You're missing the core point though, which is that for the vast majority of end-users, OS is irrelevant. People don't actually care about it, they don't care about whether or not they can study and modify the source code; the one item that the open source communities value the most is something that most people don't actually give a flying one about.

Android is a great example of this. I see figures of something like 500 million devices, but of those, what percentage of users actually bought into it because of a Linux kernel? I'll give you a hint - it's less than 1. The important criteria for a smart phone are something more like: can it make calls? Can it send SMS? Can it take photos? Can I sync it with my email? Can I browse the web on it? Can I play Angry Birds on it? Answer "yes" to those and it doesn't matter if it runs on Unix, Windows or magic jellybeans - you've got a sale.

Android is not a victory for Unix; it's a victory for the applications and services provided by the platform, and if they weren't there it would have crashed and burned. None of this is about Unix versus Windows versus iOS versus whatever tomorrow's flavour of the month is; it's all about the applications and services you give to the user, and those run on the OS, they are not the OS itself.

Let me get this straight. So android isnt a "victory" for unix, everything that runs on ARM devices either, top 500 computers either, then what on Earth would you consider a "victory" for an OS? By your own terms, no OS is successful because no one cares about OSes, not Windows, nor iOS, nor anything.

A victory for an OS is to be the preferred platform to use when developing something, because that is what it is, a platform, what makes the hardware usable. And if being everywhere and used by everyone on every kind of hardware isn't a "victory", then what you would consider a victory? I just don't get it. What makes other OS for you more successful if you stated that OSes don't matter for the user?

You seem to dismiss everything on the basis that "no one cares", but that way it would be the same if we talked about Windows, Linux, Darwin, OS/2 or whatever thing out there that the end user doesn't sees directly. And if you're measuring success on a field where every single thing fails, maybe you should consider another field for comparison altogether.

Unless that is you wan't OS developers try to make the end user care about their OS that is...

End-users don't care about their OS though - that is the thing. They care about what the OS lets them do, but the OS itself and as an OS - nope.

That makes Android a successful implementation of a modified Linux kernel with a decidedly non-Unixy front-end on it, but it shouldn't be read as being successful because it's Unix; it's successful because that non-Unixy front-end lets them do the smartphone-type stuff they want to do, and because Google did an awesome job of marketing it.

[quote name='mhagain' timestamp='1352051415' post='4997245']
You're missing the core point though, which is that for the vast majority of end-users, OS is irrelevant. People don't actually care about it, they don't care about whether or not they can study and modify the source code; the one item that the open source communities value the most is something that most people don't actually give a flying one about.

Android is a great example of this. I see figures of something like 500 million devices, but of those, what percentage of users actually bought into it because of a Linux kernel? I'll give you a hint - it's less than 1. The important criteria for a smart phone are something more like: can it make calls? Can it send SMS? Can it take photos? Can I sync it with my email? Can I browse the web on it? Can I play Angry Birds on it? Answer "yes" to those and it doesn't matter if it runs on Unix, Windows or magic jellybeans - you've got a sale.

Android is not a victory for Unix; it's a victory for the applications and services provided by the platform, and if they weren't there it would have crashed and burned. None of this is about Unix versus Windows versus iOS versus whatever tomorrow's flavour of the month is; it's all about the applications and services you give to the user, and those run on the OS, they are not the OS itself.

Let me get this straight. So android isnt a "victory" for unix, everything that runs on ARM devices either, top 500 computers either, then what on Earth would you consider a "victory" for an OS? By your own terms, no OS is successful because no one cares about OSes, not Windows, nor iOS, nor anything.
[/quote]

Well, the kernel might be a modified version of the Linux kernel but nothing that is visible to the end user has anything to do with Linux. You got a useful UI in Android where you can do everything you need the O/S to do which is exactly the opposite of most, if not all, Linux distributions.

Well, the kernel might be a modified version of the Linux kernel but nothing that is visible to the end user has anything to do with Linux. You got a useful UI in Android where you can do everything you need the O/S to do which is exactly the opposite of most, if not all, Linux distributions.

I know, I'm not saying that the user cares or should care. All I am saying is why measure the success of something by saying "x group doesnt cares about y, thus is a fail" when the "x group" arent't the ones supposed to care?

You should measure OS success by asking about the people who actually deals with the OS. Hardware manufacturers, device designers, driver developers, etc. It shouldn't be surprising that Linux (or any other OS) "fails" to matter when its put against the end user, which is exactly the group of people who shouldn't care about that stuff.

The user doesn't cares about Linux. Yeah. So what its the point of that? Why is that considered a "fail" for Linux? The end user doesn't develops on top of Linux so why they should care anyway? It's a void statement. Like saying that cars fail because they can't fly. Well, they never were intended to fly in the first place, it may be an awful plane but it's a good car nevertheless.

I think you're showing some binary thinking there - it's not a success for Linux qua Linux (by which I mean the common understanding of "GNU/Linux" the OS), but that doesn't automatically make it a fail either. What it definitely is however is a great example of one potential way forward - take the solid underpinnings and put something on top of them that actually is important and attractive to end users. So it's some kind of weird netherworld which is neither success nor failure, nor any point inbetween really - you don't see Android users going around using vi to edit .conf files, or fighting over which is better: KDE or Gnome, so they are not Linux users in the traditional sense but they are most definitely using the Linux kernel, even if they may not know it nor care about it.

It's clear that there are two different markets being discussed here; one is the Android market where we do have a successful implementation of the kernel + a new front-end, the other is the trad desktop market where by all standard criteria Linux can be said to have failed (although whether the kind of success this is measured by was ever an objective is another matter entirely).

These types of threads amuse me; people cheering the 'death' of microsoft and the rise of other platforms.... or as I like to think of it 'lolol fragmentation' because that's what you are looking at here.

And the amount of work and debugging which goes with it.

This is both desktop and mobile space too; Android devices of varying quality and spec seem to be vomited out at every turn and Apple seem to push hardware 'forward' at a faster rate (two iPads in a year, iPad mini and iPhone 5) with varying quality of OS updates to go with it.

At times my head hurts at the prospect of creating a renderer for 6 fixed hardware devices and iOS - god alone knows how this fragmentation is going to effect things.

I'm starting to think this is going to cause a slow down in the software industry more than anything and we'll end up replacing MS with Valve, Epic and Unity Technology being the only way independant developers can afford to push forward.

Competition might be good but mass fragmentation of the market certainly isn't...

Steam is supposedly working on their own linux-game-platform-computer-thingy, no?

I think you're showing some binary thinking there - it's not a success for Linux qua Linux (by which I mean the common understanding of "GNU/Linux" the OS), but that doesn't automatically make it a fail either. What it definitely is however is a great example of one potential way forward - take the solid underpinnings and put something on top of them that actually is important and attractive to end users. So it's some kind of weird netherworld which is neither success nor failure, nor any point inbetween really - you don't see Android users going around using vi to edit .conf files, or fighting over which is better: KDE or Gnome, so they are not Linux users in the traditional sense but they are most definitely using the Linux kernel, even if they may not know it nor care about it.

It's clear that there are two different markets being discussed here; one is the Android market where we do have a successful implementation of the kernel + a new front-end, the other is the trad desktop market where by all standard criteria Linux can be said to have failed (although whether the kind of success this is measured by was ever an objective is another matter entirely).

Hm... Fair enough. Though in my view, the fact that Android users aren't messing around with .conf files is a good thing, since what is a Linux user shouldn't be defined by how much you have to twist the OS it to make it work (as long as you're free to do so if you want of course). Besides, while Android is pretty popular, and the PC Linux power user may be the "traditional user", there are a LOT of less popularized uses for Linux kernel, specially since its the OS of choice to run ARM stuff on (and I'm not talking about smartphones). The idea that a Linux user should be a living xorg.conf parser for him to be a "real" Linux user is pretty silly.

Steam is supposedly working on their own linux-game-platform-computer-thingy, no?

Yep. Its going to be released along Half Life 3.
[spoiler]Joke of course [/spoiler]

Well, editing .conf files was just an example off the top of my head to illustrate a point, and was not intended to be taken so literally.

I'm actually quite excited by Valve's Linux plans myself, and wish them huge luck with it all, even though it's of no personal relevance to me. Antipathy to a platform should not be confused with opposition to it, I guess.

[quote name='MA-Simon' timestamp='1352184641' post='4997903']
Steam is supposedly working on their own linux-game-platform-computer-thingy, no?

Yep. Its going to be released along Half Life 3.
[spoiler]Joke of course [/spoiler]
[/quote]
I'm going to laugh if this happens for real.