this product for X years, to Y people per month at price Z, giving me a return of (X*12)*Y*Z". In the second, you need to get Y people to sign up to a price Z such that Y*Z at least covers the cost of development.
By taking on the risk of the cost, the first scenario can stand a much lower Z than the second because it has the X*12 to counter-balance.
I'm not saying it doesn't happen, but if you think of any product where people can order to measure or buy off-the-shelf, the custom job is almost always more expensive. There are no economies of scale, the price may not be fixed, and you have to wait for it to be made. From the point of view of the entrepreneur, although it's more risky there are also chances of better profits, which is essential.
I also don't think free software is any kind of economic tool: while the ability to share notionally does change things, in practise people copy proprietary software so much I don't think it changes much at all.
This is quite true of individuals and SMEs. It is much less true of bigger companies and governments.
That's probably right, but to bigger companies and governments, the costs of software licensing are pretty irrelevant save for examples such as Microsoft.
E.g., SAP are obviously a proprietary company. SAP also costs a lot of money. But when you buy it, you get a hoard of business process consultants come and tweak it all up for you, and what you have is pretty individual. The ongoing support and maintenance of that software is pretty much what you pay the money for. SAP are also extremely profitable (last time I looked, anyway ;)
Why is there no free software competitor to SAP? The licensing costs don't really come into it - I can speak from experience. I would also argue that the licensing of it makes no difference the majority of customers because the various installs are so custom: it's really that aspect the customer is paying for, and there is no benefit to being proprietary in that model, the lock-in exists by nature.
To be honest, it's very similar with finance software. The software is pretty easy to put together, and is generally pretty breathtakingly expensive. There is only limited value in the software itself, since the tax rules and all the other bits are what really matter. It could easily be implemented in free software, and customers won't care one way or the other because they're basically buying a service. Again, though, free software has made no significant in-road into that market.
It's pretty easy to say that monopoly-style conditions exist with proprietary software, and in obvious cases that's clearly true. But there is also a lack of competition from free software in markets which on the face of it would suit that model *extremely* well. Yet, it doesn't happen.
There is a real issue with the extent to which free software scales over enterprises, and I would argue the majority of free software at the moment is mostly suitable for small/medium sized organisations. People at enterprise/governmental level aren't just not buying free software because they don't like it; they're also not buying it because it doesn't exist.
Cheers,
Alex. End -----
---AV & Spam Filtering by M+Guardian - Risk Free Email (TM)---
Just pulling out one topical point in this recent discussion:
On Wed, 2007-11-07 at 18:51 +0100, David Picón Álvarez wrote:
Why does it make more sense for X to buy it at monopoly pricing, having close to zero input in the software's feature set, having no freedom to get someone else to improve the software at competitive prices, than pay in advance, perhaps in installments, pooling with other users, and potentially backed with an insurance if the software ends up not being produced, after which A gets an unencumbered copy of the software which has at least more of a chance to being taylored to their needs?
Via LWN, I found out that there is a new games company which is trying to run their business on a model close to this:
http://lwn.net/Articles/257457/
The idea, as I understand it, is that they will release a demo of the game and a pledge people can pay into. Once the pledge reaches a certain amount, and people pay up, the game is released as free software under the GPL.
The comments from LWN posters are pretty much what I would expect: asking for $40,000 seems to be asking too much. Yet, that would only buy ~6 man-months here in the UK, if that, which is practically nothing for a new game. And that would be at cost, too - and you can't run a business without turning a profit.
If they sold the game on a proprietary basis, at maybe $8 a pop, they'd only need ~5000 sales to break even and I doubt anyone would be complaining about the price. Psychologically, there is a lot going on there in terms of price, even before you get to "why should I pay for something so others can have it for free?".
It will be interesting to see how they do. Pledges to liberate existing software have been relatively successful, and a number of games have become free software after having a proprietary existence without the need of a ransom at all.
I'm of course not saying that the success of this particular company will dictate whether or not the model can work: it might be that the games they produce are crap, and the business would fail under any model. But it will be interesting to see if they can make it work. I don't think they can; but I'd be quite happy to be proved wrong.
Cheers,
Alex.
On 08-Nov-2007, Alex Hudson wrote:
The idea, as I understand it, is that they will release a demo of the game and a pledge people can pay into. Once the pledge reaches a certain amount, and people pay up, the game is released as free software under the GPL.
The comments from LWN posters are pretty much what I would expect: asking for $40,000 seems to be asking too much. Yet, that would only buy ~6 man-months here in the UK, if that, which is practically nothing for a new game. And that would be at cost, too - and you can't run a business without turning a profit.
If they sold the game on a proprietary basis, at maybe $8 a pop, they'd only need ~5000 sales to break even and I doubt anyone would be complaining about the price. Psychologically, there is a lot going on there in terms of price, even before you get to "why should I pay for something so others can have it for free?".
Your comparison seems to point to the way out of this knot. Instead of saying "We need a total of $40 000", why not say "We need $8 each from at least 5 000 people; if you can pay more, that's welcome"?
If the latter *still* isn't attractive, then I don't see the point of that particular comparison.
On Fri, 2007-11-09 at 08:46 +1100, Ben Finney wrote:
On 08-Nov-2007, Alex Hudson wrote:
If they sold the game on a proprietary basis, at maybe $8 a pop, they'd only need ~5000 sales to break even and I doubt anyone would be complaining about the price. Psychologically, there is a lot going on there in terms of price, even before you get to "why should I pay for something so others can have it for free?".
Your comparison seems to point to the way out of this knot. Instead of saying "We need a total of $40 000", why not say "We need $8 each from at least 5 000 people; if you can pay more, that's welcome"?
If the latter *still* isn't attractive, then I don't see the point of that particular comparison.
No, I think you're missing the point. With a pledge system, you're making a promise that the proprietary vendor isn't: you're saying "when I get $40k, I'll free the game". You're placing an overall value on the game.
Nobody bought Quake from id software on the expectation that they would GPL the game, but offering to free the game as part of the value proposition changes the equation. And that's why I don't really think it will work.
The other problem is, when do you release? Do you:
1. release immediately as proprietary software, release as GPL once the target is reached; 2. release as GPL once the target is reached.
If you do 1., you're likely to never reach the target because the closer you get the more people will decide to hold off until the GPL release is made (or, you keep the fund information private and face accusations of fraud when the target isn't reached :). If you do 2., you give the people who paid the money no benefit above non-paying people.
It's a really tricky scenario to get right. From history, we know that delayed GPL releases tend to work pretty well: people are willing to pay up for apps which are later free'd, and that's a model where pure software licensing sales can fund the development. However, it does seem to preclude community involvement pretty effectively, which could be an issue. And without that time lag, you face all sorts of problems about how and when the code is released under a free license.
I would love for these guys to succeed, I just really don't think they will.
Cheers,
Alex.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Alex Hudson wrote:
Nobody bought Quake from id software on the expectation that they would GPL the game, but offering to free the game as part of the value proposition changes the equation. And that's why I don't really think it will work.
The other problem is, when do you release? Do you:
1. release immediately as proprietary software, release as GPL once the target is reached; 2. release as GPL once the target is reached.If you do 1., you're likely to never reach the target because the closer you get the more people will decide to hold off until the GPL release is made ...
I don't think game production at a scale and with the budgets of modern commercial games will ever be possible in a comprehensive GPL framework.
I rather think (along the lines of what Stallman said of the matter, years back), that we should consider a game as consisting of two things: Software, the engine that drives the gaming, and art - the artwork, music, world design, animation, gameplay and plot of the game.
The *software* should (or must, as Stallman would say) be free as in the four freedoms, but the art doesn't need to.
As a matter of fact, I think the commercial gaming industry would benefit tremendously from a number of free game engines. But the game as a whole would still be sold as a package you are allowed to run the game and modify, change and distribute the software - but not necesarily the artwork (the game's content) with it.
As far as I can see, it makes perfect sense. You still have to pay for the game and have to pay a license to redistribute it, but the payment for the redistribution license is for the game's *content*, not the software, which is free software.
This should be satisfactory frm a software freedom perspective and could coexist with the modern gaming industry - as well as with Creative Commons-licensed, Shareware and GPL versions of licensing of game contents as well.
best regards Carsten
- -- IT-Politik - http://www.modspil.dk/itpolitik IT og teknologi - http://www.modspil.dk/itogtech