Windows 8 “Secure Boot mode” locks Linux out of ARM tablets
Some time ago, it emerged that Windows 8 would support a new “Secure Boot” mode. In this mode, only operating systems with a cryptographic key signed and stored in the machine’s firmware would be bootable. Linux users and kernel hackers like my old pal Matthew Garrett feared this would lead to a wave of machines that Linux was not bootable on, as it’s unlikely to ever have a blessed boot key.
Ed Bott had stern words for these folk:
Microsoft has specified that this feature must be enabled by default for new systems that are sold with Windows 8 to qualify for logo support. OEM sales historically represent more than 90% of all Windows sales, making this a crucial requirement. If this feature has to be enabled manually by users, or if OEMs have the option to install Windows 8 with this feature turned off, the security feature is meaningless.
So the real question becomes this: Will PC makers make it possible for end users to toggle this option in the UEFI settings? And the answer is painfully obvious: Of course they will. They would be insane not to.
A non-trivial percentage of PC buyers will want to replace the installed operating system with either an older Windows version or an alternate operating system (like Linux). If they are unable to do so, they will call the manufacturer’s support line asking why this seemingly simple task cannot be accomplished
And Ed was right — Microsoft confirmed that it wasn’t going to force the issue, and it seems most or all OEMs will implement a disable switch for Secure Boot that the user can use at their discretion. Ed was also perfectly correct when he described the improved security that Secure Boot allows (read his post for a summary of these arguments).
However, it’s now emerged that Microsoft’s certification requirements for Windows 8 (via Aaron Williamson) says:
On an ARM system, it is forbidden to enable Custom Mode. Only Standard Mode may be enable. … Disabling Secure [Boot] MUST NOT be possible on ARM systems.
So companies designing systems with an ARM CPU — used in small, lightweight systems like tablets and smartphones — will not be able to be “Windows 8 certified” unless they require Secure Boot. Not being Windows certified is tantamount to commercial suicide, so it looks like Microsoft did indeed just lock some number of new machines down to boot Windows only. If you accept the argument that the iPad and its progeny is going to change the face of personal computing, the affected new machines are also the most interesting ones to run alternative OSs on. If you buy a tablet that runs Windows 8 but decide you’d prefer to run Android, you will be out of luck.
This is not good. Cory Doctorow (as is often the case) verges on the hysterical, but I think he’s on to something with his post the coming war on general purpose computing. Computers are becoming more and more locked down. Consider two recent neologisms: “jailbreaking”, to mean “unlocking a device to run any application we want to install”, and “sideloading”, to mean “installing an app that didn’t come from the device vendor’s approved list”.
These were features that were taken so much for granted, were ingrained so deeply into our ideas of what a computer was, that we didn’t need words for them during the first 35 years of the personal computer revolution. Now we must invent words for them. As someone with a doctorate in computer science, and as a nerd with going on for three decades of enjoying computers and computing, I’m worried about this trend.
Windows Phone 7 “spiffs”
Apparently, Microsoft is now offering resellers a $10-15 bonus, or “spiff”, for each Windows Phone 7 device sold. John Gruber doubts this strategy, saying “Obviously this isn’t sustainable in the long run, given that $10-$15 per phone is probably the most Microsoft could be making in licensing fees.”
He’s probably right that, with any non-trivial spiff, each WP7 device sold will lose Microsoft money — unless the Nokia deal includes some sort of kickback above and beyond the OEM fees, anyway. But if you doubt Microsoft’s willingness to buy its way into a competitive position, consider the Xbox programme. The first Xbox lost four billion dollars and manufacturing faults with the Xbox 360 cost another billion. Yet in 2011, the Xbox business turned $1.32 billion in profit, with more to come.
The risky strategy worked; Microsoft bought its way in. Against all the prevailing market wisdom at the time, Microsoft turned Sony’s entrenched PlayStation brand upside down to assume a comfortable second-place position in the global games console market. If Nintendo hadn’t implemented a brilliant strategy shift to mainstream markets with the Wii, the Xbox 360 would probably be number one.
Clearly, Microsoft has the stomach for long and expensive fights when it thinks there’ll be an eventual payoff. With a $233bn market cap and $56bn cash on hand, it has the deep pockets too. WP7 was late to the market, but it’s winning admirers nevertheless. I think, and I hope, that we’ll continue to see it grow.
Chris Morris on what Microsoft should and shouldn't do with the next Xbox
Great post for industrygamers.com. I agree with pretty much everything Morris says.
The big question mark I see that he hasn’t mentioned is the processor architecture. The Power PC chips used in the Xbox 360 have been in an evolutionary dead end since Apple stopping using them, but moving back to Intel’s x86 (as was used in the original Xbox) will break backwards compatibility for all existing games. Microsoft could address this with the same binary patch system used to make some Xbox games run on the 360, but that’s clumsy and can only offer spotty support. On the other hand, with prominent digital distribution on the platform like Xbox Live Arcade and the Xbox Originals, I’d argue that consumers will be expecting those titles to carry on working on a new Xbox branded console. Disappointing those consumers will not help the new console find an audience.
I’d imagine this is causing, or has caused, some serious headaches for Microsoft’s engineers.
On Microsoft and Nokia
Steve Ballmer, Microsoft CEO, on the Microsoft-Nokia deal:
Ecosystems thrive when fueled by speed, innovation and scale. The partnership announced today provides incredible scale, vast expertise in hardware and software innovation and a proven ability to execute.
According to Gartner, Nokia sold 461.3 million phones world-wide in 2010. This is an enormous figure – almost one in three phones, world wide, sports a Nokia logo. For every ten phones Nokia sells, Apple manages to sell just one iPhone. Meanwhile, despite shipping a lot of phones to wholesalers in the numbers Microsoft chooses to share (and make a lot of noise about), there’s strong anecdotal evidence that WP7 has not clicked with consumers yet.
There seems to be some belief that Nokia can somehow bring their enormous sale volume to bear on Windows Phone 7. Take Microsoft’s promising OS (hampered by poor sales) and wed it to Nokia’s huge sales figures (hampered by lack of a desirable OS). Marriage made in heaven, right?
Well, maybe not. I think many people are overestimating the influence Nokia can have on Windows Phone 7’s immediate future.
Windows Phone 7 launched in October 2010 with ten handsets from four OEMs: five from HTC, one from Dell, and two each from LG and Samsung. These are not companies with no idea about designing modern smartphones. Samsung and LG are the largest makers of Android phones, by volume. HTC are the design shop that managed to almost single-handedly keep Windows Mobile 5 and 6 alive, as well as the company who designed the influential Nexus One handset for Google.
These are not idiots – in fact, you could say they have “vast expertise in hardware and… a proven ability to execute”, and they designed a decent clutch of phones for WP7. Nokia’s handsets are constrained by the same rules of physics, component supply, and material costs as these everyone else. It is foolish to imagine they are going to immediately come up with some blockbuster handset to reverse WP7’s fortunes in the marketplace.
So what about Nokia’s huge sales volumes and brand strength? Well, let’s revisit those 461.3 million phones in 2010. Gartner state that only 111.5 million are smartphones – so about 76% of Nokia’s sales were featurephones and dumbphones. Gartner don’t state how they differentiate “smartphone” OSs but I suspect they are classing Symbian Series 30 and Series 40 as non-smartphone OSs, which isn’t necessarily unfair. S40 is a pared-down OS with no multitasking, a highly restricted web browser, or third party API for native apps beyond some very limited Java ME support. It’s barely more capable than the old embedded OSs Nokia has been using for more than a decade.
These non-smartphone sales will be primarily to cost-conscious consumers and those in less developed countries. It’s a healthy market segment for Nokia, but I’d argue that it’s also a market segment that is entirely inaccessible to Windows Phone 7. WP7 has exacting hardware requirements, including:
- a 480x800 resolution screen with capacitive touch sensors
- 1GHz ARM7 CPU
- DirectX 9 capable GPU
- 256 MB RAM, 8 GB Flash memory
- Various sensors – accelerometer, compass, ambient light level, proximity, assisted GPS
- 5 MP camera with LED flash
- FM radio tuner
- 6 hardware buttons
Clearly, it’s going to be some time before even a relatively upmarket $99-upfront no-contract phone can hope to meet those specifications, let alone the huge variety of models Nokia sells below this price point. Indeed, a review of Nokia system specifications even shows that most of the 111.5 million “smartphones” Nokia sold last year are far below this specification.
Conclusion
Nokia maintains a massive market share world-wide and vast numbers of talented designers, but the bulk of their strength is in areas that do not intersect with the high end smartphones that Windows Phone 7 is designed for. Samsung, LG and HTC failed to make a Windows Phone 7 handset that set the world alight; it’s far from guaranteed that Nokia can succeed where others failed. Indeed, an early leak of Nokia’s WP7 prototypes shows Just Another Modern Smartphone.
Contrary to many industry watcher’s opinions, I think it’s far from clear that this move will elevate Windows Phone 7 to the level of mindshare enjoyed by Android and iOS. Ballmer may be parading in fancy new clothes, but they’re cut from the same cloth as everyone else’s.
