Google revives ‘network computer’ with dual-OS assault on MS

Original URL: https://www.theregister.com/2010/12/13/google_revives_network_computer_again/

Chrome OS injects new life into netbooks and thin clients

Posted in Channel, 13th December 2010 11:34 GMT

One of the great ironies of this year is that Google and Oracle – now owner of Sun and Java – are locked in legal combat. The irony stems from the fact that, even as they bicker, the concept they did more than anyone else to create is back in the limelight. This is what we used to call the thin client, which then morphed into the netbook and now the cloudbook.

In previous iterations, the vision was stymied by the lack of reliable broadband connectivity everywhere, and effectively hijacked by Microsoft. Will the Windows giant, this time around, lose out to the approach conceived by Sun, Oracle and Google – a stripped-down device with long battery life and minimal local storage or apps, connecting for its data and services to the cloud (which we used to call the server)? Google pitched its latest definition of the thin client, with the launch of Chrome OS and a next generation netbook, just after Microsoft shipped its latest – and probably strongest – attempt at finally gaining a position in the mobile world, where the cloud will increasingly have its heart.

The thin client reinvented

Back in 1997, Oracle chief Larry Ellison and Microsoft‟s Bill Gates went head-to-head on stage at a technology conference in Paris, with Ellison unleashing the new approach to computing he had cooked up with Scott McNealy, then head of Sun, and designed to kill the traditional "fat" PC with its growing software, memory and storage burden. Oracle and Sun floated the concept of thin, internet-connected clients and were joined later by Google, with its broad vision of putting every app and every piece of data and content in the cloud, to be accessed by a widening range of always-connected, slimline and mobile gadgets.

“A PC is a ridiculous device. What the world really wants is to plug into a wall to get electronic power, and plug in to get data,” said Ellison at the time.

Back then, the client device was called the "network computer", but it largely failed because of user wariness of letting their data and apps out of their own hands (still a major factor, which sees smartphones gaining ever larger memories), and because always-on connectivity was not available, so the concept was largely chained to the desk. Ellison said at the launch of the network computer: "We'll see hundreds of thousands of machines shipped in the first year. Very quickly, we'll see entire industry move to this model. By the year 2000, NCs will outsell personal computers."

This prediction worked out like most such forecasts for the latest big thing in devices.

But the idea did help shift thinking, and it has been revived multiple times since, each time with greater impact as ubiquitous, broadband-class connectivity inched towards reality, and with it, the notion that the cloud could actually work. We saw Sun's network business appliance Corona in 1998, its Sun Ray consumer device in 2000, and a new Javastation after that. The netbook was the most serious attempt so far in form factor terms. It was conceived as a thin client, with long battery life, high portability and constant connection to the web, but limited local storage and no hard disk.

In reality, there were technical flaws in the plan: the scarcity of reliable multi-Mbps 3G; teething problems with translating the ARM/Linux combination of the smartphone to a larger device; the entrenched position of Windows in the netbook's business base; and the lack of a really workable Linux choice, with Android still immature and geared to small screens. So most netbooks were, in reality, low cost and low power Wintel mini-notebooks with hard drives – useful enough for their portability, but not a radical departure, and strictly a companion product rather than a PC killer.

Chrome OS makes its debut

Microsoft had won again, but the more mobile the cloud becomes, the weaker its advantages are in the evolution of the client. So now we see the NC/netbook concept reinvented again, this time as the cloudbook, running Google‟s new platform, Chrome OS. This could create a two-pronged attack on Windows. One on hand, Android on phones, and moving towards larger screened tablets and other devices with the release, next year, of version 3.0 or Honeycomb. And on the other, Chrome OS, initially for keyboarded products, mainly a new generation of netbooks – and looking towards a whole new breed of devices, with various form factors, but all basically providing a browser appliance for optimised access to cloud services and streamed content.

Google showed off the first tablets running Chrome OS this week, a year after first announcing its second operating system. The search giant has often sent mixed messages about where its new platform fits – a rival to Windows on netbooks; a base for a whole new class of web-oriented devices; an alternative to Android more in tune with its carrier-free, open web vision. The answer is probably that it will be all three, in time, but for now its initial aims are to attract a strong developer community, preferably from Windows, and to get the netbook concept back on its cloudy track, and away from Microsoft.

The initial Chrome OS devices look like netbooks, and the aim seems to be to evolve that form factor into something that can regain hype lost to the tablet, and firmly geared to always-on cloud/browser services. The resulting cloudbook or Chromebook is yet another product that, like tablets and smartphones, could boost overall use of web services (and adverts).

The first hardware demonstrated was a prototype, called CR-48 – just a plain unbranded netbook that developers and partners can use. The actual notebooks and netbooks promised by Acer, Samsung and others are now delayed until mid-2011, at least six months later than originally expected.

Verizon muscles in

As with Android smartphones, Verizon Wireless has seized the initiative and strengthened its deepening ties with Google. It will provide the embedded 3G connectivity for all Chrome notebooks and netbooks, at least in the first phase, offering each user 100Mbytes of free data per month for two years. There will also be various plans on top of that, including a $9.99 day pass. In many ways, this is the most interesting aspect of the launch, pushing the embedded wireless model, in which carriers are so interested, a step further – but it certainly does not reveal a Google that is achieving its aim of ending dependence on entrenched cellcos.

Other features of the OS include the ability to have several user IDs on the same machine, plus a guest mode with "incognito" or private browsing. Set-up, log-in and user interface are the same on all Chrome devices because everything is synced in the browser.

For security, always a delicate issue in cloud-based devices, the browser is tied directly to the hardware with auto updating, sand-boxing at the OS level, and encryption of user data by default. There is also a "Verified Boot" feature that ensures the OS is in the read-only firmware of the computer, so no software can modify it. Google claims this is "the most secure consumer operating system that's ever been shipped."

The Chrome Web Store is now live, and apps from The New York Times, Electronic Arts and Citrix were all demonstrated at the launch, as well as one from Amazon, despite its head-to-head battle with Google over ebooks. The new stores sees a revival of the Google Checkout payment system for click-and-pay downloads. When Google went over to PayPal for Android, it was thought Checkout, which has seen limited uptake, was on its way out. Chrome Web Store is optimised for the Chrome browser but will work with other browsers too.

The attack on Windows

Over time, new device categories will be designed for Chrome OS and a year from now, some of its disruptive potential could be on view. For now, Google is playing it fairly safe, despite the anti-Windows agenda, which will require more radical thinking once the platform is established. One Google executive with high hopes on this front is Linus Upson, VP of engineering and head of the Chrome browser project, who recently claimed that 60 per cent of businesses could immediately replace their Windows machines with computers running Chrome OS. He also said he hoped this would put corporate systems administrators out of work because software updates would be made automatically over the web.

Like Ellison's claims for the NC, such views are almost certainly overblown. But it is noticeable that the software experience has now taken the helm in driving new device types. It is Google now in the hot seat, rather than Sun with its Javastations.

Along with the spread of fast wireless networks, the rise of open web software technologies have been most important in making the NC concept viable – not changes in hardware design, which have not gone much further than removing the physical keyboard (sometimes) and shrinking the screen of a PC. This emphasis on web software puts Google into a powerful position – and potentially Oracle because of Java, but Sun has allowed Java, over the years, to lose its pre-eminence as the cross-platform, thin client framework. So the emergence of Ajax, HTML5 and the open source LAMP stack have all been more important in creating workable web apps.

Google CEO Eric Schmidt claims to have brought all those trends together with business class performance and security. “Chrome finally broke through architectural frameworks with respect to speed and security. It is now finally possible to build powerful apps on top of a browser platform,” he boasted at the launch.

Logically, of course, the use of web apps, with a browser that can sync personal data, eliminates the need for a specific computer at all. Users can just sign on from any product, driving the adoption, pricing permitting, of multiple browser-based gadgets whose hardware (notably screen size and network type) is geared to various behaviors.

Android moves towards tablets

For now, for more conventional devices that still use a lot of downloads and local processing, Google has Android, and its next frontier will be the tablet.

VP of engineering and Android chief, Andy Rubin, previewed the next generation of tablets. Showing off a Motorola device running Android 3.0 or Honeycomb somewhat stole the thunder from Monday's launch of version 2.3, or Gingerbread. Although 2.3 has better support for larger screens than its predecessors, it is not clear whether it will stretch a decent user experience beyond seven inches, and it is Honeycomb that will bring real optimisation – even for 10-inch displays.

Rubin said Honeycomb, which will ship in the first quarter of 2011, will have APIs that allow applications to be split into multiple views on the same screen.

Each release of Android should improve the experience on large-screen devices, and this will bring new launches into the field. The Motorola product on show may well have been a preview version of the promised slate it is making for Verizon Wireless, which will be integrated with the FiOS IPTV service. It runs on the Nvidia Tegra2 dual-core processor, which is scoring strongly in next generation tablets, and is geared to video content and video chat support. LG is also expected to release its first Android tablet in Q111, once Honeycomb is ready.

So this begs the question of whether tablets, as well as cloudbooks, could eat away at the traditional PC. So far, small-screen tablets like Dell Streak and Galaxy Tab behave mainly like enlarged smartphones, while large-screen versions such as iPad have the potential to be used instead of PCs for some tasks (but mainly ones PCs aren't very good at – anything involving touch, of course, plus viewing videos or books). They are not really replacements for the PC in business tasks that are heavily driven by a keyboard, and though they are likely to take the notebook's place in emerging economies – and in functions for which the PC was not well suited to start with – in the first generation they remain companion devices.

Can tablets eat into PCs?

In the medium term, though, some believe they will start to cannibalise the PC market – big PCs that is, not just netbooks, whose popularity as the leading companion device has already been eclipsed by the tablet. This means the tablet becoming the primary computer rather than an additional luxury. Gartner analyst George Shiffler is one of many who believes that, despite the hype, tablets will not kill the PC. He thinks they will displace about 10 per cent of the total by 2014 because the newer form factor will still lack extensive content creation and document production facilities.

Shiffler says tablets will primarily cut into market share for PCs that were designed specifically for on-the-go data use such as netbooks and small notebooks – though a touchscreen cloudbook could reverse this trend.

In its own study, Citigroup said 20 per cent of US and UK consumers wanted tablet-specific features such as touchscreens, but most wanted their next computer purchase to include more PC-oriented elements such as Windows and physical keyboard. Only 14 per cent said they would consider a tablet as their next device purchase, and then not usually as a PC replacement. Also, while 75 per cent of those currently considering a tablet said they would buy an iPad, 52 per cent of the total base really wanted a tablet running a Microsoft OS, and 40 per cent wanted one made by Sony, which has not launched yet. However, according to DisplaySearch analysts, if the iPad is categorised as a mobile PC rather than an oversized iPhone, Apple is now the largest mobile PC vendor in the US, with 12.4 per cent share of third quarter shipments. Within that figure, the iPad accounted for two-thirds of Apple's sales in this category – or 8 per cent of the total segment – with the MacBook Air the other contender.

In emerging economies, the iPad will have tougher challenges, as the iPhone has. This is partly because of its pricing, but also content. Hidetoshi Himuro, director of IT market research at DisplaySearch, said in a statement: "Localised content in non-English speaking regions is sparse, and iPad owners must have a PC for downloading content from iTunes. As a result, penetration in developing regions will be slow."

WP7 progress

The tablet and netbook may not be threatening extinction for Microsoft in the portable client world just yet then, but the Windows giant still faces the prospect of phone-like gadgets – including smaller tablets – increasingly taking the place of notebooks. So it must get Windows (or, if it had more sense, a new OS) into a strong position on cellphones and new form factors. It won the netbook war, but has so far blundered around on tablets (insisting on Windows 7 rather than the more suitable WP7). And of course it has made multiple attacks on the smartphone OS, with limited success.

Its latest attempt, WP7, received strong developer and consumer response and brings the welcome relief of a user interface that is not designed to look like Apple's, but gives customers a choice of experience. However, it has a huge mountain to climb because of its poor reputation on handsets, and Android being at the peak of its popularity. Microsoft is tight-lipped about reports that early WP7 sales have been disappointing, but an executive admitted it would be a long process to make the new operating system a big hitter.

Arch-rival Google echoed that sentiment, as Android chief Andy Rubin patronisingly called WP7 a "good 1.0 product" (a double-edged compliment given Microsoft's famous reputation for getting things right only at release 3.0 or later).

Speaking at the All Things Digital: Dive into Mobile conference in San Francisco, Rubin said WP7 would find it very hard to catch up with Android and Apple iOS. "There's some stuff that's 20 years old in Windows Phone 7. You have this package of stuff that was invented before the internet. It gives us a speed advantage. We can adapt and be more agile," he said, as quoted by The Wall Street Journal.

At the same conference, Microsoft's VP for WP7 program management and design, Joe Belfiore, said that Android probably contains code that is also two decades old, because of its roots in Linux. He said: "It is true that we have a kernel that has been around for a long time. I don't think that's a bad thing."

However, he would not give any early sales figures for the WP7 handsets that launched recently, saying it was too soon to disclose sales. He denied that meant sales had been slow, but admitted that it could take "a couple of years" before WP7 gained significant market share. This would require a wider range of devices, some at lower price points. The WP7 app store now has about 3,500 products.

Developers have shown considerable interest in WP7 and feedback has often been positive, though many programmers believe it is still, in effect, a beta platform with problems to iron out. Microsoft and its handset partners have run major marketing campaigns in many territories, but one challenge is to get retailers and operators to give WP7 devices high prominence. Recently, UK vendor Carphone Warehouse was cold about WP7 sales, saying that HTC Android, iPhone and BlackBerry handsets were the hot items – but observers found virtually no WP7 products on display in Warehouse outlets.

Microsoft seems to realise it will have to move quickly to increase its share, and WP7 is already due for a new year update, according to sources. Chris Walshie, part of the team that created WP7's first jailbreaking tool, claims a "massive update" is due, sufficiently radical that the vendor "could have called it Windows Phone 8".

He said: "MS took three months to do what Apple did in three years." More sober forecasters tried to predict likely enhancements – copy and paste has already been promised, and as WPCentral speculates, there could be multitasking for third-party apps, turn-by-turn directions from Bing and custom ringer support.

Wireless Watch is published by Rethink Research, a London-based IT publishing and consulting firm. This weekly newsletter delivers in-depth analysis and market research of mobile and wireless for business. Subscription details are here.