3 ms·

Right, but look at history: MySpace died out (mostly), and was replaced, but by what? Facebook, another centralized, proprietary service. And then came Twitte

Right, but look at history: MySpace died out (mostly), and was replaced, but by what? Facebook, another centralized, proprietary service. And then came Twitter, which was yet another centralized, proprietary service. When has an open, decentralized product or service ever supplanted a centralized, closed, proprietary one? Never, that I can think of. Linux never replaced Windows, not for the vast majority of users. It did become very popular for servers, but really it took over from proprietary UNIX there, and really created its own market because back in the early 90s there was no such thing as a webserver; the whole use-case is different with Linux on servers. So UNIX was never really entrenched in exactly the things that Linux servers are used for now. Then there's smartphones and embedded devices like that, but there again Linux really created its own market; before Linux, small devices generally didn't have enough power to run a heavyweight OS like Linux. Android (semi-open, uses Linux) came about pretty much simultaneously with the iPhone. So here again, there wasn't some single proprietary thing that was dominant and entrenched in a mature market. There's the Apache webserver, but here again, there was no big dominant proprietary player. LibreOffice hasn't had much success in supplanting MS Office. MySQL is very popular, but here again there was no dominant proprietary web database; MySQL created that market (though PostgreSQL has become a serious competitor to it, but it's also open). So in summary, the only time I've seen open, non-proprietary solutions win out is when the environment changed, frequently because of their presence and availability, making them the natural pick. I've never seen them beat out an entrenched dominant player in a mature market. So I definitely don't expect to see it happen with Facebook. What I could see happening is something coming up which uses a non-proprietary or decentralized system which creates a whole new market, rendering one of those things obsolete. I have no idea what that could be though, but then again 20 years ago I wouldn't have imagined that people would want to spend massive amounts of time on a "social network" sharing inane pictures and videos and making idiotic comments to each other about them.

ktRolster 11y ago
That's getting close to a "no true Scottsman argument." You go through the list, find every successful open source project, and try to explain why it doesn't count. Some of your examples are flat out wrong: Oracle was the big dominant proprietary player when Yahoo decided to risk going with MySQL (which was great, and I can't wait for MySQL to get replaced by Postgres). Now think how many websites are running on WordPress. It pains me to say it (because the code quality is so bad), but that's an example of successful open source project. Git is up there too. Don't like Github? Switch to Sourceforge. Don't like Sourceforge? Run your own server. So if you define the category narrowly enough, anyone can say, "no open source project with these exact characteristics has ever had success," and be right. But because it's been limited so heavily, the example from the past will have no predictive power for the future. You need to find a different way of analyzing the problem.
- Grishnakh 11y ago
  > You go through the list, find every successful open source project, and try to explain why it doesn't count. Wrong, I explain why in each case, they've completely changed the market, they didn't just come in and take over from a successful and dominant proprietary player. >Oracle was the big dominant proprietary player when Yahoo decided to risk going with MySQL No, they weren't. Oracle was never big in web databases. Ever heard of the LAMP stack? There was never a "LAOP" stack. Oracle was never used as part of a stack with Linux and Apache and Perl or PHP. Oracle was used in a bunch of big-iron systems back then (and still are), but they were never a serious player in small webservers. MySQL helped create that market. This is something like claiming that Harley-Davidson took over part of the market owned by the Ford Model T truck. >Now think how many websites are running on WordPress. It pains me to say it (because the code quality is so bad), but that's an example of successful open source project. Which proprietary product did it out-compete? That's right, none. WordPress created a whole new market. It's just like all my other successful FOSS examples. It never successfully out-competed a dominant proprietary product. >Git is up there too. git is probably about the only FOSS product, now that you mention it, that's made any real inroads against proprietary products, and even here, not that much. It's used all over the place for FOSS development of course, but there was never a dominant proprietary product used for source control for FOSS development. BitKeeper was the closest to it, and that was only used for a single project AFAIK (the kernel), and using a special gratis license. FOSS development has always been done with FOSS VCSes: CVS, SVN, Mercurial, etc. Now, git has made some inroads against ClearCase and Perforce in corporate software development (as have all the other ones: SVN, CVS, Hg, etc.), but usually only for new projects. ClearCase has been very hard to dislodge in the companies where it's been standardized on (even though it's an absolutely abysmal version-control system that only succeeds in hindering progress). >So if you define the category narrowly enough I've already stated my requirement. Show me a good example of a FOSS product completely taking over for a highly successful and dominant proprietary one, instead of just creating a new market or expanding the market (while not displacing the proprietary product at all). I don't think you'll be able to come up with a single one.

ktRolster 11y ago

That's getting close to a "no true Scottsman argument." You go through the list, find every successful open source project, and try to explain why it doesn't count. Some of your examples are flat out wrong: Oracle was the big dominant proprietary player when Yahoo decided to risk going with MySQL (which was great, and I can't wait for MySQL to get replaced by Postgres). Now think how many websites are running on WordPress. It pains me to say it (because the code quality is so bad), but that's an example of successful open source project. Git is up there too. Don't like Github? Switch to Sourceforge. Don't like Sourceforge? Run your own server. So if you define the category narrowly enough, anyone can say, "no open source project with these exact characteristics has ever had success," and be right. But because it's been limited so heavily, the example from the past will have no predictive power for the future. You need to find a different way of analyzing the problem.

Grishnakh 11y ago
> You go through the list, find every successful open source project, and try to explain why it doesn't count. Wrong, I explain why in each case, they've completely changed the market, they didn't just come in and take over from a successful and dominant proprietary player. >Oracle was the big dominant proprietary player when Yahoo decided to risk going with MySQL No, they weren't. Oracle was never big in web databases. Ever heard of the LAMP stack? There was never a "LAOP" stack. Oracle was never used as part of a stack with Linux and Apache and Perl or PHP. Oracle was used in a bunch of big-iron systems back then (and still are), but they were never a serious player in small webservers. MySQL helped create that market. This is something like claiming that Harley-Davidson took over part of the market owned by the Ford Model T truck. >Now think how many websites are running on WordPress. It pains me to say it (because the code quality is so bad), but that's an example of successful open source project. Which proprietary product did it out-compete? That's right, none. WordPress created a whole new market. It's just like all my other successful FOSS examples. It never successfully out-competed a dominant proprietary product. >Git is up there too. git is probably about the only FOSS product, now that you mention it, that's made any real inroads against proprietary products, and even here, not that much. It's used all over the place for FOSS development of course, but there was never a dominant proprietary product used for source control for FOSS development. BitKeeper was the closest to it, and that was only used for a single project AFAIK (the kernel), and using a special gratis license. FOSS development has always been done with FOSS VCSes: CVS, SVN, Mercurial, etc. Now, git has made some inroads against ClearCase and Perforce in corporate software development (as have all the other ones: SVN, CVS, Hg, etc.), but usually only for new projects. ClearCase has been very hard to dislodge in the companies where it's been standardized on (even though it's an absolutely abysmal version-control system that only succeeds in hindering progress). >So if you define the category narrowly enough I've already stated my requirement. Show me a good example of a FOSS product completely taking over for a highly successful and dominant proprietary one, instead of just creating a new market or expanding the market (while not displacing the proprietary product at all). I don't think you'll be able to come up with a single one.