Vista and Office 2007 - Microsoft's big hitters

Jon Honeyball | Oct 10, 2006 10:37 AM
At Microsoft’s grand unveiling of Office Beta 2, Vista and Longhorn, we ask why it’s being so bullish about this release.
Recently – the first day of the WinHEC (Windows Hardware Engineering Conference), a relatively small Microsoft conference held in Seattle – Microsoft went through one of those inflexion points that happens maybe once a decade.

You might think I’m overstating the case, and maybe I am, but the simultaneous launch (in Beta 2) of Office 12, Vista desktop and the Longhorn server platforms demonstrates a crushing level of code deliverability. I can’t think of any other company in the world that could slap that much new stuff on the table in one go. Anyway, you can download all of this stuff now from the Microsoft website, so I’m sure you’ll dive into the details over the coming months, and I’ll be there to point out the significant bits and pieces, and other issues of interest.

Why do I think Microsoft is so bullish about this release? The answer is simple – it has to be bullish. The merest hint of a wobble of confidence and the whole thing falls apart. You see, although it’s perfectly possible to do an incremental upgrade to this ‘brave new world of Microsoft’, the company really wants you to end up with the whole package. That means Office 12 running on Vista desktops and laptops, and with Longhorn Servers running the Office Server products. Obviously, this is too much to bite off even for the biggest players, so each part can be used on either the current or previous version of the platform. Office 12 will run fine on Windows XP, for example, and Office Server runs fine on Windows 2003 Server. But there’s no doubt that Microsoft wants us to take the lot.

There are lots of reasons for this wish. By forcing us to buy into rolling licences, Microsoft hopes Vista will make it possible to get the last remaining companies onto the treadmill of server licensing by usage rather than by outright purchase and then not coming back for more for another five years. That’s the situation we’re in at a number of my smaller clients – they bought into new server systems back in 2000, based around the new 2000 Server with Exchange Server 2000. Their desktops are now mostly on XP and Office 2000. Getting such customers to move to 2003 is hard because there’s little perceived benefit in the move – okay, there are some nice features here, a few improvements there, but nothing worth the big financial outlay. They need more than this to justify the move, and with the Office 12 and Vista launch Microsoft hopes it’s done enough work on every part of the platform to provide that justification (and to make them realise that a rolling licensing programme might make financial sense at the same time).

Ready for Vista?
Microsoft has now published the necessary specifications for a PC to run Vista, and as you’d expect there’s a base-level specification and a higher-level one, depending on whether you want to run the more advanced graphical features like the Aero Glass interface. For the basic level, you’ll need an 800MHz processor, 512MB of RAM, SVGA at 800 x 600 pixels, 20GB of hard disk space and a CD-ROM drive. Frankly, I can’t think of a more inappropriate or unpleasant machine to upgrade on – it would be many years old by now, and the thought of an 800 x 600 screen makes me shudder. For the full package, you’ll need a 1GHz processor, 1GB of RAM, an Aero-supporting graphics adapter with 128MB of graphics RAM, 40GB of hard disk space, DVD drive, audio and Internet access. The move up from 800MHz to 1GHz isn’t a big deal, but the system RAM and the video adaptor are the biggest differences, as you’d expect.

Microsoft is releasing a tool that allows you to check your PCs for Vista compatibility, which does the scanning and checking for hardware requirements and also for a bunch of software features too. Rather amusingly, the EULA (end user license agreement) for this tool says you can’t deploy it onto production machines because it’s a beta, but this will be fixed for the final release of the tool, due before Vista is finished. I had a chat with the team responsible for this tool and for the whole user setup and state migration from XP to Vista. Yes, you will be able to upgrade, and the upgrade path is supported in the Beta 2 product, so my concerns of a month ago that Microsoft was holding back on the upgrade route have proved to be groundless.

The team pointed out to me, quite rightly, that Vista installs as an image loaded onto the hard disk in parallel with the existing OS, so it’s much easier to recover if there are problems doing the upgrade. It would have been so much worse if the upgrade tool started overwriting XP files with Vista ones, so that any problem caused you to have a dual-headed monster on your hard disks. This has made me considerably less worried about the upgrade process, especially for the XP Home market. I expect, in the XP Professional market, folks will do a full wipe of the hard disk and OS replacement anyway, all under remotely managed control. There’s absolutely no need for a company’s desktop machines to be touched as part of the upgrade scenario.

We talked about checking target machines for viruses and spyware before doing the upgrade, and they told me it was their intention that quite a thorough scan of the machine would be done before the Vista setup program started doing the main work. They had to check not only for nasties living in RAM, but also to report any possible hardware-compatibility issues and driver problems too. I proffered a plea for them to add another check – to scan the hard disk for files whose archive bit hasn’t been set or is older than, say, six months. This would clearly identify those machines where no meaningful backup has been done for a long while, and thus the machines most at risk from any upgrade disaster. They thought this was a good idea and will consider it for the final version of the tool. There’s considerably more scope for such hand-holding in the home upgrade arena than among business professionals, because
a professional product needs to be more hands-off and prescriptive. In the home market, you can be considerably more interactive, and I sincerely hope Microsoft takes this opportunity to ‘deep clean’ people’s PCs as part of the upgrade process.
click to view full size image
Microsoft's all-in-one security and cleanup tool has launched in the US.
Driver signing
I also sat down with members of the Kernel team at WinHEC and bashed out the issues concerned with driver signing, and they’re finally going to take the plunge into mandatory signing of all 64-bit drivers.
I can’t exaggerate how important such a step is going to be. I asked whether it was possible to get around this unsigned lockout, maybe by using some Registry or Active Directory tweak, and I received the very firm answer that it was not possible. There are only two ways to make an X64 Windows computer accept unsigned drivers – there’s a switch you can apply in boot.ini, or you can install a debugging facility into the core of the OS. I specifically asked whether a vendor would be able to get around this, and was told no, although I wouldn’t be amused to see a major vendor using the boot.ini switch in the future on a production machine.

Why does this matter? Because too many vendors have been relying on customers to do their beta-testing for them, and in an unstructured fashion – whenever you go to a driver website, it’s easiest to download the ‘latest’ driver even if it’s a beta. The fact is, though, an end user has neither the debugging tools nor the skill set to do anything other than send a Dr Watson report to Microsoft, so Microsoft has been on the receiving end of a lot of driver crash information and has now decided that enough’s enough. I applaud this move and only wish it had been done sooner.

At least with the 64-bit platform, vendors can’t claim there’s a user and hardware migration issue, since jumping from 32-bit up to 64-bit means a clean install across the board, of course.

Allchin retires
Microsoft hosted a Vista reviewers’ workshop on the day before WinHEC started, and at the end of the day, Jim Allchin came along to take part in a Q&A session. His full title now is ‘Co-President, platforms and services division’, and so he covers almost anything that isn’t development tools, Office or the games stuff. A few months ago he announced he would be leaving Microsoft at the end of the year and handing over the reins for this huge chunk of Microsoft to others. He’s been at Microsoft since 1990, when he came from being the chief architect at Banyan Vines (an operating system and directory system that was years ahead of its time). Back in 1992, he became responsible for putting together a computing architecture for the future that was presented first at the 1992 Professional Developers Conference. Called Cairo, it was a vision of an object store, distributed networking system, object desktop and so forth.

I’ve covered what happened to Cairo over the years in this very column (one advantage of being a columnist hereabouts is that I can point you to issue numbers years ago where this stuff was microscopically described). Suffice to say, Cairo as a project went off the rails in the cold hard light of day, as the reality of bringing such a solution to market sank in. However, the thinking behind it has underpinned a great deal of the Microsoft’s effort ever since.

I took the opportunity to ask Allchin how well he thought he’d delivered on his vision over the last 15 years, taking into account the practical realities of the times in between. He said he was proud to have delivered most if not all of it - document storage, integrated email, object-based desktop, strong networking and directory services and so forth. He admitted that he’d been rocked to the core by how badly the XP Home experience has gone, and recalls Steve Ballmer dropping a friend’s computer on his desk that was riddled with spyware and other nasties. He said it had come as a major shock to Microsoft just how bad things were out there, and that this shock had driven the work done for SP 2, and now for OneCare.

Did he strike me as feeling the job was accomplished? Well, yes and no. Back then, the vision had exceptional clarity, while today’s reality is still too reliant on bolting bits together and pretending they work. Is that good enough? It depends on your perspective: if they’d gone headlong into development of the Cairo model, then they’d have been even further out of step with the Internet revolution and it might have killed the whole company. Pragmatism rules at the end of the day. It will be interesting to see how the division changes over the next year or two. Will Allchin’s heirs do a better job of bringing Longhorn and WinFS to market? Only time will tell.
Jim Allchin hands oer the reigns at Microsoft.
Jim Allchin hands oer the reigns at Microsoft.
This article appeared in the October, 2006 issue of PC Authority.