Belief and open source

Belief and open source

Summary: As a business model, in other words, open source remains incomplete. While the machine runs great for enterprises who know they have IT budgets, its utility is limited for consumers, artists, and small businesses who need income, not just a cut in expenses, to make it through the day.

SHARE:
TOPICS: Open Source
12

Mark Asay writes that he believes in open source, because it works, and because it follows economic dicta. (Photo by Peter Cunningham for WordWiseWeb.)

I don't believe in open source any more than I believe in science. Open source a process, which reveals its truths in the market. What works survives, and what doesn't work fails. True for businesses, true for concepts, and true for economies overall.

Right now open source is working great for enterprises. Is it working for consumers?

There the results, so far, are mixed. Sure we have Google and other Internet companies built on an open source platform. But would they not exist if they were built on a proprietary platform?

Google and the other providers of Internet services also demonstrate the big problem with open source for consumers. They cut prices to zero, but they also take value from our contributions to the Net. (See my tax return for one example out of millions.)

Even the promise YouTube has made to start paying its contributors will deliver them a tiny fraction of the value Google gains, and it's not a negotiation.

Open source is doing even worse in the desktop market. Apple and Microsoft are eating our lunch there. Sure, there are great open source applications like Firefox and Open Office, but where's the business model going?

As a business model, in other words, open source remains incomplete. While the machine runs great for enterprises who know they have IT budgets, its utility is limited for consumers, artists, and small businesses who need income, not just a cut in expenses, to make it through the day.

That's what the market is saying right now. If you don't agree, prove me wrong. Science and the market are like that, because belief is not the basis for either.

Topic: Open Source

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

12 comments
Log in or register to join the discussion
  • good piece

    you hit the nail on the head with this one, Dana. Many in the industry (mylsef included) are prone to pontification about why open source is so wonderful, but the evidence is, in some cases, thin.

    Let me offer one thought to your point about YouTube, where you say "Even the promise YouTube has made to start paying its contributors will deliver them a tiny fraction of the value Google gains, and it's not a negotiation."

    I don't disagree with you a bit, however, by starting to pay contributors, Google is opening up a new competitive front with rival outlets, like bliptv (http://blip.tv), that, if they are smart, rivals will sieze upon by one (or two, or three)-upping Google in the percentage they give to contributors. Further, I think this reinforces your broader point that the market decides winners and losers, not the peanut gallery. However, it does raise another question about economies of scale. Can open source/user-contribution-based markets sustain more than one major "provider" or "hub?" I think the jury is still out on this...
    gtewallace
  • Adopters and adapters

    If acceptability of open source is determined by willingness to modify what's received, then the market for open source is smaller than that for the alternative.

    I think that's the assumption behind your comment:

    "As a business model, in other words, open source remains incomplete. While the machine runs great for enterprises who know they have IT budgets, its utility is limited for consumers, artists, and small businesses who need income, not just a cut in expenses, to make it through the day."

    Software used rarely creates income. But it can reduce expense, at a cost ;-)
    (Someone used that phrasing today.)
    Anton Philidor
  • The name is Matt, Dana

    ...not Mark. Not sure where you got it that my name is Mark, but it really isn't. :-)

    I would disagree with you on open source not making it on the desktop. Firefox,
    Adium, Handbrake, etc. There is only one proprietary application that I use on the
    desktop: Office. It's not because I'm "blinded by belief." It's because over time
    I've found these open source projects to be better than the proprietary
    alternatives. I'm sorry that you haven't (yet) discovered the same. Most people
    haven't. But they will. It just takes an itch ("I really need an app to rip movies to
    my hard drive - what is available? Oh, and I'd like it to be free.") and then Google
    helps you scratch it (Handbrake, in this case).

    That said, maybe the desktoop market never will be consumed by open source.
    Maybe it won't be kept by proprietary software, either. Google and Yahoo and
    their ilk are increasingly making it look like it doesn't have a strong future,
    anyway.

    So, perhaps, open source is consuming the markets that matter.

    You should try selling it for awhile. I think you'd believe, too. I've been selling
    open source software since 2000, and there has been a MASSIVE shift in buying
    behavior in the past three years. It's easy to miss when you're writing about open
    source; it's impossible to miss when you're selling it/making a living from it.
    mjasay
    • blinded by ________

      "You should try selling it for awhile. I think you'd believe, too. I've been selling open source software since 2000, and there has been a MASSIVE shift in buying behavior in the past three years. It's easy to miss when you're writing about open
      source; it's impossible to miss when you're selling it/making a living from it"

      So how much did you earn from selling open source software. What was the average selling price. What were the products sold. Could you give more specifics. Anyone can make any claims. However giving specifics and backing it up with proof is sorely lacking from your claims.




      "There is only one proprietary application that I use on the
      desktop: Office. It's not because I'm "blinded by belief." It's because over time I've found these open source projects to be better than the proprietary alternatives. I'm sorry that you haven't (yet) discovered the same"

      Again, specifics. Are you a software developer. Specifics would surely be appreciated.
      Read up the excellent blog by progon
      http://talkback.zdnet.com/5208-10535-0.html?forumID=1&threadID=29654&messageID=551722&start=-45
      zzz1234567890
  • From a software developer viewpoint

    Open source provides software tools for development that can only be described as ancient. Open source has certainly caught up to Windows 98 but in spite of being based on a full server OS and in spite of being based on the best ideas from last century open source has to get a central set of themes that make it useful to the masses.

    Windows NT, for example, is a server OS. It solved the problems in the Windows world that were there from Windows 98, etc, being just a user land OS. This, of course, created a large stack of applications that were both corporate and user land based.

    Windows XP does both corporate and home user stuff very well. It is even more secure than any of the linux versions out there if it is managed properly. Microsoft filled most of it's sercurity holes through a shift in user practices. The OS itself is quite easy to secure but the Windows users were not taking the actions to secure the OS. By the time service pack two came along, the user base was forced into a default secure system.

    The problem with most of the older software in the old Windows was that it is written in C or C++ like all the other OSes of that era wrote software in. Old school Windows had an open and huge application programming interface that took years to master. These hooks into the C and C++ based system were the source of most of the faults in the old windows. Microsoft said that if you follow their software practices and not make your software use unpublished or othe API hooks that were subject to change you could write software that Microsft could guarantee to be used in all future versions of Windows.

    The recent flap of quickbooks not being written to these specifications is a prime example of old school software development practices that make the software in question both brittle, requiring re-writes when the old software API vanishes or gets changed to deal with new hardware and other issues.

    But, when Microsoft went to the .net framework and limited the API calls to the new .net API, which is a whole lot smaller and safer set of API's they also moved away from the down to the metal compiled applications that cause the security holes in the first place. This new approach makes it possible to have a secure system on the internet. By moving away from a code base filled with the flaws in C and C++ applications like buffer over flows or underflows, the problem base in Windows has vaporized and the security of the newer Windows systems in Vista cannot be touched by anyone using the down to the metal C and C++ application approach.

    Unix is all about creating systems that are secure but the problem is that no C or C++ application is really secure until it has been used under fire. This kind of application development process is the source of most of the flaws in all operating systems that use this approach. C and C++ applications require an incredible amount of work to get them to work and to get them secure. Most of the time in the Unix world is devoted to the needs of finding and fixing the inevitable problems in a C or C++ based application.

    If Sun, for example, had released Java to the open source community like it promised back in the 90's it would have made the need for the new Windows API moot. Now Sun has been made moot by open source instead. The problem with Sun is that they cannot hire enough developers to find all the bug induced holes in their own stuff. The open source movement provides millions of new bug catcher eyes but every new application written in C and C++ requires millions of eyes to just find the bugs in them.

    This is why I say that open source has managed to replicate and finally caught up to Windows 98 and possibly caught up to the first version of NT. Until the unix crowd gets linux totally hardware agnostic and move away from the old world software development approach they use Microsoft will keep eating them alive. This all applies equally to the Macintosh, which is also another freeBSD variant of linux. The cost of development of compiler based software is the cause of the biggest cause of develoment costs at Apple as well.

    Apple is old school and still has one small group of people determining the software and hardware that has to be used in order to be a Mac user. The move to intel hardware only means that Apple is now another linux. Sun had the same problem. They can't even give their stuff away so they had to open source it. The world is not going in the direction of networked computers managed by a single vendor. The people in spite of moving to more and more gizmos still need that data served up to them. Only Microsoft is totally gizmo agnostic as well. The blend of realities represented in Vista of the web world mixing with the personal computer is a brilliant move by Microsoft and will make their OS even more relevant over time here than any unix based model can ever accomplish here.

    The blending of the web on a personal computer is what Microsoft has done. The web 2.0 stuff takes advantage of just one of the new elements in the .net world. The rest of the .net world makes it easy to write secure software and to use software right from the web with no extra securty software since in the new Vista world the front end is based on the xaml script language and not hard coded it is possible to do applications with a mix of local server stuff and internet server stuff and allows a very rich set of applications that are web seamless. That is what .net is all about and in spite of the claims by the web culture that they are superior to the hard personal computer is only possible because of advances in the Microsoft world and not caused by open source at all.

    How would we build a .nix (dot nix) world that is equally smart? Well, a web browser can have any kind of front end you want. The xaml approach in Vista is a very smart one and when the unix crowd finally figures out that the old paradim of C and C++ based applications is dead they can easily move to a scripted front end and lose most of the hassles found in the down to the metal programing pardigm they currently use. What happens if you move the entire front end experience in unix to a core set of processes managed by just a few servers instead of multiple graphical front ends? Well, you get your cake and eat it too. Currently, there is no way to break up the love of either Gnome users for Gnome or KDE users for KDE and yet it is quite possible, using xaml in Vista, to create the exact look and feel of either OS front end with very little code. The Mac look and feel can also be scripted in Vista as well.

    They don't even need xaml to do this, either, just come up with a whole system web browser approach used in .net and put in safe and secure and limited API interface to these scripting engines and they can innovate as much as they want to for any kind of user experience they want to use.

    It will take a little extra work in the unix world to catch up here but the worst tool out there is Java. It cannot be made better and the unix world is far better off coming up with their own c# variant instead. Ruby is a really good smalltalk variant and is quite capable of being made to serve as a core script language. Python, Perl and others can also be used but the real trick will be to get a web browser based front end like xaml. They might even be able to use xul, too. But, that is what will be required to catch up to Vista and keep the open source movement from becoming moot themselves. It is time to get away from the old paradigm in unix and take advantage of it's server heritage and create a better front end and back end to unix before the open source crowd can ever make the claim that the unix approach is better since it is not.
    progon
    • Have you looked at Linux in the past few years?

      [i]It will take a little extra work in the unix world to catch up here but the worst tool out there is Java. It cannot be made better and the unix world is far better off coming up with their own c# variant instead.[/i]

      *cough*[url="http://www.mono-project.com/Main_Page"]Mono[/url] *cough*

      Pardon me, I seem to be catching a cold. ;)

      Anyway, I don't think you should write off C/C++ so easily. It's still necessary for low-level stuff like drivers and the kernel itself(this applies to ALL operating systems, including Vista). Not to mention it's still the lingua franca for embedded devices.

      And as far as your comments on breaking up KDE vs. Gnome, why do that? A Linux system can have both desktop environments, and run KDE programs in Gnome(and vice versa). And with most default Linux installs, C/C++ development tools are NOT automatically installed, as scripting languages with GUI libraries are preferred. As a case in point, Ubuntu and Fedora Linux both prefer using Python with GTK or Qt bindings for their own apps. And with Gnome, Mono-based apps are becoming the norm.

      The Linux crowd is already moving toward a more .NET-like development model, but the change is taking a bit slower than it did for Windows, because the difference between the old Win32 API and the *nix C/C++ application APIs is very large. Have a look at the full code for a program that makes a dialog box that says "Hello, World!" in Win32 API and an equivalent Linux program written in GTK or Qt, and you'll see what I'm talking about.
      Tony Agudo
      • oops, link was messed up

        *cough* http://www.mono-project.com/Main_Page *cough*

        There ya go. :D
        Tony Agudo
    • RE: From a software developer viewpoint

      Wow progon, that was very loquacious. It looks like you used proper grammar for the most part. It, however, is mostly wrong and misleading.

      As I don't know who you are, I don't know how you came to your conclusions but you should really research some.

      <i>It is even more secure than any of the Linux versions out there if it is managed properly. Microsoft filled most of it's security holes through a shift in user practices.</i> First off, Linux is capitalized as it is a proper name. Second, how can you even say that WindowsXP is more secure than that of any Unix variant? The default install of Windows leaves the users as administrators, which is required to run many apps in windows, and that's with all updates and patches installed. In no way is Windows secure by default.

      <i>But, when Microsoft went to the .net framework and limited the API calls to the new .net API, which is a whole lot smaller and safer set of API's they also moved away from the down to the metal compiled applications that cause the security holes in the first place.</i> You make it sound as if Vista was written in .NET or maybe C#. It wasn't. Security is a process and not a function of what language you use. C, C++, FORTRAN, COBOL, VB, or even Java can be used to write extremely secure apps. Of course, that depends on the developer, not the language. In fact, I'd bet that many an argument could be made that coding quality decreases as ease of use increases for languages.

      <i>Most of the time in the Unix world is devoted to the needs of finding and fixing the inevitable problems in a C or C++ based application</i> Replace Unix with any OS and 'C or C++' with any language and the statement holds true.

      <i>Until the unix crowd gets Linux totally hardware agnostic and move away from the old world software development approach they use Microsoft will keep eating them alive.</i> What??? Windows runs only on x86 based HW. Linux runs on everything from x86 to Cell to arm to just about every architecture you can name. But maybe you mean to say the peripherals need to work better on Linux. That is not the fault of Linux. The manufacturers need to provide drivers like they do for Windows. In short, Linux developers can do very little about this.

      OK, I have to stop quoting you. It's taking to much space. I'll continue with a short list of some other issues you have. I'm pretty sure that others will carry on with the critique...

      OpenBSD is not a variant of Linux. Let's just lay it out right here, Linux is not Unix, BSD and its offshoots are.

      (SUN)Solaris is a Unix System V based OS. Again, more closely related to BSD than Linux. In fact, SunOS was BSD based. :)

      All applications of any size require lots of work to make them secure and find all the bugs! (duh!)

      And my last little tidbit is that the open source world already has a .NET equivalent. It's called MONO. Whether it becomes widely used is another story.
      jtiner
  • how does a company compete against products that are given away for free

    Most people would say that its because Microsoft is a monopoly and ......

    Open Source is unable to compete because of a flawed business model.

    However in my opinion---
    Microsoft is competing, even when the competition is giving away products for free. The reason Microsoft is able to compete and thrive is because of
    a) leadership
    b) excellent work ethic from everyone including the top management
    c) good employees

    If you could imagine another company in the position that Microsoft is in but with a few things different such as different leadership, different company culture, most likely they would be loosing to open source and other competition.

    Claims of monopoly status has nothing to do with success of Microsoft.
    zzz1234567890
    • You must not live on this planet

      Microsoft lock-ins have nothing to do with their success you say? Odd, the need to be compatible with other Microsoft software has always been the driving force for getting Microsoft software my entire career.
      mosborne
  • he's talking about the market issues

    Matt,

    the products you name are first rate, but they return ver little cash to their
    creators/investors. As you well know from your time at Novell, service is the
    name of the game for Open Source.

    If your product is so good, or so simple (like Adium - love it!), then there is very
    little service money thrown off by it. Sure, you might be able to customize Adium
    for a specific corporation. But really, there is very little cash on the table.

    So while there may be gems out there that make it through a labor of love - the
    spit and polish that we see on the commercial products take a commitment that is
    costly. And nothing in the technology revolution has changed the old adage "Time
    is money".

    morganew
    Morgan_Reed
  • Artists and open source

    > As a business model, in other words, open source remains
    > incomplete. While the machine runs great for enterprises
    > who know they have IT budgets, its utility is limited for
    > consumers, artists, and small businesses who need income,
    > not just a cut in expenses, to make it through the day."

    I think your view of artists is skewed. It is so tough to make a living from creative work these days most of the artisans I know have day jobs. And a lot of them are in IT.

    Yes, some of them use Windows or MacOS at home, but even some of the latter understand Open Source and are able to incorporate it into their work. I'm talking artisans, of course, not computer artists who for various vendor lock-in reasons would have more reason to use Windows products.

    By the market's nature, a professional artist is likely to be underreported in any survey. Few are accessible to surveyors.
    jplatt39