| |||||||||||||
Free is too expensive (Economist)Free is too expensive (Economist)Posted Apr 7, 2012 0:25 UTC (Sat) by rqosa (subscriber, #24136)In reply to: Free is too expensive (Economist) by Cyberax Parent article: Free is too expensive (Economist)
> > But users like those are unlikely to contribute to the development of the programs they use, and the entire purpose of FLOSS is to make it possible for every user to become a developer. > Wrong. I can download VisualStudio Express (for free!) and start developing Windows Phone applications in 3 minutes. With nice tutorials, great help system and one of the best IDEs. But that doesn't let you develop Windows Phone itself, nor the platform API libraries bundled with it, nor the majority of applications for it. And it really is true that the core purpose of FLOSS is to let every user become a developer of the software that they use. As a consequence, in FLOSS, "user-friendliness" doesn't just mean making it easy to use the software, it also means making it easy to develop the software. Because (in all but a few of the biggest projects) no one is being paid to work on the software, the project needs to recruit new volunteer developers from its user base, or else it will die. That's why having a Windows-like situation of "you must support old technology (often decade old technology) thus development is nightmare" in the name of easy backwards compatibility would be bad for FLOSS, even if it did succeed in attracting more users. > 1 million users would be able to sponsor (say) 100 full-time professional developers, artists, testers, help writers. I won't believe it's possible for a FLOSS project to fund 100 full-time professional developers with donations from end-users until I see it, and I haven't seen it yet. (The project that's closest to what you describe is probably Linux itself, which has hardware manufacturers employing developers, but that business model won't work for the vast majority of projects.) > one day you might find yourself without open hardware capable of running your OS. That's a real problem, but the fundamental cause of it is that the majority of users don't see why locked-down hardware is a bad thing for them — it prevents them from developing the software they use, but that's not something they would do anyway, so why should they care? (They actually should care, for reasons such as avoiding lock-in, having a larger base of developers, having the possibility that the developers may fork the software in case of a bad mantainer, … but these reasons aren't readily apparent to someone who doesn't know anything about software development.) You seem to be suggesting that attracting users who don't care about being able to develop the software they use to Linux will help ensure the future availability of open hardware, but I don't believe it — I think it would more likely lead to a locked-down hardware platform that runs Linux (and maybe this even already exists, in the form of locked-down Android devices).
(Log in to post comments)
Free is too expensive (Economist) Posted Apr 7, 2012 11:26 UTC (Sat) by khim (subscriber, #9252) [Link] That's a real problem, but the fundamental cause of it is that the majority of users don't see why locked-down hardware is a bad thing for them Of course not! It's good for them! This is what they want! Let's forget about FOSS for the minute and think about consumer devices. Radio, TV, phones (mobile and not mobile). Once upon time all these decides were combined from parts which were easy to replace and it was easy to tinker with them. Some even included principal schemes in documentation! Today they usually have undetachable connectros (components are soldered on using surface-mount technology instead of sockets) and in general are non-serviceable (usually you need specialized tool to even open the cover). Why is that? Well, these are cheaper. They are more robust, they rarely need fixes and if one of them will break it's usually simpler to replace it rather then to fix it. Some people need/want to tinker with electronics - and we have specialized shops and devices for them. But these devices usually sell some of the same parts (sometimes differently packaged)! If you'll try to create separate ecosystem "just for tinkerers" you'll quickly find out that it's not sustainable - there are just not enough of them! They actually should care, for reasons such as avoiding lock-in, having a larger base of developers, having the possibility that the developers may fork the software in case of a bad mantainer, … but these reasons aren't readily apparent to someone who doesn't know anything about software development. You asking users to give up concrete advantages for nebulous freedoms (which they don't want and don't need). It just does not work. You seem to be suggesting that attracting users who don't care about being able to develop the software they use to Linux will help ensure the future availability of open hardware, but I don't believe it — I think it would more likely lead to a locked-down hardware platform that runs Linux Sorry, but this is the only choice. Most of hardware in the future will be locked - simply because users don't case. This battle is already lost: smartphones outsell PCs already - and you can not replace OS on most of them (even if the bootloader is not locked you often don't have an image to use with it). Now we are at the next battle: make sure there are some unlocked hardware which can be used if you want to tinker with it. If most devices are locked yet use FOSS-friendly components then there will be some unlocked (or unlockable) devices for tinkerers. If devices are build around proprietary standards then there will be no FOSS-friendly devices at all. Maybe this even already exists, in the form of locked-down Android devices. Android is good example. Before Android most phones were tightly-locked and it was basically impossible to run your own OS on mobile devices. Openmoko tried to solve this problem in the fashion you suggest, but failed miserably - and it was obvious it'll fail from the onset. Android gave us CyanogenMod and plethora of the devices you can use it with. Sure, you may argue that some imaginary world where all devices are free will be better, but this is not in the cards. Today Linux desktop survives on coattails of server market (where "freedom to tinker" is still important and will be important for foreseeable future), but it looks like Microsoft is finally wising up to the problem. If Microsoft will split standards for the desktop (fully locked up with some concessions to the enterprise - see how Apple does this with iOS) and server then this will be the end of desktop Linux (and Intel will be very happy indeed because people will finally stop using cheap desktop components for servers).
Free is too expensive (Economist) Posted Apr 7, 2012 13:47 UTC (Sat) by rqosa (subscriber, #24136) [Link] > Of course not! It's good for them! This is what they want! If someone actively prefers (rather than just doesn't care) general-purpose computing hardware to be locked down, like the iDevices are, then that person is fundamentally opposed to the core ideology of Free Software: users deserve to have complete control over the software and hardware that they use. The FLOSS community shouldn't be trying to bring those people into the userbase, unless they first convert to the FLOSS ideology. (And if you actively prefer locked down devices, which it sounds like you do, then you are obviously an enemy of this community!) > You asking users to give up concrete advantages for nebulous freedoms (which they don't want and don't need). That's what the FSF has always done, since the very beginning. > It just does not work. If it didn't work, then FLOSS would never have gained any significant amount of users in the first place. Clearly that's not the case. One reason why it does in fact work is that the freedoms aren't "nebulous" like you say, but instead there are real concrete benefits to them. For example, back when I used Windows 3.11, often something would not work (or would mysteriously stop working), and I would have no idea how it might be fixable, but now that I use Linux, I'm more often able to fix problems on my own. > Sorry, but this is the only choice. Most of hardware in the future will be locked - simply because users don't case. It sounds like you're trying to convince the FLOSS community to accept Linux-on-locked-down-hardware, because it's more important to have Linux gain lots of technically-illiterate users than it is to preserve user freedom. But we'll never accept that, because we don't care about "Linux" as such; freedom is all that matters to us. Also, "Most of hardware" doesn't matter — what matters is that enough unlocked hardware remains available on the market, at low enough prices. And right now there seems to be plenty of such hardware, including desktop/notebook/server PCs (as they always have been) as well as the likes of the BeagleBoard, PandaBoard, IGEP, and the upcoming Vivaldi tablet.
Free is too expensive (Economist) Posted Apr 7, 2012 15:45 UTC (Sat) by khim (subscriber, #9252) [Link] > You asking users to give up concrete advantages for nebulous freedoms (which they don't want and don't need). Sorry, but facts don't support your claims. First release of GCC supported VAX and SunOS. Later Cygnus developed extensive system which brought GNU tools to users of proprietary OSes and it was enthusiastically endorsed by FSF. And even when Linux finally made GNU/Linux OS usable support for users of other platforms continued. RMS and FSF may sound like stuck-up zealots at times but they are very practical when it's important. > It just does not work. See above. All the FOSS “success stories” (server, embedded, Android, etc) are in places where people care about normal users and not just about FOSS lovers. On desktop, where distributions reject such people Linux is confined to aforementioned 1% and lives at mercy of proprietary brethren: most (if not all) hardware for desktop is created without Linux in mind. Linux support is added later if at all. Compare support for server-related hardware (CPUs, NICs, etc) and for Desktop-oriented one (GPUs, WiFi, etc) This is dangerous situation to be in.
Free is too expensive (Economist) Posted Apr 7, 2012 16:39 UTC (Sat) by rqosa (subscriber, #24136) [Link] > First release of GCC supported VAX and SunOS. Later Cygnus developed extensive system which brought GNU tools to users of proprietary OSes and it was enthusiastically endorsed by FSF. And even when Linux finally made GNU/Linux OS usable support for users of other platforms continued. None of which negates my point. Using free software has always involved giving up certain "concrete advantages" (in the early pre-Linux days, the disadvantages would have been forgoing first-party support from the system vendor, and spending extra effort to installand keep up to date a bunch of third-party software across all systems in the administrative domain) but getting other concrete advantages (such as the ability to fix problems yourself, the "with many eyes, all bugs are shallow" effect, the likelihood that a project with a bad maintainer will be forked rather than killed off entirely, etc.) in return for the ones given up. > All the FOSS “success stories” (server, embedded, Android, etc) are in places where people care about normal users In the examples you cited above (Cygnus, GCC on VAX and SunOS), how did "Joe Average"-type users matter at all? The people who decided to switch over to GNU tools on those minicomputer/workstation platforms were developers and system administrators, not technically-illiterate end users.
Free is too expensive (Economist) Posted Apr 7, 2012 17:20 UTC (Sat) by khim (subscriber, #9252) [Link] The people who decided to switch over to GNU tools on those minicomputer/workstation platforms were developers and system administrators, not technically-illiterate end users. “Joe Average” != “someone dumb and stupid”. “Joe Average” is Joe, you know, Average. VAX and SunOS users typically knew how to compile things and for them Cygnus solution was “good enough”. In fact it was easier to use them many alternative commercial offers. Today Joe Average no longer knows or cares about compilation of programs from scratch. Thus the same solution is rejected. In the early pre-Linux days, the disadvantages would have been forgoing first-party support from the system vendor, and spending extra effort to install and keep up to date a bunch of third-party software across all systems in the administrative domain. Bullshit. GNU software was always parallel-installable and never required “forgoing first-party support from the system vendor”. But getting other concrete advantages (such as the ability to fix problems yourself, the "with many eyes, all bugs are shallow" effect, the likelihood that a project with a bad maintainer will be forked rather than killed off entirely, etc.) in return for the ones given up. Bullshit again. Most users of GNU software in these days cared about additional features, not about source availability. The fact that bash included nice command-line editing facility was more important then the fact that it included sources. The fact that GCC was free (while SunOS compiler was expensive) was the driving force, not the code availability. Sure, at some point a lot of users have “looked under the hood” (because they had the ability) and some even become contributors, but it was never (or almost never) the motivation for the initial installation of GNU tools.
Free is too expensive (Economist) Posted Apr 7, 2012 18:55 UTC (Sat) by rqosa (subscriber, #24136) [Link] > VAX and SunOS users typically knew how to compile things I know firsthand that many VAX users didn't have anywhere close to that level of technical expertise. For example, people used to use dumb terminals on a VAX (or other minicomputer) for email at work (I remember one such place that used Pine on VMS, for example), or for using their employer's in-house applications, or even for searching public library catalogs. These users knew less about computers than most people today, and yet they probably outnumbered (by far) those who did know how to compile things. > Bullshit. GNU software was always parallel-installable and never required “forgoing first-party support from the system vendor”. Yes it did — the more that someone was using GNU software (or other third-party software, generally), the more they'd have to rely on other sources of help than the first-party system vendor. It's just that back then they were still using plenty of software from the system vendor, whereas with GNU/Linux on commodity hardware now, the proportion of non-system-vendor software has increased to almost 100% — though of course, some high-end server hardware vendors actually have first-party support for GNU/Linux now. > Most users of GNU software in these days cared about additional features, not about source availability. […] it was never (or almost never) the motivation for the initial installation of GNU tools. But that just goes to show that it's entirely possible for FLOSS to compete against proprietary software on practical features, and win. And it's not just a coincidence, either — the greater functionality of GCC and coreutils was a direct consequence of the ability of skilled users to join the developer-base, thus giving them a larger and more meritocratic developer-base than their proprietary counterparts had.
|
|||||||||||||