Slashdot Mirror
The Open Code Market
The Open Code Market
The peer reviewed magazine FirstMonday has published one paper I wrote on the business possibilities of Open Source titled "The Open Code Market".
In short, The Open Code Market aims to become a Free Market for software, as well as a market for Free Software. The OCM introduces into the Free/Open Source movement an economic incentive, to help align the priorities of Free/Open Source developers with those of the end users.
Between the final draft and publication, I received many valuable comments on the idea. I am very grateful to all those who dedicated some of their valuable time to read the paper and make insightful comments.
Some of those comments led me to an earlier similar effort by Brian Behlendorf named SourceXchange (originally at www.sourcexchange.com). After contacting Brian, his comments were probably the most valuable, since he had gone through the the actual proces of running a project similar to what I am proposing.
His main line was that:
a) while the market made profits, it did not reach the levels of profitability that were expected in the middle of the dot.com boom, and investors looked for other (more lucrative) endevours
b) That the dot.com boom also addedd costs (high salaries, etc) to his attempt and
c) that it is difficult to commoditise software creation due to the uncertainties over time / effort required to write the software, and the difficulty of the role of "project manager".
My line on all this is that
- The project may be easier to develop now as expectations for profits (and costs) have decreased, and that sourcing on low-cost countries (i.e. India, Vietnam, etc.) would reduce costs significantly.
- I also expect that the market will take off only as Free Software/Open Source (Linux in particular) moves to the mass market of the desktop, thus generating the necessary economies of scale, visibility and consumer-mass. In my idea, the target end-users of the market are not IT companies, but mainly individuals with no IT knowled (nor desire to acquire it) and Small and Medium Enterprises with small or no IT departments.
"
I think I see a new class split. The do-it-yourselfers that appreciate the quality and flexibility of open source vs. the people who just want a working computer without configuration woes. Guess which one will spawn the new millionairs and monarchs?
Always going forward, 'cause we can't find reverse.
I could see this kind of thing working as a try before you buy market - use the code and pay when you release the product.
-- "Can't sleep, clowns will eat me!"
try explaining the good side of open-source programs to people who have never written a program in their life... you'll see what I mean...
I can't see much revenue comming from ordinary computer users who would see this as a 'hacker's' network or something. where they should direct themselves is towards the programming communities and (mostly) businesses...
Did he mean the site didn't run with big losses and an unsustainable business plan?
.com boom.
I that was what was expected during the
The obvious answer is of course that the code can be seen and updated by the OSS community but IMHO most don't see the value in that and would rather stick with their own development team to do their work for them (think the Microsoft argument of having someone to fire).
My guess would be that it's effectively the same for them (since they could commission patches much the same as they could pay their programmers to update software) but again, management probably doesn't like it that they have to pay someone out of house to do it.
It's about time that someone created a method such as this to make the creation of open source code generate some funds for the programmer. If companies can see that keeping these programs open can be beneficial not only to them in the support of the OSS community but the the industry as a whole then I think everyone should be benefitted by something like this.
It's better to vote for what you want and not get it than to vote for what you don't want and get it.
- E. Debs
It sounds like you are proposing something akin to Experts-Exchange, but with a few differences. OCM would use real dollars, instead of points. OCM would also encourage the creation and reuse of code, not just answers to IT questions. OCM might incorporate private collaborative project spaces to help a shifting group of workers create commissioned code.
Two wrongs don't make a right, but three lefts do.
Is it just me or does she look like Rosie's great grandmother? Don't tell me you don't know you Rosie is you Hanna Barbara hateing WB freek.
I'm working on an open-source location search (Mobilemaps) which aims to create a collaborative market for developers/portals using it.
The right service in this case is geotargeted advertising for local businesses wanting to gain some visibility on the net. It's an extension of a proven business model used by search engines like Google.
The plan is there's no need for developers to be paid for their code, they just earn money by participating in the advertising service by operating the location search for their local area.
Suppose right now today my company wants a custom feature or a bug fixed in some OS software. We can pull out our checkbook and hire X to do Y for $Z. If it's OS work, that's just part of our contract. AFAIK, this happens already.
So what is the innovation here? To create a "market" that is basically a consulting company for OS work? Doesn't seem to be...after all, the article specifically says this is not limited to GPL (and that didn't seem to be followed up with "or other open-source license"). It says open code, but does that mean the results must be open or simply the basis must be open and the end results can only link to open libraries if they are closed?
Or is it to create a consulting company that uses contractors? I don't believe that's it either... contract consulting work is already routinely outsourced and sub-contracted.
Or is it to organize a commercial venture based on the work of OS? This might be the real heart of the issue. There is where I get mixed:
Surely funded development of OS software advances them, but this could turn sour quickly (think Red Hat!). Now, with varied customers with individual requests I'd like to think that the market could not take on a direction with its own agenda, but I would like to hear some arguments for and against this. The model indicates this is for-profit. Thus, requests from the highest bidders would probably get preference. Supppose MS throws a few billion at time-consuming features that add little value to the product. Why wouldn't this market eat that up? But think about the cost of this. While developers add features such as scrollbars changing color through a gradient to indicate how far down you've scrolled, they are not working on fundamental issues that need to be addressed before they blow up in our faces. Suppose SCO hires a company to request a feature which would infringe on SCO's IP such that there is a genuine IP infringement that they don't appear to have set up directly.
I, for one, would rather have talented developers working on things like IPv6 implementation, that commercial investors probably won't want to fund, or at least wouldn't be able to compete with people who use the market as a tool against OS.
I think any time something this scale is set up, it should be very carefully critiqued. Unfortunately, there's nothing stopping anyone from doing it if it turns out to be a set-up for OS disasters.
The only thing more dangerous than a file named -rf is renaming it -rf\ /
Mark Pratt of Beehive, Germany, has relaunched the SourceXchange idea:
Open SourceXperts.com
Complete with the lame eX. It only launched on 10 November, so be gentle with yor 'how quiet here' comments.
"The truth shall make ye fret" -- The Truth, Terry Pratchett
Did anyone else read that as "dot dot com"?
Err, hello? RH desktop...dead. SUSE...wounded by Novell. SCO is bleeding any OSS-involved company with litigation. Who's going to take linux to the desktop or "mass market" as suggested here?
I'm not saying this guy is wrong, but the OSS world has got some damage control to do. I'm confident it will pull through (my paycheck depends on it) but I would rethink using this as a justification for this project's market readiness.
You never saw a fish on the wall with its mouth shut.
I would hope that OCM would be better than "try before you buy." I have reached a stage in life where I want products that do what I want them to do. I don't care to spend time trying software in a production environment because I find that so much of it is a waste of time.
OCM would seem to support the commissioning of software projects or modifications. Thus I could go the Market, submit a request (e.g., "I want a real-time transparent version of CVS") and various groups might make counter-offers of price and functionality. A bit of public interchange would would lead to a mutually agreeable specification and price. In some cases, the software provided might be off-the-shelf and in other cases it might be written from scratch.
As a venue for the commissioning and creation of software, I see OCM providing a valuable venue for both software creators and software consumers. Software consumers could aggregate to share the costs of new or modified software or form a mutual-support community around a given piece of software.
Two wrongs don't make a right, but three lefts do.
Not only did I have this idea, I even wrote it up in a paper for a college class a year ago. It's the same not only in spirit (a commission system for building programs and adding features, the result being open to all), but in the particulars as well (regulation of the market by a for-profit "third party" entity and so on). Not that I'm complaining; rather, I'm glad to see that somebody has the guts to try it.
As for the challenge of getting Linux and OSS on desktops, the standard approach appears to be to try and sell Linux and OSS as a bundle, and to get people to shift over all at once. I personally don't think this is the right approach. The easiest OSS programs to convince people to use are the ones that have Windows as well as Linux versions.
Right now I use Mozilla for browsing, the GIMP for image editing, and free (though perhaps not Free) programs for AVI reencoding and archive compression and retrieval - and I'm a Windows user. If I were able to use Free software for everything I normally do in Windows, and all of those programs also had Linux versions, I'd switch.
However, I'm stuck in Windows because I'm a gamer, and a lot of games don't have Linux versions (or Mac ones for that matter). If the zealots (you know who you are) went after the game designers to port their code, you'd get me on Linux no problem.
Of the several problems F/OSS faces today, one is the fact that it is free. Not as in freedom, but as in free beer. Both private users and business are at a loss when it is explained that there are thousands of people developing Free Software, for free. The immediate reaction is of disbelief, followed by the assumption that the software must be of low quality...consumers and businesses work for money, and they mistrust people who do not.
Except the software is increasingly being used, and recognized as high quality. I just don't see the problem here. If there were not, in fact, thousands of people willing to develop good software for free, then we would have a problem. If the software were, in fact, lower quality because it's free, we would have a problem. Inaccurate perceptions and cultural mismatches are not a problem. They are correcting themselves quickly enough.
In fact, this all sounds like a leftover argument from 1996...these days, the argument is commercial vendors screaming that free software is ruining their business model.
Wasn't there a project recently (OSDN-related), that did this. Someone posts an order, people "bid" to work on the GPLed code, etc... it kinda flopped.
I guess his primary difference is the OCM is the contractor.. "subcontracting" the coders instead of some completely un-regulated bid-market.
meh
Market failure can also occur when people who have not paid cannot be prevented from consuming the good, a condition called nonexcludability. Public goods by definition exhibit both jointness of consumption and nonexcludability.
Ergo, it may be a resource, but it can't be called a market.
Always good.
What are the connotations of this applied on a large scale though?
Blogzine
clifgriffin > blog
Maybe it's just me, but I find this sentence extremely difficult to parse. I think it's a special case for off-duty bookeepers....
I believe that a project like this misses the point. As I see it, most Open Source developers do not write code because they want to Make The World A Better Place. Their motivation is not the satisfaction of users.
Programmers write code because that's what they love to do. This is why there are dozens of editors, mailers, etc, rather than a single well-maintained application of each type. A coder says to himself "I feel like writing yet another mp3 ripper GUI". He proceeds to do so despite the fact that numerous examples already exist because his motivation is his own pleasure, rather than a desire to fill some gap.
Trying create a carrot to dangle in front of these people is pointless. They don't want your carrot. They want to write the code that they want to write.
The article is *not* compelling.
The OCM introduces into the Free/Open Source movement an economic incentive, to help align the priorities of Free/Open Source developers with those of the end users.
The one thing Microsoft has done well is focus on end users. They are able to hock woefully inferior file, database, and mail server solutions because they make sure that Word's macro engine can cook your breakfast for you if you want it to. If users want a feature, it will be in the release product - performance, stability, and security be damned.
I don't necessarily think we should follow MS in this regard. It doesn't seem to have really worked out for them. Let's focus on what is really attracting people to free software - choice.
who are those slashdot people? they swept over like Mongol-Tartars.
TopCoder Software has being doing what this article describes for over a year:
http://software.topcoder.com/
In short they are trying to put a structured, process-oriented community development model into a design / development competition format (with cash incentives, both upfront and as royalties) for creating new software. The resulting work is marketed as a component library, and the community itself is marketed as a "no-shore" development resource.
Check it out, it is a pretty good system -- the results are surprisingly good. I regularly participate in these projects.
+1 Perceptive
The communities are not separated. The high level of communication between developer, tester, and user is the greatest strength of OpenSource software. They are sometimes the same person! By introducing a "clearing house" you are adding a middleman and thereby reducing efficiency.
+1 Perceptive
This is a difficult problem. Really difficult - and essentially unsolved by traditional development models. OpenSource itself is a solution to this problem. Code reviews, while one of the most useful tools for preventing these problems, are also one of the most labor-intensive (read "expensive").
Here is a question I have had about open source code since taking a computer ethics course and also wondering whether we as computer professionals should be certified.
If company A is using an open source product and said product causes a customer of company A to lose millions of dollars whos is liable? Some may say company A is liable which may or may not be true, but lets remove company A now. Many businesses want someone to be "responsible" for the software they use. If that software is written inhouse or by a third party then the liability is obvious. If open source is used would you go to source forge (as an example) and get the user ids working on the project and sue them for liability? You see businesses can't just think about what is the best piece of written code, but also perhaps which piece of code will cover its backside the best. Is that the best situation? I do not know, but it is a situation that exists, why do you think major corporations and even small business owners have lawyers and liability insurance?
The do-it-yourselfers that appreciate the quality and flexibility of open source vs. the people who just want a working computer without configuration woes.
This is the big problem with current Open Source world, IMO -- the current Open Source world seems oriented toward DIY'ers creating code for other DIY'ers. Its a BOF community of computer and software-loving people. The result is code that only another coder would love. What fraction of Open Source software packages "just works" without configuration, reading a manual, etc.
The trick to going mainstream is to create incentives and drivers for writing easy-to-use, "just works" software. That means bringing in people who have an incentive to maximize the number of users (meaning they really want to go after the mainstream desktop market). It means bending the programmer's interests away from cool code and toward ease-of-use. Currently, only companies with copyrights seem to do this with their combination of sales commissions, threat of firing programmers, etc.
So how can OCM help incentivize programmers to make mainstream code, not coder's code? I would argue that by supporting the formation of user syndicates, OCM creates an environment in which programmer compensation is proportional to the number of people wanting the code. If a million people pay $10 to join a syndicate that delivers the next version of some desktop application (e.g., an Office substitute), then the syndicate has real money that can motivate real programmers to create code that satisfies that mainstream.
The only problem? This model forces OCM to move to a restricted-access software model. Why would a million people pay money to join a syndicate if they know that they can download the results for free the day after it comes out?
Two wrongs don't make a right, but three lefts do.
This might come off as a trol, so I will post anonymously... For a long time I have the feeling that all of us, developers/coders, are lemmings. To those who does not know, a lemming is a little mole-like creature that once in a while commits a mass cuicide. Why don't we see any other professional worker , let it be a builder, a doctor, a writer or whats not give his work for free ? Why is that only coders are being asked to code and let their creations go ? for free ?! We are commiting mass cuicide. True, lots of people enjoy linux (as I am) and contribute to it (as I am) - but this does not mean we need to lower the salary standard in our profession to *zero*. The open source advocates will say its about the "service" but hey - if you didn't realized it so far - there is no "service" to give - the end users are asking (natuarally) for everything to be simpler, more user friendly, with help screens and such and we give it to them (see "kmyfirewall" as a great example on how to make the cryptic iptables into human enjoyable expirience for the whole family), the outcome will be that no one will need coders for anything else other than creating programs, friendly programs, for free. So? don't you think there is some truth here? (I know this looks like a trol, but it REALLY not)
With the creation of client syndicates, desired features can be had at a fraction of the cost of paying a software house to do it. Several companies can pledge smaller amounts rather than one company paying the lump sum.
tasks(723) drafts(105) languages(484) examples(29106)
Sounds like he's proposing what I've already done, pubsoft.org. We take tenders for software, whether for entire new pieces of software, or improvements to existing software, and seek to put together developers and users. Multiple funding models are supported: matching grants, whole grants, or the Street Performer's Protocol. We've funded the up2us program (which never made it out because of the difficulty of tracking up2date development), and are currently funding FreeS/WAN, and GNU Radio.
-russ
Don't piss off The Angry Economist
"This is the big problem with current Open Source world, IMO -- the current Open Source world seems oriented toward DIY'ers creating code for other DIY'ers."
You say that as if it were a bad thing. Why do you feel this is a problem? I find it to be a solution.
If the purpose of any program it to be useful, then this is great. If the purpose of any program is to sell a billion units, then I suppose this would be a problem.
For me, it would be a "it-makes-me-feel-good" job. A lot like the job I have now. I don't tutor in a computer lab because there's a lot of money to it (there isn't. $6/hr), but because I enjoy helping people learn about computers, especially the UNIX classes.
Contributing to OSPs has a similar feel to it. I've grown to love Open Source and what it does for me, and would love to be able to regularly make real contributions to it. Getting paid for it would make that possible.
tasks(723) drafts(105) languages(484) examples(29106)
So, ease-of-use is not cool?
I was trying to say it's Munoz, but as you see, slashdot removes the & ntilde ; and places a "n" instead :(
"The result is code that only another coder would love."
What are you saying here? - surely the only people qualified to appreciate code are coders.
"...away from cool code and toward ease-of-use."
There's no link between the quality of the code and how easy to use it is. An easy to use interface is not indicative of the quality of the code and vice versa.
"...make mainstream code, not coder's code?"
If by "coder's code" you mean good elegant code that's not cranked out to meet commercially driven deadlines, I'll take that anytime.
I suppose an important question in an organization like this is: Who's copyright goes in the license? Does the OCM get put in, or does the developer who did the actual work?
tasks(723) drafts(105) languages(484) examples(29106)
This initiative is doomed to fail if they do not have very strong policies on IP. The GPL provides no guarantees on IP; for example, I could implement a patented algorithm and release it and users could get sued.
The article does not even mention the words "intellectual property." Something to consider!
"This is the big problem with current Open Source world, IMO -- the current Open Source world seems oriented toward DIY'ers creating code for other DIY'ers." You say that as if it were a bad thing. Why do you feel this is a problem? I find it to be a solution.
Your point is excellent! Different types of computer users want different types of software. Code for coders is beautiful.
I agree with you that "useful" is important and that software does not have "to sell a billion units". But I would argue that coders create more value if their code supports the needs of a wider audience. The total value of a given bit of code would seem to be the product of its functionality ("usefulness") and its marketshare ("units"). Simplistic, lite versions of software may ridiculed by real computer professinals, but they seem to provide great value to plain old computer users.
What I do worry about is the long-term marginalization of the Open Software community because of its code-for-coders emphasis. I can easily see Microsoft creating barriers to open software in the guise of security -- requiring software publishers to license expensive DRM, trusted-software technologies. Without the financial resources to buy Microsoft-provided security codes, Open Source would die because Windows would prohibit its installation or reject interfacing with "untrusted" open software. And without a billion units of Open Software out there, too few people will fight for its survival.
Two wrongs don't make a right, but three lefts do.
I'd like to ask the anti-corporate nitwits who played right into corporate cost-cutting hands, what's next? You've successfully transferred dulled-edge computer technology into the public domain and made it 'good enough' that all your jobs are going to be done in India and China. You've increased the need for accountants and managers; how nice. So what is the next step? What are pseudo-engineers supposed to move on to?
I'm not sure about the overall fraction of open source, but I think that the majority of the most "popular and common" open source apps "just work." Here's some examples from my personal experience based upon what I use.
Everyday apps:
KDE - just works. Plenty of tweaking can be accomplished by right clicking or through the control center.
Mozilla - just works.
Evolution - just works.
OpenOffice.org - just works.
When you lose something irreplaceable, you don't mourn for the thing you lost, you mourn for yourself. - Harpo Marx
This is the big problem with current Open Source world, IMO -- the current Open Source world seems oriented toward DIY'ers creating code for other DIY'ers. Its a BOF community of computer and software-loving people. The result is code that only another coder would love. What fraction of Open Source software packages "just works" without configuration, reading a manual, etc. Don't confuse Open Source with the Linux community -- the two biggest Windows compatible Open Source projects, Mozilla and Open Office, require no more configuration than their MS counterparts. And with Mozilla programs, you can do some heavy duty customization by going to Mozdev and installing extensions directly from the web -- so easy, even an AOL monkey could do it.
Les Miserables Volume 1 now up with my reading of
Okay, first off I am a technology consultant that focuses on helping SMB's (I guess SME's is the new buzzword of choice to describe small businesses) intergrate opensource software into their armada of software. Sometimes its as easy as a package like phpESP, other times its Linux on the file server, and a couple times, top down use of Linux from servers to desktops.
Linux will not be making it onto many desktops down here anytime soon. Why? RH's ending support and SuSE's future is at best wait and see. SMB's want to have a number to call for support and know that there are experts out there besides myself. With no corperate support for the desktop, they will choose Windows over Linux everytime. Why? Well all the software they often needs is written for Windows and they don't mind to pay so long as it works.
As you said, you would expect the market to increase as OSS moves into the mass market. Without the commerical support, its not going to happend. RH's move last week has impeeded adoptation of Linux by SMB's on the desktop by several years. Our company had been making a large push towards the platform, and the moves last week has undermined our credablity. SMB's want something that is supported and not "here today, gone tomorrow" which is what, in their eyes, has happened.
OS X from Apple has been the real killer of Linux on the desktop. A great many of us that switched from Windows to Linux made the switch to Apple. Why? We have our native *iux development enviroment, access to a great many of the OSS applications we liked, AND have the commerical support from vendors like Adobe and Macromedia.
"The problem with socialism is eventually you run out of other people's money" - Thatcher.
However, I have to wonder...what will existing maintainers of a given project do when a company pays someone else to add a feature to their application? I suppose it would depend on the maintainer, but would there be a definate trend?
... The original developer will lose control of the future direction of their creation. Not that this can't happen now, it just doesn't that often. But this new fiscal dynamic could changes things. And since it is open source there is nothing you can do to stop it either.
Bingo, and that is the flaw. There is no way you can compete on cost with developers in Russia, India, China,
This Free Market/Software Exchange idea is basically a global version of a eBay that streamlines the off shoring of software development. My feelings are mixed. Is this really a good thing?
The article presents an intriguing idea, and it seems to have a lot going for it. But let's consider the individual or small business owner who needs specialized software. The article mentions dentists and architects as examples.
For example:
If a dentist needs has an itch that can be handled by some completely new software, he or she can pay some OCM to commission a developer to create the software. The OCM tells the dentist the project will cost $8000. Should the dentist pay the $8000 knowing the program will become OSS and all the other dentists can get it for free, or at a severely reduced cost? Or should the dentist wait and see if one of his/her competitors pays for a similar program first and then get that program for free?
On the other hand, the maker of proprietary software can charge the first dentist and the hundredth dentist the same fee. More importantly, this fee can be far less than the $8000 requested by the OCM. A typical dentist is much more willing to pay $100 for proprietary software.
Economists talk about how the "First Mover Advantage" gives a huge boost to the first person that moves into a market or adapts a new economic efficiency. The OCM becomes a "First Mover Disadvantage". If the OCM presents a huge disadvantage to the person who pays to make F/OSS, then nobody will pay for it. Result: Fewer F/OSS projects and more proprietary software.
Disclaimer: Philosophically, I prefer F/OSS over proprietary software. Perhaps the OCM can overcome the "First Mover Disadvantage". I hope so.
I think the idea in itself is a wonderful one, and it's something that I thought about years ago as a solution to the (then) problem of finding good developers and reducing costs. But I see a problem and that is from commercial developers who now make expensive software that is well liked by by end users. Think of Adobe's or Macromedia's products or even products such as games.
Let us assume that company X wants to make a product, let's call it PhotoX, which it would like to sell in competition to Photoshop. If the development is all done under the GPL (as opposed to say, the LGPL or BSD licences) how could the company be in a position to make a profit from the product, since the GPL requires that sources be made available?
I see the problem remaining that there would be no one willing to invest in such models for enduser software such as the above mentioned. Given that such companies are very profitable today, I find it hard to see an opensource alternative today unless the BSD for example is also used.
An alternative idea is to introduce a community currency: (this was presented in the Future of Money summit)
Non-free (low cost) software that is entirely open source, where developers can submit changes to the source code. If the changes are accepted, the company pays the developer some bounty based on the amount of work.
The company should pay fairly, because that will encourage others to submit code changes to make the program better.
Is anyone doing this? Does this idea sound appealing, or am I smoking my socks?
I can't believe that the First Monday peer review process accepted that article without pointing out another piece they published a couple of years ago, entitled The Wall Street Performer Protocol: Using Software Completion Bonds to Fund Open Source Software Development.
There are also some additional references to related market-forms here. Google Answers also works on a somewhat related principle.
Fixing copyright
why a customer would pay a developer to write OSS code when he can pay for proprietary code and not help out his competitors.
I've tried for the last week to successfully enable the licence on a piece of shareware, sending emails to the author almost every day, and getting another .dll, another licence code etc. from the author almost every day. I've installed and uninstalled the software on two seperate machines each time I get an update. The original demo worked, but once licencing software has been thrown into the mix then software will no longer do anything but pop up error dialogue boxes. The software is niche scientific software, and nothing else resembles it - so there are no options other than to get the thing to work. The price is trivial when compared to the wage cost of trying to get the thing to work.
If the author could get money without using the licencing software, then there wouldn't be the problem of a working demo turning into nothing but space on a disk. If the source code was available I may have been able to work out what my machine requires to get the software to work - if it isn't just the licence manager software that is the problem.
Your $10 could get you something other than the final software. It could get you:
Basically, you could pay a bit of money to have some control over the development process.
Why haven't gnome set up a way to add money to each item?
I wan't to pay over Pay Pal or some other means to encurrage the things _I_ (and my company) want's.