commentary IT remains a lively, exciting and suprising place. That makes predictions particularly foolish, but here are my best bets for the winners and losers of the next twelve months.

Good Year: Google
Not the hardest of predictions, but we've had six months of Google backlash and the boys from the 'Plex are still looking good. What remains most exciting is that for all we've been wowed by the technology, Google remains a media company. Its money is coming from the deflating corpses of old media -- newspapers, radio stations, TV companies -- as the online advertising beast gets fed.

Because Google thrives on data, not selling technology, it can give technology away as much as it likes providing only it gets more data in return. If it decides to create a free online Web service-based office suite, it will only help its cause. Were Microsoft to try and do that, it would attack its own revenue stream at its most vulnerable point. Will that happen? At some point, it must.

Or Google could use its immense and ever-growing server infrastructure and interests in communication and wireless to move into media distribution, online gaming, TV, corporate data management -- its limiting factors are its ability to manage its growth and ambition without breaking the 'don't be evil' brand value it's produced. That's a better class of problem.

Bad Year: Apple
The iPod is at the top of its game now: it cannot get better. It cannot get smaller, more colourful, or sound nicer. The Queen and George W. Bush have one, yet it's still seen as an icon of the counter-culture: there is literally nowhere for it to go. Except as part of Apple's integrated media strategy -- which is, alas, missing. Together with Apple's mobile phone strategy and enterprise strategy.

There's the move to Intel chips, which in the currently fashionable phraseology involves transplanting an entire developmental ecosystem to a new planet. One, moreover, already populated with a huge number of very hungry Intel-based monsters capable of destroying margins in a microsecond.

For all it has created a grossly expanded niche, Apple remains a vertical market company in the guise of a horizontal. It needs more pillars if the roof isn't going to fall in.

Good Year: Mobile Broadband
Forget the arguments about WiMax versus Wi-Fi versus 3G versus HSDPA versus a thousand other acronyms describing stuff that even its inventors aren't sure actually exists. What people want is broadband on the move -- and that's broadband like they get at home, affordable, connected to the Internet and with no messing about.

This year, that's what we'll get. Mobile operators are making money with uncapped, fixed price services in Scandinavia, and that's with ordinary 3G. They'll make more money with HSDPA, and that's being rolled out overseas -- and trialled in Australia -- as you read this.

Things will really get going once mobile broadband becomes cheap and efficient enough to threaten the fixed-line market. Not this year, but maybe next.

Still, look on the bright side. We can expect the launch of Office 12, which is undeniably the most exciting twelfth version of any productivity software, and a plethora of back-office products. Then there's the new strategy for Media Center, which is finally showing some respectable numbers on the basis of bundling the software on anything that might one day get a TV tuner plugged into it. And then there'll be all those Google innovations to try and keep up with.

Good Year: AMD
AMD has played a smart game of late, and it's getting ready to do battle in the mobile world. Intel's got good products but AMD's kept a lead in areas such as 64-bit and chip interconnect buses -- and it's been particularly smart in taking full advantage of the way Intel has grown the chip market into new areas. The better Intel does, the better AMD does.

At the top end, AMD's at no disadvantage either in raw CPU power or in its willingness to do deals. The move towards massively multiprocessor supercomputers may not be soaking up chips in mass market numbers, but it provides some substantial numbers for sales to point to and engineers to back up their pitches with.

Will Dell succumb? That would make 2006 an exceptional year for AMD, so don't expect Intel to let it happen.

Bad Year: Itanium
This is a bit like predicting a bad year for the Ukrainian gas industry. With AMD producing more power per pound at the top end and the new low-energy server chips due to pour out of Intel Israel and smack the lower end up a bit, Itanium is more irrelevant than ever. It is possible to attend five straight Intel briefings in a row that mention enterprise computing and not hear the I word mentioned once by the blueshirts. But then, the project's been around in some form or another since 1988 and you've got to really, really want to believe in something that doesn't fly for that long. It's currently selling around seven thousand chips a quarter -- by some distance the least popular high-end server processor -- and Montecito's final arrival in 2006 may not make that much difference.

That's if it get launched. With some customers actually going back to the HP PA-RISC chip it's supposed to be replacing, and management changing within Intel, there's a bullet with Itanium written on it just waiting to be bitten. The company can be tough -- it shot Timna through the head mere weeks before launch -- and it needs to be.

Broadband is the technology that finally makes sense of the digital home, and if the media companies themselves can avoid mucking everything up then Viiv should become the platform of choice among developers. Of all the companies who've been dabbling in the idea, Intel may have got the timing right. That counts for an awful lot.

Bad Year: Open Source
The wind is out of the open source sails, as some of the big areas of interest show signs of losing momentum. Cisco's thrown open source out of Linksys' routers, because a closed proprietary system means cheaper hardware and, one presumes, fewer managerial headaches. Firefox seems to have hit its natural take-up limits. Even Microsoft's started to be intermittently nice about open source, which just cannot be a good sign, and while the nuclear option of European software patents hasn't happened yet there are plenty of skirmishes and no shortage of fingers on triggers.

There are still positive areas, such as Ubuntu doing so much work on usability and packaging, but it's hard to avoid the feeling that some of the structural issues propping up the Linux edifice are going to make it tough to beat it into mass-marketable shape any time soon.

Good Year: Open Standards
If open source is looking a bit battered, open standards are on the up. They've always been the key in hardware to making a market where innovation and competition can thrive and you only have to look at cases such as UWB where they've broken down to see how badly that can affect everyone's revenues.

The Massachusetts case has stirred up a lot of mud and heat, but it has got the main issues for software open standards out in the open. The arguments made there by the proponents of openness are strong and will be used in many forums around the world: the arguments against are unlikely to cut much ice with those who aren't already signed up to the proprietary way.

One area in particular is absolutely desperate for open standards: health care. It cannot afford to do things any other way, but remains a bastion of the closed and outdated.

Bad Year: DRM
The more they push it, the less we like it. While Sony is committing slow brand suicide by getting its name involved with the worst kind of corporate arrogance, the rest of the IT media industry isn't far behind. It shows no sign of realising that consumers hate DRM, which means that it will impose it as hard and as fast as it can. This is the perfect condition for a serious backlash, especially as the punters ask awkward questions about why their high definition DVD doesn't seem to work with their pricey high definition home cinema system when a cheapie DVD works just fine with their telly from Aldi.

Don't expect much more movement in the corporate sphere, where the concept of 'trusted computing' has run out of steam faster than you can run a spreadsheet showing how much capital expenditure is needed to ensure you can't send e-mails to your suppliers.

Nearly 30 years after he spat it out on a San Francisco stage, Johnny Rotten's "Ever get the feeling you've been cheated?" stands ready to serve as the epitaph for DRM. Get those safety pins sharpened.