标题: What's Wrong with the Open Source Community? [打印本页] 作者: cyan 时间: 1.12.2003 19:18
Quote from <a href='http://www.linuxworld.com/story/38073.htm' target='_blank'>http://www.linuxworld.com/story/38073.htm</a><br><br>James Turner: 5 problems with the Open Source community<br>There’s no question that the Open Source community has a lot going for it. Besides a staggering amount of developer power that can be turned against important problems, the Open Source movement also has a passion and commitment to its work that the commercial software world often envies. But sometimes, the Open Source community can be its own worst enemy. Here are a few reasons why. <br><br><br>1. Too many developers “scratch the same itch.”<br>We hear that Open Source developers come up with new ideas because they “had an itch to scratch.” In other words, there was some need they had for a new application, and they “scratched” it by coming up with a tool. The problem is, it’s not uncommon to end up with two or three (or more!) different packages doing the same thing. For a specific example, look at what’s happened with the Linux sound systems, where there are now several competing packages that have to be supported by each distribution. Or in the Java world, look at how many competing MVC frameworks there are now for JSP development. <br><br>A little competition can be a good thing. After all, Linux is all about offering a competing vision for the operating system domain. But when too many competing visions exist, and aren’t winnowed down to a small number of options over a short period of time, you end up with a mish-mash of conflicting standards, and a user community that ends up having to download and install a plethora of different packages that all do the same thing. <br><br>A perfect example of the “too many itches” syndrome is the absurd number of Linux distributions that exist out there. There’s absolutely no reason for there to be more than two or three distributions. And because each one does things slightly differently, we’ve ended up with the problem that applications and drivers are rarely made available in binary form, because there are too many versions of too many releases of Linux to support. <br><br>As an application developer, you would have to provide 5 - 10 different binary installs, one for each distribution. Now multiply that times the five or more active releases of a distribution that may be in active circulation, and you see why so few packages are available as anything but source (especially the most recent releases of packages that have not been compiled and included into Linux distributions yet.) <br><br>The next question to consider is, why don’t we see more consolidation of technology? The answer: because… <br><br><br>2. Open Source developers love a good feud.<br>BSD vs Linux. Gnome vs KDE. Debian vs Red Hat. For every interesting Open Source technology, there are two bitterly feuding camps that spend as much time taking potshots at each other as in improving their own products. <br><br>It’s hard to imagine how much better a lot of Open Source software would be if these groups cooperated and consolidated their efforts, rather than act like the Hatfields and McCoys. Unfortunately, the downside of personal commitment to projects is that people come to use them as a measure of self-worth, and it becomes increasingly difficult for rival groups to admit the good points in each other’s efforts. <br><br><br>3. Open Source developers often scratch the wrong itch. <br>The problem with commercial development is that the developers often aren’t the consumers of their products, and thus don’t feel the pain of their mistakes. The problem with Open Source development is that the development community often doesn’t fix problems or develop new features that aren’t directly interesting to them. <br><br>Usually, this isn’t a problem, because the developers (as users) encounter the same problem set as their user base. Unfortunately, one way that Open Source developers are different from a general user base is that they have significantly more technical training. This means that they are willing to put up with the need for a much higher degree of technical savvy to use something than a non-technical person might. <br><br>Restated by example: an Open Source developer might think nothing of requiring users to create and configure an XML file to make something work, where an end-user might require dialog boxes. <br><br><br>4. In the Open Source Community, you’re either “with us or against us”<br>A typical complaint of the Open Source community is that proprietary software vendors use legal means to stifle criticism of their detractors. But the Open Source community can be just as unforgiving of internal critics. Attempts to point out flaws or places where there’s room for improvement in an application usually lead directly to defensive rebuttals, character attacks on the critic, or complete rejection of the validity of the issues. <br><br>Consider that recently I posted a story on the linuxworld.com Web site listing some problems I saw with the current set of desktop Linux distributions, problems I thought could severely hamper consumer adoptions of Linux in the short run. The posted responses ran in a couple of themes: “It works fine for me, you must be an idiot.” “You’re nothing but a Microsoft ass-kisser.” And the ever-popular “Windows sucks too.” <br><br>Until the community learns to listen to and internalize negative feedback (oops, almost slipped into Pointy-Haired Boss speak there…), it will be staring at its navel. <br><br><br>5. The Open Source Community has a huge chip on its shoulder…called Microsoft<br>Although SCO is also a popular a target lately too, the merest mention of MS is like a bull having a red cape waved before his eyes. All reason and sense of decorum flies out the window. And while I’m first in line to throw rotten tomatoes at Bill Gates, it’s harmful to the community. The reality is that Microsoft owns the lion’s share of the non-server OS market. If the first thing you tell all these people who own Windows is that they are idiots, you’re not starting out on very good ground to convert them. <br><br>Like it or not, the existing Windows user base may not like the dreaded Blue Screen of Death or Microsoft’s pricing and licensing, but they know how to use Windows and can usually get the applications and hardware support they need for it. Linux has a wonderful and growing suite of tools that let people migrate away, but they are going to need a lot of hand-holding to decide to make the move. They have to be told why Linux is better (and it really has to be better for them), not just why Windows is trash. <br><br>Especially unhelpful is the “who cares about X” attitude, where X is unsupported hardware, non-existent game availability, complicated multimedia support or anything else that Linux has or is perceived to have problems doing. Just because someone wants to do something that you don’t, it doesn’t mean that what they want to do is less important. <br><br>I had a number of comments when I complained that I had great trouble getting my DVD player on my laptop to view commercial films, comments that essentially said “why are you watching DVDs on your laptop?!” Some even suggested that I buy a dedicated portable DVD player. Leaving aside the hassle of having yet another piece of electronics to drag through security if I want to watch a movie on a plane, these kinds of comments are the worst kind of evasive nonsense, based on: if Linux doesn’t currently do something as easily as Windows, attack the need to do it at all. <br><br><br>To sum up, the biggest problem that the Open Source community faces in taking Open Source to the next level is not some legal challenge or Microsoft marketing campaign. It’s the immaturity and insecurity of some of the members of the community. As was once said in Pogo, “We have met the enemy, and he is us.” <br><br> 作者: cyan 时间: 1.12.2003 19:21
<br>Steve Suehring rebuts James Turner<br><br>Rebuttal<br><br>1. Too much competition <br><br>The first argument made is that too many Open Source developers are "scratching the same itch" with the end result being too much competition. The economist in me is attempting to find a reason why competition would be bad, and James does admit that a little competition can be good. The opposite of competition is monopoly. As we've seen, monopoly in the software business means little or no innovation, virtually no method to make the software secure, price gouging, and much lower quality software. <br><br>Given these two alternatives, I'd much rather have more choice. <br><br>Is there a happy medium? James calls for there to be two or three Linux distributions instead of the numerous flavors available today. I believe the market should and will play the deciding role here. The market will determine the best flavors of Linux available and those will be the most widely supported among vendors. I see no reason to stifle innovation and competition with the goal of less choice. If someone wants to build their own distribution of Linux then so be it. That is the power of Open Source. You are free to do with it as you please. <br><br>Those supporters and developers of other distributions can continue along their path and make their distro better or merge the best aspects of theirs with another. Thus the consumer gets the best-of-breed operating system. The same goes for Open Source software. I would much rather have a choice in Web servers or DNS servers to deploy than be left with one or maybe two choices. <br><br>It is this competition that is one of the strongest aspects of Open Source and it is the most widely misunderstood. I'm constantly surprised to hear seemingly smart people compare Open Source development to some form of communism. In fact, Open Source is the free market system at work. The consumer has a choice among distributions and software and is free to do with it as they please. <br><br> <br><br>2. Love a good feud <br><br>Unfortunately in his article, James believes that we don't see more consolidation because Open Source people love to fight. <br><br>There certainly are strong views among developers of competing software in the Open Source community. Again, I tend to lean towards allowing developers to compete against each other because I believe that in the end the consumer wins. The strongest points of each other's software will be adopted amongst the competing software. <br><br>I don't think the goal of having one operating system or one Web server is best for the market. At the same time, I sincerely doubt that Open Source developers work on competing products because they like to fight. I think the feuds are a side-effect of a healthy market economy at work. <br><br>There are precious few commodities where a monopoly is good - and software is not one of them. <br><br> <br><br>3. Scratch the wrong itch <br><br>James's next contention is that Open Source developers often "scratch the wrong itch". I see this point as somewhat valid but I see it being much more applicable to closed source software than to Open Source. With closed source software you get what you get. If a feature isn't in the software, too bad. Wait for the next release and maybe they'll put it in. You are at the mercy of the software vendor that you locked yourself into. <br><br>An example of the problem of scratching the wrong itch in closed source is pop-up blocking. Everyone's familiar with those annoying pop-up ads that appear when you surf to various Web sites. For years, the open source Mozilla browser has been able to block pop-up ads effectively without disabling all of the other features of Javascipt. Where has this functionality been in Internet Explorer? Microsoft is finally getting around to adding this feature in a forthcoming release of IE, years behind. <br><br>As an example in his article, James uses the argument that an Open Source developer might require a user to create an XML configuration file to make a piece of software work, as opposed to the "ordinary user" requirement - which might be for dialog boxes instead of manual configuration. Aside from the fact that there is an assumption of a graphical user interface which many Linux users don't use, there's an inherent problem with the argument. Having worked with numerous closed source software companies, I can say that the issue isn't limited to Open Source software. I've worked with closed source software where I had to edit a registry setting or manually change a configuration file in order to make the software work (don't forget to reboot the entire server if you make a registry change.) <br><br>The difference is that with the closed source software, I had to pay an extra 20% surcharge in order to receive that support to tell me to change the configuration file. All major Open Source software is documented. Further, the chances of finding an answer quickly using Google are much greater for Open Source than for proprietary software. <br><br>At the end of the day, consumers also have the source code, the one and ultimate resource for determining how a piece of software works and changing that software to suit their needs. If you're not a programmer, then you can request functionality to be included - which is the same process you'd have to follow if it was closed source proprietary software as well. <br><br> <br><br>4. With us or against us <br><br>James' article goes on to state that there is a feeling of defensiveness in the Open Source community. That when you try to take a critical look at Open Source, you are met with harsh responses. To this point I have no counterpoint except to say that it's certainly not limited to Open Source. <br><br>When I wrote a piece last summer for LinuxWorld.com addressing the fact that Microsoft didn't include an adequate firewall in their operating systems it was met with numerous personal attacks, including many sent direct into my inbox. I heard some of the same things that James did, though from a different camp. <br><br>Simply because closed-source folks do it doesn't make it excusable for the Open Source community. Neither side in the debate between Open and closed source is without reproach in this regard. I find no valid excuse for the behavior and feel that both sides should be less concerned with killing the messenger. Of course, I have a vested interest in saying that, since many times I am the messenger. <br><br> <br><br>5. Chip on its shoulder <br><br>James' piece wraps up with an argument that we Open Source people have a chip on our shoulder about Microsoft or, more appropriately from the examples, about people who use Microsoft products. I've seen evidence of this and I agree with James that it is one of the problems with the Open Source community. <br><br>I believe that Open Source software solves countless problems better than its closed source counterpart. However, as I state in forthcoming article in LinuxWorld Magazine, Open Source can win the technology battle but lose the adoption war. The technology can be better, but if we alienate the people who are just starting to use the software, we'll find that the superior technology will be usurped and we'll end up in a niche and not in the business and technology mainstream.