Note: this is a post from one of our readers. I obviously have different views om the topic, but I think it's important to share maruadventurer's views. -- Tony Mobily
#1: The Operating system is no longer important. In 2009, people develop for the Web, full stop
There are whole genre of programs that like Photoshop will probably never make it to the web. Practically any program that requires a physical interface will require an OS. CAD/CAM, topology, astrophysics, astronomy, etc. The list is long.
The examples given by the author make a major omission. That mixture of "...are now developed with a mixture of HTML, AJAX, and server side programs.", misses the fact that the Apache code that serves up the HTML sits on an OS. That the Javascript was served off an OS. That the datastore, be it flat file or MySQL sits on the OS. The fact that the client side appears devoid of the need for an OS does not mean that the Web itself is devoid of one.
The key differentiator. The Web is a file cabinet, rapid communicator and idea accelerator. So any program that works in that realm can be moved to the web off the OS. Any program that requires physicality will remain on the OS. Ironically that means the back-end of the web itself will remain on an OS. At least for the time being.
#2: Google’s market power
Google does have a lot of market clout. No denying it. But Google is not infallible. They have had plenty of failures as well -- Google X, Google Catalog, Google Video Player, Google Answers. Then there are the also rans -- Orkut, Google Checkout -- viable but not dominate in their niche.
There is also the biggest mistake they ever made -- the 700mhz wireless broadband auction. They played chicken and lost. Google should have gone for broke on that one. Having those frequencies would have permitted them a new venue of ad generation and development. They wimped out. Nor is the Android - ChromeOS exactly a bright move. Tactically both developments go after different niches. But strategically this two track methodology wastes effort and expends $$ unnecessarily. Microsoft made that decision with their consumer/business OS two track and ended up with a downstream merge called XP to resolve the problem. I see the same problem here for Google.
But Tony I have to wonder. If your supposition is correct that the days of an OS are numbered due to ChromeOS then how does one make the pitch to a OEM like Dell? One cannot in one vein tout the non-OS as conquering hero. Then in the same meeting say, what do you think of this foil logo on your Inspirion? From a marketing perspective it is an immediate disconnect.
Oh and don't assume that just because Google is involved that magically peripheral interface problems disappear. There is the little issue of IP rights that many of the peripheral MFRs will be loathe to give up. If Google cannot reach a critical mass then there will be no compelling reason for the MFRs to sing Google's tune.
#3: Google creates technically-sound systems
I won't argue the point that Google makes good technical decisions. I would point out that you are no longer limited to Java to program on Android. There is now a whole series of scripting languages for use with Android. Python and Lua principally among them.
I also feel compelled to make comment on the end user desktop issue. Were I to place a bare metal box in front of you and Windows and Ubuntu disk sets and asked you to install them I think you would be surprised. From bare metal Windows can actually be harder at times. The fact is seldom if ever does a consumer ever load Windows on their box. It came that way, preloaded. As to the peripheral issue. That's like blaming the horse why the cart won't move while a wheel is on the ground, off axle. Many peripherals won't work with Linux because the peripheral MFRs won't release the API or register codes in order for third parties to develop the drivers for the hardware. But that is NOT the fault of Linux but the MFRs choice.
"There are basically no technical specifications about Google Chrome OS. It has been announced, but anything else has been pure speculation." Which is to say I am projecting my wishes and dreams that Google solves all my problems for me. Answer: Won't happen. Should Google come out with a product based on Linux Kernel then there are a whole series of givens for choice. Many in fact, long before they arrive at which Window Manager to use. Each choice prune other choices off the options list. Choose LXDE for the display as a lightweight alternative. Fine, but if you want KDE and Gnome apps then you are going to carry some of that baggage forward too and so goes your lightweight display environment.
Google does many things right. They foster innovation. They also have failures just any any other human enterprise is want to experience. But they are not the light-bringer and have no magic spells. Let us just wait and see what they bring to beta shall we?