Windows 8 has been a long time coming. Running on from the success of Windows 7, the last of the desktop-focused operating systems, Microsoft have been planning the next big leap behind closed doors since 2009.
Whilst there are a lot of exposés and articles on the features and changes with Windows 8, I’m a bit too hardcore for that. My hard drive has been wiped, and Windows 8 will be my primary and only OS for as long as I can put up with it. This series of articles will cover what life is really like with the Developer Preview of Windows 8.
It’s important to stress that this is a very early iteration of the OS – it’s still very similar to Windows 7, but quite a few individual things have changed, both on the surface and underneath. I’m running this on an i7 2600k box with 12GB of RAM and a range of hardware and accessories, and I’ll be running everything from essential apps to games and a range of legacy and third-party software to give it a thorough test.
Windows 8 is also much more tablet-focused, but it’s absolutely critical that it works well on standard desktop setups of a keyboard and mouse. Not everyone will be buying a brand-new touch laptop or monitor for the upgrade, certainly not businesses. So whilst I may not be getting the best experience, I’ll be getting the same one that most Windows 8 users will get.
Windows 8 has been in the pipeline for over two years – before the release of Windows 7 whilst I was still at Microsoft. This, along with 30 years in the OS industry, gives them a better platform on which to make something amazing than anyone else.
In many respects, Microsoft are being brave by finally shedding the burden of backwards-compatibility, in some areas at least. In others, they are still clinging to some concepts that they insist are fundamental, rather than worth refreshing.
Having the Metro interface from Windows Phone 7 as the default, with the old-fashioned Explorer desktop and Start Menu in the background sounds like a disaster, but I suppose it depends entirely on the quality of the implementation, its performance, and that both interfaces are completely independent. An end-user needs to be able to spend their entire life in the Metro UI, never having to resort to the old desktop, control panel or (heaven forbid) command lines
Speed you care about
One of the most dramatic changes is not to the user interface or the mysterious modularity of the OS, but to boot times.
To cut a long story short – when you turn off Windows 7, it saves everything to disk then loads everything from scratch again on boot. It takes time to re-initialise drivers and devices, check they’re all working and set everything up before the user can even log on.
Hibernating speeds up the process, but then everything is saved from RAM to the hard disk (which takes up a huge chunk of space) and all the user sessions return with everything as it was before. There’s no halfway option that lets you quickly get to a fresh start.
Windows 8 solves this problem in a remarkably simple fashion. When you shut down, all the user sessions are ended as normal, but the system session (session 0) is hibernated. This takes up a tiny fraction of the disk space of a full hibernate, and means when the PC boots, it just has to reload that system session into RAM and the user can log in.
Microsoft are reporting boot speeds which are 70% faster than Windows 7 (which wasn’t exactly slow). That sounds like an impossible blue-sky ideal-situation statistic which no user would ever really experience.
But all I can say is – see the evidence for yourself:
Yes it’s optional – you can disable this permanently in the UI. You can also initiate an old-school cold boot from the command line or by hitting ‘restart’. But that should (hopefully) only be necessary for hardware changes and OS updates.
This almost solves one of the most annoying problems with Windows, the need to reboot because it happens to be Tuesday or because the weather changed. Once motherboard manufacturers switch from decades-old BIOS to newer UEFI interfaces, and once SSDs become more commonplace, this will rapidly deplete.
There is one extra step I’d like Microsoft to take, though. I have one user account on my main computers, it’s the only one that is ever used. Why can’t that session automatically be started up when my PC boots? Sure I can authorise automatic-logon, but that defeats the point of having a password. It should be able to figure out that at 99% of boots, this user logs on, so it’s probably worth loading that whilst waiting for him to finish making that cuppa and entering his password.
