Will The Free Software Desktop Ever Make It?

[mirv]

In my opinion, he is conflating issues here. Sure Ubuntu apparently doesn't provide an easy way for experts to fix issues they've caused. But the average user? They won't be messing with xorg.conf at all. So, in reality, this particular issue is a complete non-issue for the Free Software Desktop making it.

Also, seriously, I've only dabbled in making an xorg.conf and even I know that it goes Option "OptionInQuotes" "setting". He really has no one to blame but himself here, even if Ubuntu did make it hard for him to fix it.

I think his issue was more of: if something does go wrong (and it almost certainly will for some average users) that there wasn't any easy method of recovery, and there he does have a point.

[pingufunkybeat]

For me, Linux was ready for the desktop 10 years ago.

For my girlfriend, it was last year. She wouldn't go back to anything else now.

With each person, it's different. I'd say that we're on a really good way. It will probably never be ready for EVERYBODY and everybody's requirements, but quite frankly speaking, I don't give a flying crap

[?John?]

Probably my longest comment ever

Despite all that, it looks like linux won'r replace windows massively, not out of lack of anything, but really because people want what they know. Whatever.

As much as I love GNU/Linux, I still have to admit that there are some real deficiencies that take a little too long to be addressed to say at least.

I think we're missing the point here. In order to become a good desktop operating system, Linux needs 2 things:

1. A stable API for its kernel and its main subsystems, namely the graphics and audio stack. If you want people outside the open source movement to support Linux, you'll have to give them stable APIs so they can write their modules without the worry that they won't work with the next release of the package. This would also ensures some degree of backward compatibility towards old, non-open binaries that are no more supported by the developers. Yes, we would all like 100% free software on our machines, but this is not a perfect world.

Never gonna happen - see what Linus and Greg Kroah-Hartman have to say about this and as far as I'm concerned, they're perfectly right. You may not agree with them and dislike their opinion all you want but that's pretty much all you can do about it besides forking Linux and seeing how it works out for yourself. Good luck with that - perhaps the good blobbers whose interests you're trying to defend can help you out by willingly sharing some of the pain but I really doubt that.

1. One (and only one) thing for every critical task of the OS, so that people can find a similar environment on any distro they may try: one graphical server, one audio framework, one package management system, one GUI toolkit, etc. I know, Linux should be all about users' choice, but when it comes to critical parts of the OS we shouldn't have the luxury of choosing which parts fits us best. With an uniform system framework across distros, people (users and developers alike) would get more attracted to Linux, IMHO.

I agree that much more investment in any efforts to polish and standardize these areas would almost certainly help us with the serious lack of manpower we're facing right now. But again: we absolutely have to avoid tying our hands too much with backwards compatibility or we end up creating yet another pile of inconsistent junk held together only by a crapload of duct tape. At that point, I really don't care if it's free (as in freedom) or not.

If we can achieve these two goals, then Linux on the desktop will finally emerge from its niche.

All I'm trying to say is that I think it's much better to take all the time we need to make things right instead of rushing right into the hellmouth even if that means not emerging from the niche any time soon.

This is very true, I have been using Linux as my main OS for about 1 year now, but only out of necessity because windows was wigging out on me on both my desktop and laptop. Linux (Ubuntu specifically) has been treating me very well so far.

Same here, just a little longer (since Dapper Drake or so).

I have noticed that in the past 2 to 3 years Linux has become MUCH bigger than it used to be. This charge being led mostly by Ubuntu and its various derivative distributions that have branched off of it. I CAN'T WAIT to see where things go in the next 2, 3, 4 or 5 years. Linux won't be the top dog in that time. But with out any doubt some BIG things will be happening, especially in the standardization for lots of aspects of Linux usage.

Yeah. I even used to install new versions on my boxes with every release but now I tend to be a little more conservative, especially with my clients.

So, Linux fails to make simple things simple, but succeeds in making complicated things simple, and that's why we like it.

On the other hand, Windows succeeds in making simple things simple and complicated things even more complicated .

Yeah, something like that. But I still think it's a little more complicated than that.

Two deadly scenarios have to be addressed for Linux to reach masses:
1. Updating a 10.04 installed through Wubi causes "error: no such device" + "grub rescue". The average user commits suicide for less than this.
2. Configuring xorg.conf is like defusing a bomb. You cut the wrong wBOOMMM!!

Man, you're just cracking me up. So true.

I think at least 10% of Windows users could use Ubuntu nowadays, and that they would benefit greatly from its greater speed, much faster maintenance, much greater security, and quite easier use (for not bomb-defusing cases, I mean).

The reason for the current 1% instead of 10% is that Ubuntu was not pre-installed on their machines.

An additional 5% would be possible with more papercut fixing (progressing at good pace) and the damn Photoshop working seamlessly on wine (progressing at good pace also).

That's exactly what I think and right now there are quite a few years of real-world experience backing that up. I was installing Ubuntu on one of my client's laptop just yesterday. He was forced to buy a new one because someone stolen the one he had right from the trunk of his car. The new machine came with "7 sins" (not to mention the crapload of trial versions of some worthless junk nobody ever asked for) because of corrupt vendors' stupid bundling practices. There was no way to connect that piece of crap to the company network because of what seems to be decades old bug now simply being obscured. He'll be using Ubuntu for daily work (and pretty much everything else he's doing) and the only reason why he still keeps Winblow$ around is because he might need to run some WinCrap here and there so ditching that piece of crap completely doesn't make much sense (although confining it to a virtual machine is a whole different story and the only reason why it's not done yet is a lack of time).

I can understand if someone gets angry when something does not work like it should or like someone think it should work.

I understand when someone gets pissed off because stuff doesn't work the way it should, but I just can't stand it when poor artist is blaming his tools. It just plain sucks when someone is blaming something that he obviously isn't even willing to learn using properly. Unfortunately as a designer you have to assume just that.

Then it's always this linux has this and this small issue, but windows is perfect (at least the much problems with it get not quoted).

It's probably better to get used to it because I seriously doubt that attitude is gonna change anytime soon.

So software is never perfect it has always bugs in it. At least if it gets bigger then 100.000 lines.

That's right. You can screw up even "Hello, world!" if you don't concentrate enough on what you're doing. Now imagine you have to keep all those bugs and bad design decisions around indefinitely for the sake of backwards compatibility and trying to improve the thing at the same time.

I think linux is today on the desktop. Not in future, today. That not 99% of the users use it is no point because most of the people who used both systems would agree that MAC OS X is the better than windows but has also not that big market share.

History clearly shows us that even technologically superior solution doesn't have to win. That's how M$ became the monster it is today in the first place. Who knows how things turned out if OS/2 didn't flop.

Windows works on most computers, and it is installed on most computers you can buy and most people used it for several years. Especially the last point is the point why they dont will change if windows will do their job in nearly enough as they want it.

Yeah. That's how good old FUD combined with serious cash used for all sorts of dirty business practices and quite a few valid technical reasons works.

I know its written to another context but it fits the problem here too:

"and accordingly all experience hath shewn that mankind are more disposed to suffer, while evils are sufferable than to right themselves by abolishing the forms to which they are accustomed."

* United States Declaration of Independence

"Different context" doesn't necessarily mean "irrelevant".

Btw. that's also the point why we have most of the other important problems on the earth.

That's words of true wisdom right here!

For all of you whiners out there: if you dislike Linux OS so much, simply STFU and go install Windows or whatever, and then tons of other bullshit software just to make that OS secure and usable.

I really don't think any OS can be made secure any other way than by constantly evolving it in order to be more and more secure. That's because even if you completely ignore that it's impossible to come up with a 100% reliable malware-detection algorithm, then by the time the alarm goes off, it's already too late.

Anyway - the bottom line is I think it always comes down to choosing whatever set of annoyances and inconveniences you can handle the best.

[pejakm]

It just plain sucks when someone is blaming something that he obviously isn't even willing to learn using properly.

So true! I have a bunch of friends negatively criticizing Linux OS just because they used to work with Windows. After I show them it ain't such a big problem, and that (usually simple and basic) functions they need on a system are practically the same or even easier on Linux, I hear ovations and approving. But as soon as a simple problem occurs (ie. they can't find a movie player shortcut in START menu) the same outcry happens again. I had installed Arch Linux on my cousin's laptop a few years ago. He did not had any touch with computers before at all, so the Arch Linux I've installed was his first contact with an operative system at all. And guess what? He has never had any problems getting used to Linux OS. Also he didn't have any problems with it.

Anyway - the bottom line is I think it always comes down to choosing whatever set of annoyances and inconveniences you can handle the best.

Agreed 1000%!

[glxextxexlg]

With good driver support binary blob or not...

...linux desktops should have gained %10 usage by 2010, and many popular proprietary applications-games should have been ported to linux. The answer why companies don't port their applications is because they don't want to mess with customers asking for help solving a problem that has nothing to do with their software but is indeed caused by a still buggy ATI driver, an unsupported audio interface-soundcard etc. This is the major reason of linux desktop's setback and companies like AMD, creative are to blame. Praising AMD for just releasing specs to their hardware and leaving all the daunting work to opensource developers won't solve anything... but not purchasing their hardware and encouraging people to do the same will.

[deanjo]

Praising AMD for just releasing specs to their hardware and leaving all the daunting work to opensource developers won't solve anything...

In all fairness it was the community that was asking for the specs and documentation and from there they said they would handle the rest. The community then quickly realized it seems that it was a lot more work then they could chew and now they expect AMD to do all the hard work.

[pingufunkybeat]

Praising AMD for just releasing specs to their hardware and leaving all the daunting work to opensource developers won't solve anything... but not purchasing their hardware and encouraging people to do the same will.

Which hardware should be purchase then? Intel?

You won't be playing many games with that.....

[Tudhalyas]

Never gonna happen - see what Linus and Greg Kroah-Hartman have to say about this and as far as I'm concerned, they're perfectly right. You may not agree with them and dislike their opinion all you want but that's pretty much all you can do about it besides forking Linux and seeing how it works out for yourself.

Don't get me wrong, I don't want to defend people who release binary blobs here. I'm talking as an "average user" who wants to use a program he has bought (assuming that commercial, closed-source applications like games will be commonplace once Linux will emerge on the mainstream desktop) even after he updates his/her kernel or the X.org server to the next "minor" version. Am I wrong?

I agree that much more investment in any efforts to polish and standardize these areas would almost certainly help us with the serious lack of manpower we're facing right now.

Exactly!

But again: we absolutely have to avoid tying our hands too much with backwards compatibility or we end up creating yet another pile of inconsistent junk held together only by a crapload of duct tape. At that point, I really don't care if it's free (as in freedom) or not.

I completely agree on this. Backward compatibility must be held if it's humanly possible. If you can't mantain that without infernal hacks and whatnot, then it's time to change the API... but if you change it way too often, it means also that you're not designing your software so well

All I'm trying to say is that I think it's much better to take all the time we need to make things right instead of rushing right into the hellmouth even if that means not emerging from the niche any time soon.

Again, I agree 100% to this.

[blackiwid]

...linux desktops should have gained %10 usage by 2010, and many popular proprietary applications-games should have been ported to linux.

I don?t know if that would be really good. I don?t really want unfree software under Linux. I would only maybe tolerate them if they would run in a vm which protects the rest of the system for adware-features like spying or rootkits and stuff like that.

The answer why companies don't port their applications is because they don't want to mess with customers asking for help solving a problem that has nothing to do with their software but is indeed caused by a still buggy ATI driver, an unsupported audio interface-soundcard etc.

That did not stop companys to realese software that did at the beginning work just with big problems or not even startet with Ati Hardware/Drivers (they did not send AMD a Demo or something to make it working) so just 1-2 months later amd made it work.

This is the major reason of linux desktop's setback and companies like AMD, creative are to blame. Praising AMD for just releasing specs to their hardware and leaving all the daunting work to opensource developers won't solve anything... but not purchasing their hardware and encouraging people to do the same will.

They do not just give the specs free and doing nothing they made the fglrx driver better (faster), and parallel they also pay some developers to make a good opensource drivers.

So there is right now some big changes in the linux-graphics-development. So all takes a bit longer (majorly gallium3d). But the time between new hardware-releases and working drivers even with advanced features like 3d are get always smaller. And all the supportet hardware gains from the new features we read each month about here on phoronix ^^

[glxextxexlg]

Which hardware should be purchase then? Intel?

I have a laptop with nvidia corporation's g210m gpu, and I can do everything with it the hardware is manufactured to do. And at least Intel supplies its customers with opensource

drivers

not just opensource text specs with magic numbers on them