Quelle: http://tim.oreilly.com/articles/paradigmshift_0504.html Date: 2007-01-30 Comments to the text Tim O'Reilly: Open Source Paradigm Changes ========================================================== * GIS user input 2004-09-29 08:21:32 tallbill ========================================================== The problem with GIS user input seems to me to be that the user is a person who may not want to put input into the system. For example a person who is a hiker might consider writing a book about his favorites spots in the mountains. But then after publishing the book he regrets revealing his secrets to the public because he now finds that his quite spots become overrun with new hikers. Why? Because he told them about the places in the book. The same would be true about GIS for motor traffic. Publishing the data on what routes are faster to get from hither to tither results in those faster routes being advertised. People who need to go between the two places now know a 'faster way'. However if they all try to use that information at the same time, then there is suddenly a traffic jam. So when we have a short cut and we tell everyone to use it, then it might not be a short cut anymore. Also, the idea of having user input involves tracking a user which may violate his privacy. When I was working with some construction crews I thought about this a lot. I realized that these guys did not want to have their lives tracked as they didn't always do what they said they did or go where they said they were. They didn't lead 9-5 lives, and no one needed to know. They got their jobs done or they didn't get paid. And yet having a GPS and transmitter in the truck to allow for an easy truck inventory for the fleet was very desirable. If we take tracking of motion to an extreame, one would could set up tracking that would be very percise, but then we would have a flood of data. If we can track a persons finger movements, we can track their keystrokes. We can track their handwriting. Fortunately the bandwidth to do this for everyone is huge. The ideas seems perposterous now. However, to do it for a small group perhaps wouldn't be impossible. I believe that On-Star must have a plan to do something with all of the data that they collect. But if they do they could be compromising the privacy of the car owners. Obviously this is an area where caution and vigilance are needed to assure that new tyrranies do not arrise from emerging technologies. I am not worried because if people feel they are violated, then they will disable the tracking. However, as this data is so useful for insurance accident investigations, and police criminal investigations, there might be a time when this kind of tracking is forced into acceptance. The British in London are using GPS to charge people to drive into the city. At the point when police use GPS to blanket issuance of speeding tickets there will be public outcry and this issue then will become of major importance. It hasn't happened yet but given the huge number of police organizations it seems like just a small time before some one some where tries to do this. The question of owner ship of tracking data comes up. When actors are strapped to sensors and their motions control an computer generated animation they are considered to be acting for the camera. Do they then own the performance? If a person is driving his secret short cut in his car and giving this data (unknowingly perhaps) to OnStar, does OnStar then have the right to publish this and sell it to others? I would think that they should not have the right. I would hope that law allows that a persons motions and activities are his 'performance' and that publishing them would have to meet with approval of the performer. This is a lot to think about. But clearly there are some new issues here that need to be addressed. * Personal rant 2004-07-08 10:08:36 Although the Tims position is not new but he presents it in a very interesting way. Perhaps the most interesting aspect about collaboration is giving enough people the ability to participate. Making the physical communications easy is only a part of the story, making a common language (I do not mean simply the spoken language e.g. English) but the subjects sub-language, for example in a Electronics discussion forum you have to be familiar with the terms and concepts of Electronics. Herein lies the the fundamental problem that I see - the successive sub-specialization a forking of all disciplines. The general participation that we see on discussion boards I will take Linux kernel development as an example - an arbitrary choice. The first observation to make is that the cost on entry is that you have to be a C programmer, and CS degree definitely helps. I am sure it was relatively easy to be a participant in Linux kernel development 10 years ago, but as the components have become ever more complex the fewer percentage of potential contributors are able to participate. The whole point of this is to also suggest that the other paradigm shift we need with open source is to put particular emphasis on simplifying principles that again open up the field for greater participation. Perhaps it is time for the Kernel to take Andrew Tannenbaum's early criticism of the kernel more seriously. Again the purpose of it is to suggest that simplifying and separating out components will allow greater collaboration by enabling a greater number of people to participate. Perhaps even more than the kernel, projects like XFree86 show the danger of open source with open collaboration, despite the many advantages of the X model, the graphical interface on Linux/Unix has faller far behind OS X and even Windows XP. The beauty of the Unix model is that encourages loosely coupled cooperative systems, but even those cooperating systems, in time, become large, monolithic, and ossified. Ossification is the fate that any system suffers once it fails to easily allow new participants to to be easily introduced to development process. ========================================================== * We are in the middle of a paradigm shift 2004-06-29 18:22:02 webetize ========================================================== We are in the middle of a paradigm shift I will present several short topics that cover major issues from 10 miles high. Short background to the agreement SUN provided several major stepping stones in the march forward within the computer science/industry Microsoft saw this and had to backtrack to get out of what had become (becoming) suns world, thus the agreement. Otherwise Microsoft would be tied up in legal battles for the next 15 years and needed to be free. Given this Microsoft could work with SUN and gain some semblance of control. I The Universal GUI The whole area and major industry discussions about thin and thick clients is a tempest in a tea pot. The short answer is: Given the ?right technology? everyone can chose the optimum balance between thin and thick clients. Why don?t we have the ?right technology? Short answer The one who controls this technology controls the ?Computer1? (see foot note for background lead). Worst case scenario (agreement does not work) Microsoft takes over the Web GUI by dumping JavaScript, replacing it with it?s own ?browser? scripting language. AND MORE? Related sub topics, Universal GUI, 3270 al la 2005, Web GUI==All GUIs!!! II The Universal API The whole area and major industry discussions about Web Services, REST/SOAP, HTTP/TCP etc all boil down to the adoption of the new UNIVERSAL API. We are not there yet. GRID technology, the all forms of distributed processing all are an aspect of the ?Computer? as described in footnote one. All these API discussions boil down to a set of solutions adjusting to the ?connection speed? whether it is called a bus or the Web. Microsoft had developed the best API ever OLE, COM, DCOM all had solved the issues of Component Composition at least with in the PC industry. Then Java comes along and offers an alternate solution. OLE/COM/DCOM are broken in one swoop. See the Small Talk language which combined messaging in a language. Major refactoring is triggered as a response to the new choices offered in the new API. Worst case scenario. Microsoft can dominate thus replacing OLE/COM/DCOM with another proprietary API. III XML becomes the new ASCII Text becomes king. The collapse of the old AI paradigm, NLP and SGML at the same time made text king. HTML beat the pants off SGML, SGML comes back with XML to save their butts. XML becomes the ASCII of the new century, thus making the need for NPL more distant. IBM clung to IBCDIC for years, XML became king over night IV The Universal Computer Language JAVA as a language is not important, Java as a ?Web enabler? is very important. Java forced the computer industry to see what was obvious in other circles. The process Code-Compile-Link was obsolete. There is no mystery here; this development was driven by the tremendous increase in CPU speed. The ?Libraries? and ?API? made famous by Apple and Java were the real paradigm shift. The libraries were the glue that made webetizing a reality. As the level of complexity increased these libraries became the new ?level? of development. The holy grail of ?Component Software? while not reachable through the API, could be approached by supplying large libraries. Microsoft sees this and betters JAVA with NET. Notice that they did not come with a language but NET, which was the real strength (Web) masked as a language V Universal machine code JAVA as a language is not important, Java as a ?machine code? is very important. Again driven by the tremendous increase in CPU speed, machine code is replaced by a virtual machine. Microsoft sees this and betters JAVA with NET. VI The Universal Code Library The tremendous growth of Open Source and industry standards is not an accident; it is driven by the need for a universal Code Library. Open source replaces Apple?s, Microsoft?s, Sun?s, everyone?s proprietary Code Libraries. The bottom line is that learning these libraries had become a necessity and learning so many proprietary ?libraries? was impossible. Thus open source filled this need. There are many other aspects of open source that I am not addressing here, but this need for a universal Code library was the driving force. Summing these points 1. The Universal GUI 2. The Universal API 3. XML becomes the new ASCII 4. The Universal Computer Language 5. Universal machine code 6. The Universal Code Library 1. XML becomes the ASCII 2. Code-Compile-Link becomes obsolete 3. Ubiquitous language (universal virtual machine) Many of the buzz words of today are acronyms for the present candidates for the universal GUI, API, ASCII, Language, Machine Code and library. As the industry battles to control these changes the users and developers struggle to keep up. Synergism, the new paradigm These six forces all combine in a major paradigm shift that is manifested in the rapid changes we see. The power of these six Universals in the form of libraries, APIs, Language XML, etc is what the Cooperation Agreement is an attempt to handle. More to come....... Keith Elkin Back ground: The Web is the computer Finally the message is getting through ? the Web is becoming a single, shared infrastructure for collaborative computing, and that changes the nature of software for ever. ============================================================ * So who makes the money? 2004-06-28 22:38:26 rawbeertoe ========================================================== Are you saying that IBM and other major IT consulting service providers will benefit from the changing of the once proprietary software market? ========================================================== * "killer apps" 2004-06-28 20:29:35 infoatneo ========================================================== From your article... >Most of the "killer apps" of the Internet, applications used by hundreds of >millions of people, run on Linux or FreeBSD. These "killer apps" are just web pages on the client side. We are still running dumb terminals but now they have pretty graphics. Client side, innovation is stuck in low gear in order to work across multiple platforms. The experience has become a lot like Television but with more channels. Why did this happen? Microsoft controls the onramp for 90% the world's Internet users. If anyone else innovates on their desktop they are purchased or crushed. Their web strategy is obvious: make sure every copy of Windows includes a Microsoft browser that encourages incompatibility while pretending to support open standards. I don't see this situation changing until main stream hardware vendors begin to offer serious desktops and notebooks with Linux. Most of them have a symbiotic relationship with Microsoft and are openly hostile. HP for example told me that installing Linux on my laptop will void the warantee. I want to walk into Circuit City or Best Buy and see high powered desktops and laptops with Linux installed. I'm not talking about anemic excuses at Wall Mart with minimal RAM. When this happens, innovation can begin again and some killer apps can be created on the client side. ========================================================== * How Open Source changes the balance of power 2004-06-28 18:30:30 sergiom ========================================================== *About Software Commodification*: Whilst Open Source has significantly increased the quantity and quality of available building blocks, software development process itself is not undergoing more commoditization now than 10 years ago. A programmer nowadays has more a less the same tools and uses more a less the same methodologies available by 1990. Still one key diffence between software and other technologies is that if your have a couple of television channels and you merge them, the result is a single television channel. But if you take a couple of software applications and merge them the result is something that simply doesn't work unless you almost entirely rewrite one of the applications. Another key problem of software development is that still the best way of developing a program is inside one person's head, or at least inside the fewest people heads. That is why best core development teams always try to keep themselves small. *THERE ARE A LOT MORE OPEN SOURCE COMPONENTS NOW, BUT WE STILL DEVELOP SOFTWARE MUCH THE SAME AS WE DID 10 OR 20 YEARS AGO.* *About Creative Destruction*: "Creative Destruction" is all right. What it is not so all right is "complete market destruction". The key point here is not how Open Source changes a market such as Windows vs. Linux but how Open Source changes the whole production and benefits share. Let me put an example: Ten years ago being an Oracle DBA was a real value. Oracle was so expensive (at least in my country) that even private universities could not afford buying it for educational purposes. As a result, students came out of the university with no idea of Oracle, and since Oracle was so common among big enterprises these people had to pay for training and then got higher salaries for their unique knowlegde. Today anybody anywhere around the world can download MySQL or PostgreSQL and learn everything about them (much more they could learn about Oracle). It is a matter of time that an Open Source database reach the performance and reliability level of a good commercial database. It doesn't matter whether Oracle or PostgreSQL or anybody else wins. What does matter is that Open Source is moving the power of innovation from more developed countries to less developed countries and this is extremely important because the whole technology wealth of the last 200 years from fireguns to microchips has been based on selling knowledge and patented materials to people who just could not do it themselves. This is happening only in the software industry and not is the hardware industry which is everyday more and more concentrated in a few points. One reasonable argument against extreme Open Source is that we'd better think twice where is our money going to come from before giving away our competive advantages. Change and paradigm shifts are all right, bloody revolutions are not. *OPEN SOURCE VS. PROPIETARY IS NOT JUST A MARKET WAR BUT A WORLDWIDE REDISTRIBUTION OF KNOWLEDGE AND POWER.* *About building custom distributions*: A custom distribution is very expensive to mantain. Maybe you can do a fork of some well know code and sell it to government agency for a very specific purpose. But basing a business in custom distribution is suicide because there are so few people with enought money to pay for it. Customizing software is not like painting a car in different colors. *ONE TO ONE CUSTOMIZATION MARKETING IS STILL NOT FEASIBLE BECAUSE SINGLE CUSTOM SMALL PIECES OF SOFTWARE ARE TOO EXPENSIVE TO PRODUCE.* *About building a strong brand*: Building a strong brand with nothing behind it is one of the best ways of burning cash. The business history is full of example of companies that spent their money in branding because thay did not know what else to advertise. *BRANDING IS LIKE COMMON SENSE: YOU SAY SOMETHING IS COMMON SENSE WHEN YOU CANNOT DEMOSTRATE IT.* *About Network Collaboration and the Bazaar Model*: Collaborative Software development is quantum leap in the way software is build, but it is not the ultimate solution to every problem. One basic thing to understand in collaborative software development is that people develop things they need for themselves and then put their work into public domain when they cannot make money out of it (this is how Stallman began writting his own printer drivers). As a consecuence it is difficult to control the output of a collaborative process. And if there is something that customers dislike that is uncertainity. Another barrier for Collaborative Software development is that we still lack of network enabled tools powerfull enougth for supporting it. Publishing your work at SourceForge and doing some bug tracking is one thing, coordinating a hundred people around the world writting ten thousand features is another one much more complicated. *COLLABORATIVE SOFTWARE DEVELOPMENT IS A PARADIGM SHIFT BUT NOT THE ULTIMATE SOLUTION TO SOFTWARE BUILDING.* *Heroes of code fork*: I really love the short story about Mark Anders and Scott Guthrie ASP.NET prototype. It demostrates that it is not just an idea but the energy you put on it what makes it work. *A SINGLE VISION CAN CHANGE THE WORLD IF YOU PERSIST ENOUGHT ON IT.* ========================================================== * Open Source Paradigm Shift politics and the Something(tm) system solution. 2004-06-28 14:20:54 ShannonBailey ========================================================== Thank you Tim for delving into the heart of the matter in your article The Open Source Paradigm Shift. I am enjoying the user comments and they have stimulated me enough to write about these issues and describe a huge project to address the issues brought up in this article. I enjoyed the history and perspective you shared which add to my experience in the computer industry for the last 20 years. To continue the thoughts and questions asked, I'd like to discuss the statement: 'we can build an operating system that is designed from the ground up as "small pieces loosely joined", with an architecture that makes it easy for anyone to participate in building the value of the system'. When I read this I felt torn between agreement and disagreement. I agree that a new paradigm is built upon much of what currently exists. But like the difference between a Monkey and a Human, both use neurons, but the organization of them differs, or should I say has evolved to produces one organism whose structure is capable of supporting a vast social structure and language far beyond the other. If one looks at computers, operating systems and applications using the architectures of today as the mind set, this previous statement seems false, for it fails to acknowledge that a new system which creates a new paradigm must be a heretofore unique collection of pieces organized in a new ways to form a whole greater than the sum of its parts. A key reason that this new whole has not been formed yet is that it cannot be owned and possessed by a "Corporation", for the aims and goals of this whole are incompatible with their goals of corporate Ownership, Individuality and Complexity. I'd like to contrast these three concepts with the three concepts promoted in the article: Software as a Commodity, Network enabled Collaboration and Customizability and Software as a Service. If those in power allow the 'Ownership' of 'Software as a Commodity' to become the property of global humanity collectively, then they would no longer be able to control and extract money from that which has become a public good. This runs counter to their interest and therefore will be discouraged by the architecture of the systems they produce. Because commercial interests have som many people involved in their production, a brain drain occurs and the models of computer use and interaction are simply built upon instead of rethought from ground up periodically. Linux while a great OS, suffers from being Unix architecturally in mind set instead of something far beyond. There is no benefit to those in power with providing us the means of acting as groups and working to help us break the barriers of our 'Individuality' of use of computers, by facilitating 'Networking enabled Collaboration'. To do so creates a system where the value of what is being produced through the group interaction is the value and not the system itself. Those in power would no longer be able to neither gain profits from the interaction nor own the content of the interaction. So integrated collaborative end user tools, which are 10-100x more difficult to create, are not deemed a priority feature. Finally, why would a corporation wish to make their system as minimally complex as possible yet very powerful, easily documented and learned and designed to be extended and made 'Customizable and Software as a Service'. To do so would allow others to copy it and facilitate many people evolving it without the original owner gaining further monetary value from those improvements nor maintain control of what is created. Purposeful 'Compexity' is a key defensive strategy used to maintain ownership of most software today. But in reality, most complexity can be a hidden part of a system without burdening users and programmers with it directly. But if the way to use and make applications for a computer became simple, cohesive and symmetric, the applications would be portable to clones of that computer and operating system, and thus ownership would again return to humanity as a global collective. And so the politics of control and a lack of a cohesive plan for an alternative system, hobbles current computer, network and software architecture. For these are not designed with humanities benefit in mind, they are designed so that a small circle of people can control and own them. My response continues with a description of the Something(tm) (code name) operating environment project description at: http://www.shamanicvision.org/simunity/paradigm.html Thanks, shannon (at) simmunity (dot) com ========================================================== * Paradigm Shift 2004-06-28 10:34:58 jmcneely ========================================================== Excellent stuff, and this is something I have been thinking about for some time. I'm also a big Thomas Kuhn fan, kudos for bringing that into this. We are in the middle of a shift, not simply open source but information technology. This is the kind of thinking that is going to move us forward. I think the comments to the article, esp on slashdot, are showing that a lot of open source guys don't get this, to them it is a MS vs. Linux thing. I have been striving to implement user control over applications in my own solutions. It is also a key insight that content generated organically through normal usage actually IS the heart of the app. for me this is actually a selfish thing, because I want the users to create the reports and such that they need without my intervention, and this has proven successful where I have put it in place for people. I think that the focus needs to be expanded from the SOURCE being open, to the entire SYSTEM being open. The system includes the contribution, the innovation and knowledge of every stakeholder. To the extent that they allow users to contribute material, that is the genius of Amazon and slashdot, and it is why they are successful. Some stakeholders are programmers, some have a knack for interface design, some are data entry people, some are data overseers, and need complex overviews of large segments of data. If the system frustrates or ignores the talents and interests of ANY of these users, it fails. Thus a software system is not comprised of the network, nor the app, but the PEOPLE who use these things. The system should simply provide convenient tools to allow people to do their thing, to follow their genius, tools that THEY quickly understand and use intuitively. Mere open SOURCE is closed to many of the stakeholders in a system, simply because they do not perhaps know how to use regular expressions or write a multi-join SQL request or whatever. We need open SYSTEMS, in which the people who need certain things from the system can create the interface to do so. Then the question becomes, how can you make a living off of such a thing. Are we even taking the expertise of programmers, as well as their ability to create the software, off of the table as a means of income? Not at all, the source of income needs to be where it should be, at the point when someone is obtaining value from the system at the expense of someone else's time or expertise. Time and expertise are not only in the hands of programmers. It is, after all, a computer system; computers excel at maintaining lists of times and usage characteristics and users. Google's ads are a great first model of what is possible with sensible tracking of usage-based revenue gathering. This could be so much more sophisticated. However, I envision networked or web apps in which the interface itself is easily created and manipulated by a normal (non-programmer) user, gathering and contributing to data hosts as if every source were part of a giant relational database, using tools perhaps to manipulate xml/xslt without needing to know how to create style sheets from scratch. ========================================================== * Open System and Open Source 2004-06-28 10:14:08 Argent ========================================================== What I've been saying for many years is that Open Source is important because of Open Systems. Which seems to be another spin on the same subject: it's the interfaces and protocols that are important, and to keep them open you need competition to keep a single vendor from controlling the interface. Open Source does that better than closed source, because it's impossible to get rid of the competition. Even if there's only one implementation, Open Source software inevitably competes with itself so even the author of that system can't close their own interfaces... if one really tried to force people to use a closed interface against their will, the community would just fork. Open Source promotes Open Systems, which is why Open Source is important. Open Systems aren't a new paradigm, though, and the way Open Source keeps systems open... that's really a change in emphasis more than a paradigm shift. ========================================================== * Stigmergy is economically efficient discovery 2004-06-28 09:37:30 mengwong ========================================================== The economic perspective you and others take is absolutely appropriate. Economically speaking, at the highest levels, the product of human effort in a free market is value. The actions of a market can be considered a hill-climbing exercise in discovering the combination of factors that provide the greatest value to the greatest number. Money has historically been tightly coupled to production. In the information age we have begun to see a decoupling. All the fascinating changes you talk about look to me like a product of that decoupling. The key human resource in the information age is enthusiastic attention. Money is useful as a quantifier of value because it affords measurement, feedback, planning, and improvement. Money is also useful as an enthusiasm generator because it directs human attention toward problems whose solutions create value. But on the Internet, which is the most frictionless medium yet invented for the exchange of ideas, ideas themselves are an attractor of enthusiastic attention. Those ideas can compete with money as an attractor. In other words, folks who love opensource programming do it even when they're not being paid for it. And, in general, ideas trump money: money follows ideas; ideas do not follow money. In fact, on the Internet, a monetized model of production in which coders work for hire can be less efficient than a non-monetized model in which coders work on whatever interests them! Why? The fundamental question in a market is "is this work valuable?" In traditional markets, we answer that question using money as a measuring stick. The answer is yes if the project attracts customers and profits are obtained. The answer if no if the project does not sell and the producers make a loss. But on the Internet, you can take money out of the equation and still obtain the same feedback dynamic. A bad idea will not attract coders. A good idea will. And in this pure economy of ideas, you don't have the overhead of payroll, timesheets, sales, focus groups, and contracts. All that stuff is friction. Necessary in a money economy, unnecessary in an idea economy supported by the Internet. That's why solutions developed in the opensource way tend to actually be better than solutions developed in the commercial way. In commerce, because enthusiastic attention has to be bought, money is friction that slows down feedback loops, and competition between ideas gets caught up in competition between companies. In the opensource world, competition between ideas can happen much faster --- evolution is quicker because there's less friction. Because the drosophila fly can reproduce in a day, it is more adaptive to a changing environment than, say, a redwood. Coming back to the economic model, we see that open models can be a better, more efficient way for a market or a society to quickly discover good ideas and good designs. When you roll market research and r&d into a single process, product cycles no longer interrupt the game. The action keeps going continuously. And that's why technologies developed in the open are better than technologies developed by companies. Now, that doesn't always mean that all technologies deserve to be developed in the open. Technologies can only take advantage of the above argument when the field is fertile. It takes a rare set of circumstances for a set of actors to overcome the tragedy of the commons in the absence of government. Collective action only works when each participant can benefit by participation. And there have to be a lot of candidates working in a field before a critical mass can bootstrap into existence. The opensource development model worked with SPF because everybody wants to solve spam. The Apache web server worked because everybody wants to publish content for free. Linux worked because everybody wants Unix on the cheap. RT works because everybody needs a ticket system. But there are lots of things that haven't worked as well: where is the best-of-breed opensource account mangement and billing system? Where's the opensource answer to Adobe Photoshop and Illustrator? (Sorry, the Gimp still doesn't come close.) I'm sure they will come one day: we just need to wait for critical mass to build. For some technologies, that critical mass may take so long to build that by the time it's there, the problem has already been solved and commoditized in the old way of commercial production. That's fine too. ========================================================== * Paradigm Shift? With links 2004-06-28 06:50:51 loca ========================================================== I am not sure that open Source per se is at the level of a paradigm shift. Indeed swapping source is as old as computer programming (see Moody's Rebel Code and the FSF site for example). And regarding theorising the age of service delivery, well again this is much more interestingly discussed in The Age of Access: The New Culture of Hypercapitalism, Where All of Life Is a Paid-For Experience by Jeremy Rifkin. I think that the social ontology of open source is perhaps the possible point for a new paradigmatic moment. But at the moment everybody is so caught up in technical fetishism that I think it will take a while for us to notice. The reconfiguration of the political as a network and the resultant new epistomologies and ideologies will be the new battleground for social and political action. See for example Libre Manifesto ========================================================== o Paradigm Shift? With links 2004-06-28 14:48:21 timoreilly ========================================================== I think you miss my point. My argument is that we need to look through the obvious elements of open source such as licenses and code sharing, and notice that open source and open standards are leading to the commoditization of many types of software -- which in turn drives value in unexpected ways. Unexpected, at least, to people who have grown up in the PC era, where direct payment for software is a paradigm shared by both those who want to profit from selling software and those who eschew it. Meanwhile, new paradigm players are making a bundle using entirely new business models. ========================================================== * Not that original and is it really at the magnitude of a Paragigm Shift? 2004-06-28 06:48:43 loca ========================================================== I am not sure that open Source per se is at the level of a paradigm shift. Indeed swapping source is as old as computer programming (see Moody's Rebel Code and the FSF site for example). And regarding theorising the age of service delivery, well again this is much more interestingly discussed in "The Age of Access: The New Culture of Hypercapitalism, Where All of Life Is a Paid-For Experience" by Jeremy Rifkin. I think that the social ontology of open source is perhaps the possible point for a new paradigmatic moment. But at the moment everybody is so caught up in technical fetishism that I think it will take a while for us to notice. The reconfiguration of the political as a network and the resultant new epistomologies and ideologies will be the new battleground for social and political action. See for example http://www.libresociety.org/Libremanifesto1-61.pdf ========================================================== * paradigmn shift 2004-06-27 10:39:41 packrat2 ========================================================== structure, function and dynamics. Neat, this functional shift to downstream networking (money), but you didn't go far enough. The function of the net (the app is the net (sun, etc) stops there; in a blaze of self-serving hype. will google burn itself out with bad news, mistakes or get censored to death? Will the lemming effect kill the net? I see my box as a work-station, a net-center and an entertainment device. who ever makes senses outta the 6 million hits and DELIVERS content will be the 3'd gen, betcha. AI can already replace judges,etc, so it'll be a massive crash+burn to get there. With built-in lie-detectors, probably. pat ========================================================== * What is this about 2004-06-27 01:43:55 musnat ========================================================== I am sorry, I have great respect for O'reilly, but what's the point of this article exactly? I see the same things I have seen so far. - Microsoft will die. - Linux will win But people have been saying this over 2-3 years now. When is this going to happen exactly? How many people really go with open source now? When I look at the sourceforge and other open source projects, the good projects seem to center on very few specific fields. You don't find too many developers that really put a lot of energy to develop real world challenging solutions out there. The fact that Mysql has dual license means that, there will be almost always someone who doesn't want to show the code. If suddenly everybody decides to embrace Open Source, if that was really the trend, then how come Mysql plans to make money exactly? It is the same with some other companies. Netscape embraced open source, it was closed down. Mozilla is out there, it is good, but the future is not clear, so many developers are not working on it anymore. Some of the core developers are working in other companies, they are not developing mozilla anymore. OpenOffice is not a good software, it is somewhat good enough but office is clearly far superior to openoffice. Why would someone use openoffice if they can easily afford ms office. You clearly have a higher productivity by using ms office. Also, Sun is supporting OpenOffice now, but it is not their core business, we just don't know how long Sun will pump money to OpenOffice. Commodozation of software assumes that there won't be much innovation anymore in software market. I disagree. There are so many more things that can be done and should be done. That means you need massive developer power, and other than Linux kernel, I haven't seen a project that can change the direction of a project in a significant way. Every project I have seen is depending on the old code, and just tries to improve upon it. When you have to start from scratch, you need funding and lots of developers. None of the open source projects seem to have that power. ========================================================== * Open Source Software 2004-06-25 11:17:00 Turing ========================================================== The problem with OSs like Linux is "which Linux?" do I use and who can use the software I develop with my version of Linux? How much do my expected users know about drivers and interoperability? Who is my audience and am I serving them or myself? A decade or so ago, IBM started doing something interesting and unheard of with its JES/3 job entry system. IBM increasingly converted open code to proprietary code because so many people jerked around with and tweaked and perverted the open code that new releases required several programmers six months to a year to implement simple upgrades. (They also wanted to force people off JES/3 so that they could retire it.) Open source code is obviously good for some things (like cudtomization) and a terrible idea for other things (like standardization). And geeks should not contribute to, but not control the debate. ========================================================== * Open Content/Processes 2004-06-22 22:49:55 enkerli ========================================================== As usual, thought-provoking ideas on a wide range of issues centered on the future of the open-source philosophy behind much of contemporary computing. One thought. Some of the killer apps are based on open content created by users. Users reviews for Amazon or IMDb, links for Google, item descriptions for eBay, etc. These are very different from the open-source tools and processes that we commonly associate with computing. Though this difference does show up in this and other texts, its implications might mean a broader influence of the "open-source" philosophy on contemporary society. Of course, with "content," there are obvious issues of intellectual property and ownership, cost and value. These are perhaps unavoidable. But just thinking of the amount of content most of us generate in a year and multiplying that by the number of users, there's more to Internet-enabled collaboration than sharing patches for software bugs. Blogs have come under a lot of radars recently. While an adequate example for a type of openness, they're almost as author-driven as traditional publishing, which might be a reason why media analysts like them. But when one thinks of not only Slashdot comments and Wikis but also all the public content posted on mailing-lists and Usenet newsgroups, one's attitude toward knowledge as wealth may change radically. Usenet might have been a convenient way to transfer patches but it was still a way to distribute actual content, including news. Much of it was a free-for-all, with a fairly low signal-to-noise ratio, but many of these newsgroups gave birth to important FAQs which, in turn, have spawned into a lot of the pieces of content users look for on the web. The same is true with mailing-lists with the added notion that mailing-list membership may be more impactful than its traffic. Nobody writes everything they think about, but reaching someone through a mailing-list not only implies a wider distribution of knowledge but the actual generation of new ideas. While these processes aren't dependent on the open-source model, seeing them through the history of open-source might lead future users in interesting directions. ========================================================== * user participation in trafficdodger.com 2004-06-22 22:44:16 yogimind ========================================================== /Now consider a counter-example. MapQuest is another pioneer that created an innovative type of web application that almost every Internet user relies on. Yet the market is shared fairly evenly between MapQuest (now owned by AOL), maps.yahoo.com, and maps.msn.com (powered by MapPoint). All three provide a commodity-business powered by standardized software and databases. None of them have made a concerted effort to leverage user-supplied content, or engage their users in building out the application./ Dear Tim, my company has a website for doing real time routing based on current traffic conditions, and I'm interested to hear how you think we might be able to introduce user participation into the paradigm. Think of it as a step up from the static routing provided by mapquest et al. http://www.trafficdodger.com Your insight into this would be most appreciated. The cost of the map data alone represents a significant cost in exploring this concept on a national scope (were currently routing localized to LA county). I'd love an opportunity to spin and grow this into something of overwhelming utility ala google. Kind Regards, Yogi