Why Google Chrome OS will turn GNU/Linux into a desktop winner

Why Google Chrome OS will turn GNU/Linux into a desktop winner


A small revolution in the IT world is about to happen, and we are about to witness it. Microsoft Windows' domination has been challenged many times: first by OS/2 (failed), then Apple (failed), then Java and network computing (failed), then GNU/Linux and Ubuntu (failed, so far). And now, Google's Chrome OS. After such a long list of failures, what makes me think that this latest attempt will actually succeed?

There is a list of factors. Let's have a look.

#1: The Operating system is no longer important. In 2009, people develop for the Web, full stop

This point is very important. The strength of an operating system is amplified by the amount of software available for it. Microsoft Windows has benefited by a herd of developers writing all sorts of programs -- big and small -- which themselves made the Windows platform important. Java tried to break that with its virtual machine: the dream of writing software once, and running it anywhere (which didn't actually happen).

Times have changed; a lot. Anybody developing today (and for the last few years) does so for the Web. Even applications that not so long ago would have been dependent on an operating system, like document management, are now developed with a mixture of HTML, AJAX, and server side programs. People don't write Windows programs anymore. People write web applications. And it can't just be a company's slogan: applications need to "run everywhere".

Some programs are the exception here: I don't really see Photoshop becoming a web application anytime soon (although I might well be wrong). But most of the programs people use every day are moving onto the "cloud"; that's not necessarily a good thing (Google Documents is free but it's not free software!), but never-the-less this move is making the operating system increasingly less relevant.

#2: Google's market power

I will keep this simple. When you buy a printer, you always see the "Windows compatible" logo and you often see a "OS X compatible" logo too (and some of us see the irony here, since OS X uses many of the technologies GNU/Linux uses for printing). You never see a "GNU/Linux compatible" logo -- ever.

No GNU/Linux vendor has been able to change this. Expectations rose with Ubuntu, but I was disappointed. I don't think Canonical did anything wrong. I just think that the amount of strength it takes to reach such a point is immense -- both money wise, and in terms of strings that need to be pulled.

Google is the only company which could well change this. It has the power to create an infrastructure where hardware makers will actually be interested, pushed, motivated and bribed, in order to get peripherals to sport a "Chrome OS compatible" logo. The change for GNU/Linux users will be immense: regardless of the version of GNU/Linux they use, they will know that the peripheral will work.

#3: Google creates technically-sound systems

We all know that people at Google know what they are doing. Most people trust their products. Android is fantastic. I am not quite willing to forgive the fact that Android needs to be programmed in Java. It hurts me to say this (and I am sure plenty of people are ready to hurt me even more in the comments below), but I don't feel GNU/Linux is fully ready for the desktop user as it is. Ubuntu is the distribution that has gone the furthest in terms of end-user usability. However, software installation in GNU/Linux is still a problem and peripheral support is still patchy at times. These are, in my opinion, the key areas which make GNU/Linux adoption slow. Google is able to solve both of them: I am sure they will create a GNU/Linux distribution which won't have the problems I list in the article above, and will push hardware makers to add support to their peripherals -- and have the GNU/Linux compatibility logo I have been eagerly waiting for.

Will it be good, really?

There are basically no technical specifications about Google Chrome OS. It has been announced, but anything else has been pure speculation.

Here is my personal wish list -- these are all the things I consider important:

  • Software installation a la OS X: one directory = one end user application. (Keeping things secure with digital signatures for applications downloaded from other sources) This is extremely likely.

  • Availability of third-party software. This is GNU/Linux. I will want to be able to run Quake, GNU/Linux games, OpenOffice, and more. At Google, they will probably try their best to keep people online and use Google Documents, GMail, etc. However, I hope third-party software won't be treated as second-class citizens.

  • Availabiity of GNOME and KDE software. It might be done through third-party extensions, but running GNOME and KDE applications should be possible. There is a lot of amazingly good software out there. It would be absurd to expect the community to rewrite all of it.

  • Use X. The announcement talks about a "New window manager". This is all speculation, but they have three options: the first one is to ditch X altogether. This is immensely unlikely: too many drivers. The second one is to use X, but run something completely different on top of it -- a sort of full screen application which manages absolutely everything, without actually running a "proper" X Window manager. I think this is also unlikely, because it would make it really hard to run "normal" GNU/Linux programs. The third option would be to run X with their own window manager and provide some default libraries, which could well be KDE and Gnome or... well, this is speculation verging on clairvoyance, so let's wait and see!

  • Release under a free license. Seeing Google's history, especially with Android, this is highly likely.

Conclusion

Exciting times. I have just bought a Dell netbook, and would be totally happy to install the very first beta version of Google Chrome OS on it.

There is a chance Google might get a few things wrong, but this is the right time to make ourselves heard -- and excited.

And... let's be polite. Thank you Google.

Category: 

Comments

xutre's picture
Submitted by xutre on

Let's see- where do I start:

1. Windows dominates- does it really- on the smartphone?; the large business server?; the internet/streaming/cloud server?; scientific computing?; massive // processing?; high-end workstation?; fire-walled modem/router?; appliance?; getting the picture yet?;

2. Java- still around and used more than ever;

3. Linux compatible logo- I'm seeing it more and more on hardware packaging esp. as Linux based distros move toward widespread use in a growing number of countries (eg Chinese made hardware).

Terry Hancock's picture

Fair comments, but...

1) Well, Tony's obviously talking about the desktop, laptop, and even netbook marketplace (Netbooks are particularly embarrassing because they started out with GNU/Linux, but are now losing a lot of marketshare to Microsoft -- I won't speculate here on why, though I don't really think it's a genuine performance or features issue). This is about end-users, and that's who Google is after with Chrome as well.

2) I think Java got a new lease on life when it was finally released under a free license. I am no longer opposed to using it or Java apps on my own systems, though I still have little interest in learning to develop in it. Still, there are plenty of people who love it.

3) This I can't corroborate: I still don't see Linux compatibility logos on desktop or laptop hardware. I do see it on networking equipment, but that's obviously because of the strength of Linux in the server market.

In other words, we still see Linux use limited to use by "computer experts." "Ordinary people" still use Windows almost exclusively (except when they don't have to install it or maintain it themselves).

Personally, I think the problem is that we still haven't quite cracked the problem of creating a truly end-user-friendly distribution system. This is no insult on those who've tried -- it's just a really tough problem. Debian and particularly Ubuntu are close, but there persist serious process flaws that result in end-user concerns being ignored. If you know the technology well enough, you can steer around these flaws, but they continue to be show stoppers for the crowd who doesn't really want to know what's going on inside their computers, but want them to "just work".

I don't think Chrome OS is going to be what we need, myself. I think the "bootloader for the cloud" theory is probably a better bet on what Chrome OS will be. But without seeing the product, this is just idle speculation.

kornelix's picture
Submitted by kornelix on

Linux is plagued with a surplus of window managers. These cause user confusion and application compatibility issues, and the poor developers must also pick the one to use (which one?). So why do we need yet another window manager from Google? The answer could be the same as the one asserted for fixing the application installation chaos: make a de facto standard to sweep away the alternatives. I would gladly overhaul my programs in return for a clear dominant standard. "Linux is all about choice". BS! It is also about chaos and confusion and fragmentation of resources.

billkendrick's picture

> You never see a “GNU/Linux compatible” logo — ever.

I've seen them on some products. And I put one up on the front page of the Tux Paint website and on the business cards I hand out to parents and teachers that I meet (right next to the Windows and Mac logos).

wgreenhouse's picture

By definition, web applications require no installation and, ideally, no locally installed dependencies other than a browser and the ability to run whatever language the web app is written in (e.g. a locally installed Java VM, etc.). Sun Microsystems explored the idea of a "thin-client" Java operating system years ago to no great effect, and as far as Google network-based applications, the gOS distro installed on Wal-Mart's low-cost PCs was essentially based around this concept. It's a horrible little gremlin of an OS.

I see no particular reason to hope that Google's involvement will improve hardware compatibility or software package installation on GNU/Linux, other than to create yet another incompatible standard which will not benefit the Linux community in general. Google has done a great deal of work supporting existing Free Software projects through their Code project, but they simply have no experience in this sort of thing and I fear they may reinvent the wheel as they did with Android--which doesn't even have X available, much less any of the usual very useful standard Linux applications.

Re: the "Linux compatible" logo, I saw it for the first time on a recent hardware purchase: a cheap, off-brand 2-gigabyte flash drive I bought at a drugstore. It's a start. :-)

admin's picture
Submitted by admin on

Hi,

I don't agree, but I guess only time will tell. Hang around here for a couple of years and we'll see!

Merc.

Terry Hancock's picture

The real reason why you don't see "Linux Compatible" on many products is not so much that they don't work with Linux, but rather that the company would have to train their tech support people how to do it. So long as the marketshare in question is very small and the number of techs with Linux experience is also small, this is hard to crack.

You see, the problem is that once you put a label on your product, you adopt certain responsibilities: products labeled with "Supports Linux" or even "Linux Compatible" mean to most customers (and therefore to your marketing department) that a customer with a Linux system should be able to call and get tech support on installing their system. That means the tech on the other end of the line has to understand Linux well enough to provide the support.

As things stand, most of them don't. Telephone support techs are usually low-paid, relatively untrained techs who've gone through maybe 2 to 6 months of special on-the-job training. They are taught formulaic solutions to common problems in a very simplistic way. As a result, most will simply get stuck if anything is different. Windows' highly uniform product functionality helps ("Click on 'My Computer'", "go to the so-and-so tab", etc -- these instructions don't change much from system to system).

By comparison, the Linux way of doing configurations, by going into /etc and changing various config files (while in some ways actually being simpler) is likely to confound these techs and leave the company with a serious training (and re-training) problem. Adding to this is the fact that there are actually several different ways that config files are commonly set up -- the "Debian Way", the "Red Hat Way", the "BSD Way", etc.

To anyone really familiar with the system, these differences are trivial, but to someone working essentially from a script, they are somewhat hard to accommodate.

Of course, within the free software community, we mostly don't want this kind of guarantee. We're mostly just interested in knowing if there is community support for the product (and whether specs have been made available to the community for support, etc).

However, this is a fundamentally different kind of expectation than what the manufacturing company's management, marketing department, and mainstream customers expect those little "support" labels to mean. Getting them to understand and promote this different sort of customer relationship is a challenge.

It would probably be better to create some kind of approval trademark along the lines of the "UL Listed" mark (used in the US to indicate electrical safety and compatibility) that would indicate approval by an outside organization -- thus eliminating the perception that the company is taking on the support responsibility.

Of course, Google has the marketing clout to make such a label happen for Chrome, if they choose to do it, and that might have spin-off effects for other Linux distributions, assuming that they are all sufficiently compatible.

ajhwb's picture
Submitted by ajhwb on

I agree that GNU/Linux is all about choices, choices means flexibility. So, I'm free to build my custom Linux system, I'm free to decide what application to use, I'm free to decide what hardware and device drivers to use. Google want to make 'live is easy' for users, so what type of users they are talking about? what type of hardwares?

I don't care about Red Hat's way, Debian's way or Google's way. I care about the Linux's itself, knowing there are Free-as-Freedom Operating System out there was enough for me. I hate to compare each Linux system (distro) and choose a winner or to make it a winner, but I'm glad to hear another Linux based system is coming named 'Chrome' from Google (in fact they don't take too much place in the kernel development) offers blah... blah... blah...

Author information

Tony Mobily's picture

Biography

Tony is the founder and the Editor In Chief of Free Software Magazine