LWN: Comments on "Free is too expensive (Economist)"

Free is too expensive (Economist)

nye — Mon, 16 Apr 2012 12:06:26 +0000

>"We must remember that, from what we can see, the KDE project's primary mission in life is providing entertainment for KDE developers, not convenience or stability for KDE users."

> Can you just stop gratuitously insulting us? It's extremely tiresome.

> You must know you're wrong

To be honest, I for one wasn't aware that that was disputed and was hence under the impression that his statement was an objective fact. I would never have guessed that it would be found offensive, so I doubt there was any insult intended.

Free is too expensive (Economist)

anselm — Thu, 12 Apr 2012 20:16:45 +0000

Oh my god. I don't think you understand how completely outside the realm of feasibility this is for the non-technical user.

The machine will presumably know (or be able to find out) about its own distribution and architecture, so it can tell the app store on the user's behalf. A method to do this is the first thing a prospective app-store proponent would want to standardise.

Free is too expensive (Economist)

khim — Thu, 12 Apr 2012 17:14:42 +0000

I don't think it's even a significant minority, but actually the *majority* of users who don't know what version of Windows they're using - assuming they even know they're using Windows at all.

Well, usually they do know that - Windows is too hyped-up to miss. What they don't know is where Windows ends and other things begins. I've seen the cases when people bought new PC (with Windows obviously) and vainly tried to find Excel there (which they have not bought).

The proportion of people who know what processor architecture they're running has got to be far, far below 1%.

This is quite obviously not true because Linux occupies about 0.5% of the market share and most Linux users know that.

Free is too expensive (Economist)

nye — Thu, 12 Apr 2012 16:55:37 +0000

>So what? You tell the app store that your machine is running Debian GNU/Linux 6.0 on an amd64 architecture.

Oh my god. I don't think you understand how completely outside the realm of feasibility this is for the non-technical user.

When tech support people bitch about users who only know that they're running 'Microsoft', and know that 'the blue e' is the internet, *they're not exaggerating*. That's not hyperbole. It's a literal description of a sizeable proportion of the user base.

I don't think it's even a significant minority, but actually the *majority* of users who don't know what version of Windows they're using - assuming they even know they're using Windows at all.

The proportion of people who know what processor architecture they're running has got to be far, far below 1%.

Free is too expensive (Economist)

anselm — Wed, 11 Apr 2012 07:53:22 +0000

This was tried many times. It does not work. People find something deficient with tiny selection of “Linux preinstalled” offers and buy something else instead.

When I bought my current computer (an HP business notebook), HP did offer one configuration with Linux preinstalled. That was the bottom-of-the-line configuration with the slowest CPU and GPU, the lowest screen resolution, the smallest hard disk and half the RAM of the one I eventually got. It does not come as a big surprise that under these circumstances buying the machine with Linux preinstalled is not the option most customers will take.

On the »plus« side, the machine I bought in the end is also very nice for Linux, with basically everything working out of the box using Debian (I've so far not missed the fingerprint reader, and somebody like HP could probably get that supported by leaning on the chip manufacturer). It does make one wonder why HP does not offer a Linux preinstall for the top-of-the-line configuration rather than the bottom one.

In my experience, installing Linux on notebooks has become a lot easier over the years. Whether this is due to improvements in Linux itself or the manufacturers moving towards supported components is difficult to tell (probably a mixture of both), but the presence of any configurations of a model with Linux preinstalled is a good sign because it indicates that Linux will probably work well on the other configurations, too. There is certainly nothing technical that prevents manufacturers from offering more Linux preinstalls – the reason why this doesn't happen more often is mostly to do with Microsoft's sleazy business practices.

Free is too expensive (Economist)

khim — Wed, 11 Apr 2012 06:05:16 +0000

If anything, a desire to create lock-in would be an argument in favour of preinstalling Linux as long as your Linux distribution is good enough and supports your machine well, because as long as you're the only one selling such a machine people will continue being your customers.

Nope. Others sell similar machines, too. Both tiny firms with full selection of models (tiny selection because firms are tiny) and large companies like Dell, HP, or Lenovo - but with few models (again: Linux is not large enough to support the large range of models).

This works for Apple – Macintoshes could just as well run Windows but people tend to stick with the OS X that comes with the machine.

Some actually install Linux and/or Windows, but that's not the point. The point is that MacOS works poorly on anything else and Apple vigorously ensures that there will be no machines with Hackintosh preinstalled. This is where lock-in scheme starts to work and this is what you can not do with Linux.

In the same vein, if a good computer came pre-installed with a good, supported mainstream Linux like Debian (rather than the low-end boxes with weird Linux distributions that hardware manufacturers tend to offer if they offer anything at all), most people would probably stay with that because putting anything else on it would be more of a hassle than it was worth.

This was tried many times. It does not work. People find something deficient with tiny selection of “Linux preinstalled” offers and buy something else instead. And it makes no sense to create as many models with Linux as you create models with Windows if you expect 10x-100x less buyers.

Free is too expensive (Economist)

anselm — Tue, 10 Apr 2012 22:48:46 +0000

You need bigger margins, not the same margins, or else the whole exercise is pointless. For that manufacturer need some kind of lock-in - that's what I'm talking about.

As a PC manufacturer, you can't »lock in« people to your hardware (unless you're Apple, but we already said that special rules apply to Apple). If anything, a desire to create lock-in would be an argument in favour of preinstalling Linux as long as your Linux distribution is good enough and supports your machine well, because as long as you're the only one selling such a machine people will continue being your customers. This works for Apple – Macintoshes could just as well run Windows but people tend to stick with the OS X that comes with the machine. The margins on Macs aren't quite like the ones on iPhones, but we don't see Apple complain.

In the same vein, if a good computer came pre-installed with a good, supported mainstream Linux like Debian (rather than the low-end boxes with weird Linux distributions that hardware manufacturers tend to offer if they offer anything at all), most people would probably stay with that because putting anything else on it would be more of a hassle than it was worth.

Free is too expensive (Economist)

khim — Tue, 10 Apr 2012 21:57:12 +0000

This is exactly what I'm saying. It's not up to the Microsoft to determine if given country has high piracy or not. But in both cases Windows wins: if piracy is high the Windows wins because of network effect, if it's low then Microsoft makes good money and spends some of it via kickbacks to promote Windows. Windows wins in both cases. Q.E.D.

Free is too expensive (Economist)

dlang — Tue, 10 Apr 2012 21:12:01 +0000

where linux and windows are both 'free', windows will win due to the network effect.

but if microsoft didn't allow for the piracy of windows to maintain this, they would not both be free.

Free is too expensive (Economist)

khim — Tue, 10 Apr 2012 20:43:58 +0000

These both are wrong questions. The answers to both are: there are more then enough computer manufacturers and they represent huge slice of computer industry (China alone is large enough), but in places where Windows is free and Linux is free Windows wins hands down. It's not even a contest. And since you can not attach any non-zero price to Linux (this will immediately make your Linux offer non-noncompetitive) the end result is that there are still no chance for “serious try”.

People somehow expect that “serious try” will be “identical push for Windows and Linux” but it just does not work: Windows in incumbent, identical push will always favor it, you need bigger push to succeed - and where money for said push will come from?

You can not even use typical bundling strategy (where producer of demo version of commercial program pays you dollar or two) because Linux distributions are typically designed to repel any and all proprietary commercial developers as we are discussing here.

Free is too expensive (Economist)

khim — Tue, 10 Apr 2012 20:35:49 +0000

There is nothing in principle to prevent someone like HP or Dell from offering Linux als a pre-installed alternative on all of its machines other than that the margins you get from moving boxes do not lend themselves to experiments.

Ever wondered why systems with Linux preinstalled are always separate models (often with the same hardware but still with separate nomenclature article)? Apparently there are just this little teeeny agreement.

The problem is to get things going in the first place; once you have a system set up, there is no reason why supporting Linux would need to be any more expensive than supporting Windows.

Not enough. You need bigger margins, not the same margins, or else the whole exercise is pointless. For that manufacturer need some kind of lock-in - that's what I'm talking about. And indeed when nettop story started vendors tried to produce such lock-in - but unsuccessfully.

In the long run it may even be cheaper.

PC business is very low margin business. Companies just don't have luxury to think about long term: if they'll start producing losses then the end can come very fast.

»Lock-in« doesn't enter into it from a hardware manufacturer's point of view because nobody is »locked into« generic PC hardware that you can get from dozens of manufacturers.

Sure. But why start expensive and complex program which may jeopardize your relationship with Microsoft if the end result are the same tiny margins you already have?

Free is too expensive (Economist)

dlang — Tue, 10 Apr 2012 18:28:13 +0000

and how many computer manufacturers build systems exclusively for those markets?

and what slice of the overall computer industry are these markets?

I suspect that the answers to both are very small numbers.

Free is too expensive (Economist)

Cyberax — Tue, 10 Apr 2012 17:29:58 +0000

I often hear laments: "Oh, if only Windows wasn't preinstalled. We'd rule the world with Linux, surely!"

They are definitely not true as I see a counter-example with my own eyes.

>It certainly doesn't prove Linux couldn't compete with Windows on a level playing field. What is easier to come by for non-geeks in Russia and Ukraine, a CD with pirated Windows or a CD with Linux?

A CD with Linux - broadband access is ubiquitous and cheap. I have 100mbit Ethernet connection for $8 a month, so I can download Ubuntu install CD in about 30 seconds. Of course, I can do the same for a Windows install DVD.

So I'd say they are on a level footing.

>What about CDs with pirated Windows games?

Actually, they are becoming quite rare. Why would you bother with buying CDs when you can download anything you want for free?

Free is too expensive (Economist)

anselm — Tue, 10 Apr 2012 17:14:06 +0000

OK. So Microsoft doesn't waste money on kickbacks to manufacturers in countries where people won't pay for Windows in the first place. What does that prove?

It certainly doesn't prove Linux couldn't compete with Windows on a level playing field. What is easier to come by for non-geeks in Russia and Ukraine, a CD with pirated Windows or a CD with Linux? What about CDs with pirated Windows games?

Also, Microsoft could easily crack down on pirated copies of Windows if they wanted to. However, that would be utterly counterproductive because it would just make Linux look more attractive in comparison. Microsoft knows very well that hanging on to their 95% market share is worth a bunch of pirated CDs in Russia.

Free is too expensive (Economist)

Cyberax — Tue, 10 Apr 2012 16:48:03 +0000

Yeah, sure.

Microsoft doesn't provide kickbacks in Russia, Ukraine and lots of other countries. You can easily buy laptops and computers running FreeDOS there. Guess what's the first thing people do after they buy them?

Hint: it generally involves a CD with pirated Windows.

Free is too expensive (Economist)

anselm — Tue, 10 Apr 2012 15:07:54 +0000

This will never be “seriously tried” because of poor ROI: if there are no lock-in then how can you recoup your expenses?

If you're a hardware manufacturer such as HP or Dell you're not interested in lock-in – you're interested in moving boxes. Anything that looks like it might move more boxes is worth looking into. The problem is to get things going in the first place; once you have a system set up, there is no reason why supporting Linux would need to be any more expensive than supporting Windows. (In the long run it may even be cheaper.) »Lock-in« doesn't enter into it from a hardware manufacturer's point of view because nobody is »locked into« generic PC hardware that you can get from dozens of manufacturers. Even if you're into high-end kit you can always get that from any of half a dozen manufacturers. (Apple is an anomaly here because Apple is no longer a computer maker, it's a life style. People are hooked on Apple in a way that they never get hooked on Dell or Asus. However, to a large extent OS X suffers from the same uptake problems as Linux; it just doesn't matter because Apple makes three quarters of its money selling stuff that isn't running OS X in the first place, anyway.)

There is nothing in principle to prevent someone like HP or Dell from offering Linux als a pre-installed alternative on all of its machines other than that the margins you get from moving boxes do not lend themselves to experiments. Microsoft has a nice little positive-feedback loop going that will ensure that as long as 95% of PC buyers buy Windows pre-installed, the hardware makers aren't keen on being the first to sink investment money into something new – and Microsoft is trying to keep things that way: The real reason Linux isn't preinstalled more is that Microsoft provides kickbacks to hardware manufacturers like HP or Dell to »recommend« Windows.

Free is too expensive (Economist)

khim — Tue, 10 Apr 2012 14:34:24 +0000

A computer manufacturer could do a lot worse than team up with somebody who will provide support for one of the major long-lifecycle distributions like Debian or Ubuntu (LTS). That would ensure a reasonable time between upgrades (which are usually seamless) as well as timely security patches, and a wide, easily-accessible selection of software from the get-go. Also these distributions are unlikely to go away anytime soon. So far this hasn't been seriously tried AFAIK.

This will never be “seriously tried” because of poor ROI: if there are no lock-in then how can you recoup your expenses? This is the case where Linux's greatest strength becomes it's greatest weakness.

Of course if the prime purpose of the computer in question is to run the newest Windows games, a Windows machine is likely to be the better bet – but that isn't going to change however much Linux is modified. It may not actually be worth the trouble.

Of course direct attack is hopeless! Linux must do something which MacOS and Windows just can't do and then grow from such niche to the full-blown desktop. There are some ideas about what exactly this niche can be - different companies play with different niches.

The problem is that it looks like all such attempts will happen with something like Android or webOS: Linux which has nothing to do with traditional Linux desktop.

Free is too expensive (Economist)

dlang — Tue, 10 Apr 2012 10:57:23 +0000

the distros selected for the early netbooks were not any of the bigger name, well supported distros. They were ones that virtually nobody had heard of, that had no significant community around them, and then the vendors provided no updates for them on top of that.

That's hardly justification for saying that such installations will require "a divorce from the world of traditional Linux distributions", In fact, it's more a matter of showing that if you do make such a divorce, it's going to take a lot more effort on your part to make things work.

Free is too expensive (Economist)

anselm — Tue, 10 Apr 2012 10:42:10 +0000

But it looks these preparations will require complete divorce from the world of traditional Linux distributions. Which is sad because said distributions did many things right.

A computer manufacturer could do a lot worse than team up with somebody who will provide support for one of the major long-lifecycle distributions like Debian or Ubuntu (LTS). That would ensure a reasonable time between upgrades (which are usually seamless) as well as timely security patches, and a wide, easily-accessible selection of software from the get-go. Also these distributions are unlikely to go away anytime soon. So far this hasn't been seriously tried AFAIK.

It is important to emphasise that these distributions already come with lots of software that people would otherwise have to obtain, possibly at very considerable expense, from the open third-party market or an »app store«. So it's not as if one would immediately need lots of buy-in from third-party software developers (as in OS X). Certainly somebody who uses their computer mainly to surf the web, to write e-mail and letters, even to deal with holiday photographs, to catalogue books or DVDs or a stamp collection, and many other things that occur in the usual home or SOHO use, could go a very long way without having to install anything from outside the distribution's repositories.

Of course if the prime purpose of the computer in question is to run the newest Windows games, a Windows machine is likely to be the better bet – but that isn't going to change however much Linux is modified. It may not actually be worth the trouble.

Free is too expensive (Economist)

khim — Tue, 10 Apr 2012 09:20:22 +0000

I would go out on a limb a bit and say that Linux is probably 100x as popular as the next closest competitor if you were to exclude pre-loaded OS installs.

Yup. Linux desktop is large fish in this aquarium. But there are the problem: master of said aquarium can displace it at any time. The question: can Linux desktop survive in the ocean?

and no, I don't consider the "distro that nobody has heard of" shipped pre-loaded on the early netbook machines to qualify as Linux being pre-loaded.

Why not? It was the first time Linux desktop was seriously pitted against older contenders. It was eaten alive. I hope people who'll try to do that next time will be better prepared. But it looks these preparations will require complete divorce from the world of traditional Linux distributions. Which is sad because said distributions did many things right.

Free is too expensive (Economist)

dlang — Tue, 10 Apr 2012 00:08:57 +0000

Is there any OS that doesn't come pre-loaded on systems that has achieved any noticeable market penetration?

I would go out on a limb a bit and say that Linux is probably 100x as popular as the next closest competitor if you were to exclude pre-loaded OS installs.

That says a lot of good things about Linux, and you have to wonder where it would be if it wasn't for the Microsoft shenanigans.

and no, I don't consider the "distro that nobody has heard of" shipped pre-loaded on the early netbook machines to qualify as Linux being pre-loaded.

Free is too expensive (Economist)

anselm — Mon, 09 Apr 2012 23:03:33 +0000

Yet Mac OS X captured more than 10% of the market, even though it requires overpriced proprietary hardware.

Yes, but that's not because people specifically want OS X and therefore need to buy a Mac, it's because they want a Mac in the first place, which is a different ballgame altogether. People would buy Macs even if they came with a GUI version of CP/M, as long as there was a big-enough Apple logo on the computer. This is because for many people, buying Apple stuff is a life style decision, rather than a technical decision. Sort of like being vegetarian.

If Apple wasn't manufacturing its own PCs and OS X was an after-market OS to be installed on generic PCs like Linux is today, it would be just as (un)popular as Linux, simply because most people can't be bothered to change the OS on their computer. It would also have the same hardware support/developer buy-in issues, only worse because there would be less free stuff available and fewer hardware manufacturers would be interested in supporting with drivers or using it for their own products.

Free is too expensive (Economist)

rqosa — Mon, 09 Apr 2012 21:14:52 +0000

> or will continue as a small 'niche' product.

There's nothing wrong with that, as far as I'm concerned. KDE has always been a "niche product", and yet it has survived for over 15 years and still has a thriving developer community.

Free is too expensive (Economist)

paulj — Mon, 09 Apr 2012 16:25:55 +0000

Sorry, but this is wrong. OPIE and GPE only had platform to run because Sharp created Linux-based PDA.

Sorry, but this is wrong. DEC^WCompaq Western Research Lab had been working on SA1100 StrongArm prototype hand-helds well before the Zaurus, with Linux.The Compaq iPaq was borne out of the Itsy work and, though it shipped with WinCE, Compaq WRL provided Linux friendly bootloader firmware and distributions, which were pretty easy to install. The only daunting step was running the WRL provided WinCE app to reflash the firmware - still easy though. DEC^WCompaq WRL remained a nexus of the Linux StrongArm handhelds community for a long time after (handhelds.org remained hosted there for years after).

The Sharp Zaurus StrongArm Linux devices came after the Compaq iPaq. Also, they were, I think, harder to find. Compaq iPaqs were in a lot of shops at the time.

Free is too expensive (Economist)

Cyberax — Mon, 09 Apr 2012 14:34:56 +0000

But that's exactly what khim and me is telling. Good switchable GPUs are a hard task to implement.

So vendors simply don't bother with Linux where it'll be useful only for a fraction of 1% of their users. NVidia hasn't even ported their Optimus technology to Linux in proprietary drivers.

In the area of switchable GPUs all we get is airled. And while he's a mega-super-developer, he can only do so much.

Free is too expensive (Economist)

nix — Mon, 09 Apr 2012 08:44:34 +0000

That's not because they're being evilly kept secret by nasty hardware manufacturers trying to destroy desktop Linux. It's because switchable GPUs is hard enough when they're *not* completely different GPUs with distinct drivers. Even the first case has only been working for a year or so.

(And the people working on these free drivers are funded by... AMD and Intel! Normally, you'll note, competitors.)

Free is too expensive (Economist)

Cyberax — Mon, 09 Apr 2012 01:40:59 +0000

Care to explain how can I use switchable GPUs (ATI and Intel - both officially supported) on my Sony VPCSE?

Right now I have to blacklist radeon driver, or it simply hangs with black screen.

Exactly...

Cyberax — Mon, 09 Apr 2012 01:03:25 +0000

You can have it. Ubuntu works fine in a chroot on my Galaxy S.

Go on, install it and use it.

Free is too expensive (Economist)

Cyberax — Sun, 08 Apr 2012 23:22:55 +0000

>Because the phone carriers are hostile to user-freedom, and also because Plasma Active is new and immature. It's not true, though, that "nothing" runs Plasma Active, and the amount of devices it supports will increase over time. And it's not even the only current FLOSS mobile UX; there's Nemo Mobile, and there's CyanogenMod.

Care for a prediction? Plasma Active won't run on more than a handful devices with by the end of 2013. It will be used by relatively few users, mostly computer geeks. Then it'll either slowly wither away and die or will continue as a small 'niche' product.

Free is too expensive (Economist)

rqosa — Sun, 08 Apr 2012 19:40:10 +0000

Oh, and:

> Well, the history repeats itself with GPU, at least.

Not so much; right now there are only three major desktop GPU manufacturers (Intel, AMD, nVidia), all of which have free drivers available for all GPU variants up to almost the newest ones, and the latter two manufacturers also have proprietary drivers for Linux. And for mobile GPUs, proprietary drivers for Linux are readily available, and work is underway on free drivers for one mobile GPU family/manufacturer.

Free is too expensive (Economist)

rqosa — Sun, 08 Apr 2012 19:20:31 +0000

Also, one more thing:

> > It's one thing to circumvent the bootloader and few unique components. It's another thing to port Linux to the hardware which has no public specification and which was never designed with Linux in mind.

What about Rockbox, then? It was quite successful on hardware that was never meant to run it or any other OS/firmware other than the manufacturer's own one, and it only declined because (same as with PDAs) the smartphone/tablet boom has decreased the usage share of DAP devices.

Free is too expensive (Economist)

rqosa — Sun, 08 Apr 2012 17:47:44 +0000

> Which basically implies that people who don't want to learn how to build Linux systems and care for them should be considered defective and don't deserve lenience.

It implies nothing of the sort.

> It's one thing to empower people by giving them access to human knowledge.

That is what I've been saying all along: if people don't have sufficient knowledge about the technologies they depend on, they are disempowered. And things like locked-down hardware and source-unavailable software have the effect of disempowering people by excluding them from access to this knowledge. Therefore, what you said earlier — locked-down hardware is "good for them!" — can't be true (and is totally contrary to the core ideals of the FSF and the FLOSS community at large).

> They failed in the sames sense Linux desktop has failed.

The way I see it, the Linux desktop is successful today and getting better all the time — and there's no reason why it has to be used by the majority to be "successful".

(There's an old saying: "Unix is user-friendly, it's just selective about who its friends are." It was true then, and it's still true now.)

> When Sharp switched to Windows CE itself in 2007 OPIE and GPE lost the momentum, too.

Like I already said, the real reason why OPIE and GPE lost momentum was because the whole "PDA" device class was supplanted by smartphones and tablets.

> It's one thing to circumvent the bootloader and few unique components. It's another thing to port Linux to the hardware which has no public specification and which was never designed with Linux in mind.

The Windows CE devices were never designed with Linux in mind, and yet OPIE and GPE ran on them.

And for another example: the BeagleBoard / PandaBoard / RaspberryPi / IGEPv2 class of devices probably would never have existed if it weren't for both:

The rise of the smartphone / tablet market that began 4-5 years ago (because these devices use CPUs and GPUs that are primarily sold to smartphone/tablet manufacturers), with most such smartphones / tablets not designed with Linux in mind until Android took the lead in marketshare; and
The existence of enough people who do care about having fully-programmable hardware devices.

What this tells us is: if enough people demand computing devices that let the user (= owner) have full control over them, then the manufacturers will meet the demand with devices made from whatever commodity hardware components are currently on the market. That's why the FLOSS community must convince as many people as possible that unlocked hardware is a desirable thing (and this doesn't have to be a majority of people, it just has to be enough for the manufacturers to take notice).

> This way was tried and it just does not work.

It did work — that's why we've got the PandaBoard et al., and the Nexus series, and probably also why the threat posed by EFI was defeated (at least it looks that way currently).

> This means that FOSS long-term survival is guaranteed only if FOSS community will learn to create toys used by general public. If they will be threatened then you you can mobilize millions if not billions in a case of danger.

The problem is that they won't be threatened (or at least they won't perceive any threat) by the unlocked hardware going out of production or the destruction of the FLOSS ecosystem. If the Nexus and whatever other unlocked Android devices all were discontinued today, the majority of Android users would barely even notice. You said so much yourself: "Of course not! It's good for them! This is what they want!"

Free is too expensive (Economist)

rqosa — Sun, 08 Apr 2012 16:30:49 +0000

> as time goes on it becomes harder to install Linux on PC, not easier.

Not in my experience. I first tried to install Linux on a PC around 1997 or 1998, and couldn't do it. Since then it's gradually gotten easier; the last few times I've installed Linux (most recent one was this past December or January), I had no trouble whatsoever.

> Either Linux desktop will finally reach general consumer or it'll die off.

That's pure FUD, nothing more.

Free is too expensive (Economist)

khim — Sun, 08 Apr 2012 11:26:00 +0000

As someone who was involved in the handhelds.org community for a long time and the current maintainer of Opie (yes, it's still barely alive) I feel that Opie and GPE are being bandied about here as if they have significant relevance to the discussion at hand.

Well, yes, is is.

It was a few years ago now but after working for some time on handheld Linux I came to the unpleasant realisation that Opie, GPE and the Linux-based operating systems that they ran on where never, ever going to reach the masses. It was never going to happen.

Hmm, that's my point exactly.

Because they never came pre-installed mass-market devices (among many reasons why not, at the time, GPL was a problem for many companies) and getting them onto existing devices was an exceedingly difficult and risky procedure even for the moderately competent - much more difficult than installing Linux on a PC.

Sure. But here is the problem: as time goes on it becomes harder to install Linux on PC, not easier. Not just things intended to close the ability to install Linux totally (like Secure Boot) - there are other efforts, too. These changes are slow because when they interfere with lives of general public general public pushes back, but the process is quite steady.

Should we want till Linux desktop will reach the same stage as OPIE today? Or, perhaps, we need to do something to make sure it'll never happen.

Note that even the reason which kept Linux niche open for years (you need some Linux-compatible hardware to develop server solutions) is no longer valid: Virtual PC works fine for that.

Unlike PCs, the hardware was almost completely closed and differed for almost every new device, and we couldn't keep up.

Well, the history repeats itself with GPU, at least.

As Cyberax said: there's that sense of fin-de-siècle in the air - the current situation is unsustainable and Something Has To Happen. Either Linux desktop will finally reach general consumer or it'll die off. And the more I look on the situation the more likely it looks like we'll have both (like it happened on handhelds/mobiles): we'll get some kind of mainstream “Linux desktop”, but it'll be some kind of deep fork which will ignore most of the efforts which happened before it. Current distributions then follow the OPIE/GPE lead on the road to oblivion.

Free is too expensive (Economist)

BlueLightning — Sun, 08 Apr 2012 10:54:58 +0000

As someone who was involved in the handhelds.org community for a long time and the current maintainer of Opie (yes, it's still barely alive) I feel that Opie and GPE are being bandied about here as if they have significant relevance to the discussion at hand. There are some parallels, but the situation was entirely different.

It was a few years ago now but after working for some time on handheld Linux I came to the unpleasant realisation that Opie, GPE and the Linux-based operating systems that they ran on where never, ever going to reach the masses. It was never going to happen.

Why not?

Because they never came pre-installed mass-market devices (among many reasons why not, at the time, GPL was a problem for many companies) and getting them onto existing devices was an exceedingly difficult and risky procedure even for the moderately competent - much more difficult than installing Linux on a PC. Unlike PCs, the hardware was almost completely closed and differed for almost every new device, and we couldn't keep up. Not to mention that building an OS for end-users for a mobile device was a gargantuan task for a group with fairly limited resources. The saddest thing of all though is that ultimately the effort was stymied by politics.

However, I wouldn't say the effort was a complete failure. We got a lot of real software development done, and out of the desire to be able to build an operating system grew the OpenEmbedded project, which flourished and has enjoyed commercial success that still continues to this day. Not to mention that developers who worked on various projects around handhelds.org had a lot of fun and learnt a great deal (myself included). This isn't particularly relevant to the desktop Linux discussion at hand, but worth noting.

Free is too expensive (Economist)

khim — Sun, 08 Apr 2012 10:11:59 +0000

If other people were actively excluded from joining that group (as is the case when copyright and/or patents and/or lack of source code exclude people who do know how to program from being able to modify the software that they use), then it might.

What this has to do with discussion in question?

What you dismiss as "crazy declarations" are nothing less than the core ideals that the 18th century French and American revolutionaries believed in.

Rilly? You must know how to program or you are not human are ideas of French and American revolutionaries? News to me.

I've said:
    Linux breaks applications all the time.
You answered:
    But you can still run them, as I've already said…
The next step was:
    I can run them. You can run them. Joe Average can not - and that's the problem.
Which prompted this crazy response:
    In that case, the real problem here is that "Joe Average" isn't computer-literate enough.
Which basically implies that people who don't want to learn how to build Linux systems and care for them should be considered defective and don't deserve lenience.

This is far cry from the “core ideals that the 18th century French and American revolutionaries”. It's one thing to empower people by giving them access to human knowledge. It's another thing to disqualify people by demanding them to learn things they don't really need or want.

It's way too soon to say that Plasma Active and Nemo Mobile are "failed".

They failed in the sames sense Linux desktop has failed. They don't come preinstalled (and will not come preinstalled in the future), they don't influence the markets they are in (hardware is designed to support Android 2.x or Android 4.x, never to support Plasma Active or Nemo Mobile), etc. The most they can hope for is something like Zaurus: niche product which will be on market for a few years mostly unnoticed and which will be later replaced with Android (or may be Windows8/9/10). They may survive as “curiosity project” like XMBC but this is side-attraction at best, this is not where future direction of the society is determined.

If that were true, then it would have been impossible for OPIE and GPE to run on hardware made for Windows CE (Jornada / iPAQ), but they did.

Sorry, but this is wrong. OPIE and GPE only had platform to run because Sharp created Linux-based PDA. And earlier efforts were also driven by companies, not by FOSS community. The same hardware was used for Windows CE devices thus it was an easy port (initially OPIE only supported Zaurus). When Sharp switched to Windows CE itself in 2007 OPIE and GPE lost the momentum, too. It's one thing to circumvent the bootloader and few unique components. It's another thing to port Linux to the hardware which has no public specification and which was never designed with Linux in mind.

And it's not guaranteed that Android being successful will ensure that unlocked hardware will be available in the future — for a while it seemed like there would be no more unlocked Android phones, when the Nexus One was cancelled.

That's separate issue. But if your hardware is using Linux-friendly components then to have free OS on it you basically only need to circumvent the bootloader. If your hardware is designed for totally different OS from the ground up then it's much, MUCH, MUCH harder.

iOS is descended from FLOSS (Mach and 4.3BSD), and yet it has no "sibling platform for FOSS-lovers".

iOS is only used by one producer which is quite explicitly is not interesting in filling all the niches. And it you can install Linux (Android) on iPhone - but it works significantly worse then Linux on Android handsets.

The only real solution is for there to be a large enough niche market of people who actively prefer unlocked hardware, regardless of whether it's desktop or mobile.

Bullshit. It just does not work. This approach was tried many times (Zaurus, OpenMoko, Nokia's Maemo/Meego efforts, etc). This niche market is just too small. It's large enough to support creation of a few devices from the components used by mainstream, but it's not large enough to support it's own separate ecosystem.

That's why it's crucially important to make the case to the public at large about the benefits of user-freedom (and in particular the ways that locked-down hardware restrict it).

“Doing the same thing over and over again and expecting different results”? That's definition of insanity. This way was tried and it just does not work.

Time to move on: accept that public at large is just too ignorant to care about software freedom… and coopt it anyway. Internet community did that beautifully when it was threatened by SOPA/PIPA: general public don't care about copyright all that much (mostly because it's too ignorant about copyright-relevant issues), but it reacts when confronted with the danger of loss of their favorite toy.

This means that FOSS long-term survival is guaranteed only if FOSS community will learn to create toys used by general public. If they will be threatened then you you can mobilize millions if not billions in a case of danger. If FOSS will be used only by some FOSS-lovers then the destruction of the whole ecosystem will just not be noticed by general public.

FOSS community may be powerful, but it has an Achilles heel: ultimately it needs hardware to run on and said hardware can only be created by large companies. It is just as stupid to pretend that it's not important as it is to stupid to pretend that FOSS is powerless.

Free is too expensive (Economist)

rqosa — Sun, 08 Apr 2012 07:31:52 +0000

> does it mean they are “small elite with have power over others”?

If other people were actively excluded from joining that group (as is the case when copyright and/or patents and/or lack of source code exclude people who do know how to program from being able to modify the software that they use), then it might. Because if society depends on something that is under the control of an exclusive elite group, they will have political power over the whole society. That's why the core goal of FLOSS has always been to increase the amount of people who have control over the software they use. (And that's also why even non-programmers stand to benefit from FLOSS — with a large and non-exclusive base of developers, non-developer users have more options to turn to when developers/maintainers go against the users' interests.)

> Whatever. You may as well declare Law of Gravity as something “inherently wrong” - it'll not care. Just like I don't care about your crazy declarations.

What you dismiss as "crazy declarations" are nothing less than the core ideals that the 18th century French and American revolutionaries believed in.

> Yes, there are enough failed FOSS projects and I'm sure there will be many more.

It's way too soon to say that Plasma Active and Nemo Mobile are "failed". I believe they'll be more successful than the ones you mentioned (OPIE and GPE) ever were — mainly because the hardware they're designed for is itself far more commercially successful than the old so-called "PDA" devices ever were.

> CyanogenMod is the only project with clear long-term perspective.

It has no more "long-term perspective" than the rest. The only reason it has a larger usage share than the others is from riding on the coattails of it's "evil twin".

> If Android will fail at some point (for example if WP15 will kill it) then CyanogenMod, Plasma Active and other simlar projects will have no hardware to run on.

If that were true, then it would have been impossible for OPIE and GPE to run on hardware made for Windows CE (Jornada / iPAQ), but they did. And it's not guaranteed that Android being successful will ensure that unlocked hardware will be available in the future — for a while it seemed like there would be no more unlocked Android phones, when the Nexus One was cancelled.

> It's either that or nothing at all. If Linux will form the platform which is used by Joe Average then there will be sibling platform for FOSS-lovers.

Again, there's no guarantee of that. iOS is descended from FLOSS (Mach and 4.3BSD), and yet it has no "sibling platform for FOSS-lovers". And there's no reason why a (desktop or mobile) OS based on GPLv2-licensed Linux couldn't be just the same — indeed, Android could easily become like that if Google and the device manufacturers chose to stamp out all unlocked hardware.

The only real solution is for there to be a large enough niche market of people who actively prefer unlocked hardware, regardless of whether it's desktop or mobile. (And I believe that Google, for the moment at least, understands that there is such demand for unlocked hardware, or else there never would have been the Nexus product line.) That's why it's crucially important to make the case to the public at large about the benefits of user-freedom (and in particular the ways that locked-down hardware restrict it).

Free is too expensive (Economist)

BlueLightning — Sat, 07 Apr 2012 20:56:24 +0000

You forgot about webOS which is open source now.

Not yet it isn't...

Free is too expensive (Economist)

khim — Sat, 07 Apr 2012 20:26:36 +0000

Because if there's only a small elite who are able to develop software, that small elite will have power over others — and people having power over others is inherently wrong.

Small? Elite? What are you smoking? Very small percentage of people know how to farm today (and food is essential for living!) - does it mean they are “small elite with have power over others”? Very small percentage of people know how to stitch boots (and in many countries you can literally die without proper boots!) - does it mean they are “small elite with have power over others”?

Contemporary society is highly differentiated and any given skill (beyond small number of basics like the ability to speak) is only known to small percentage of it. Why programming should be any different?

On the contrary: I've worked as CS teacher some time ago and it's obvious to me that most people will never be able to program. Never. That's just fact of life. You can not do anything about it. You can ignore these people or you may adopt them somehow, but the society where most people know how to program is just impossible.

There can be no argument whatsoever that it's not inherently wrong — it's a core value that needs no justification.

Whatever. You may as well declare Law of Gravity as something “inherently wrong” - it'll not care. Just like I don't care about your crazy declarations.

It's not true, though, that "nothing" runs Plasma Active, and the amount of devices it supports will increase over time. And it's not even the only current FLOSS mobile UX; there's Nemo Mobile,

You forgot about webOS which is open source now. Yes, there are enough failed FOSS projects and I'm sure there will be many more. People just refuse to learn.

and there's CyanogenMod.

This is different kettle of fish. CyanogenMod is the only project with clear long-term perspective. Because it has evil twin designed for Joe Average - regular Android. If Android will fail at some point (for example if WP15 will kill it) then CyanogenMod, Plasma Active and other simlar projects will have no hardware to run on.

You seem to be suggesting that this is what the future of the Linux desktop should look like — but the current user-base and developer-base want nothing like that.

s/should/would/

It's either that or nothing at all. If Linux will form the platform which is used by Joe Average then there will be sibling platform for FOSS-lovers. If Linux will continue to form 1% of desktop and Microsoft will succeed in separation of closed Windows-only desktop platform from server (where Linux is not in danger for foreseeable future) then Linux desktop will be extinct.

Exactly...

man_ls — Sat, 07 Apr 2012 19:00:48 +0000

Thanks for your post, it sums up my thoughts nicely.

Do not yield to the flower of the day, be it mobile or phone. I would very much prefer having debian on my phone than android on my desktop!