By now, the world and his dog knows the news: 3D Realms is dead, and with it goes its perpetual work-in-progress, Duke Nukem Forever.

The news - which has been everywhere, even making its way onto the BBC's web pages - has been greeted with a wave of surprise and disappointment - and justifiably so. Fans of Duke Nukem have been waiting patiently for the best part of twelve years for the game, shrugging off the remarks of non-believers who dismissed the project as little more than vapourware.

And like most traumatic events, the 3D Realms collapse has already sparked a number of conspiracy theories - an ex-Realms employee wrote a blog entry claiming that DNF was little more than a cover for an ongoing research and development programme for the Unreal engine, and was never intended for release at all. Others have suggested that the closure's all a big ruse; that 3D Realms essentially faked its own death and will rise like a phoenix under the recently resurrected Apogee banner. Still others have sensed an elaborate marketing ploy, citing the registration of savedukenukem.com the day before 3D Realms' closure as evidence of a viral advertising campaign.

To me, these claims sound like nonsense wrapped up in a fuzzy blanket of wishful thinking; any game developer that only releases one product in over a decade (2006's Prey) and spends twelve years developing what should have been their cash cow is bound to run into fiscal problems before too long. The notion that a company would play possum just for a bit of pre-release attention is also far-fetched to say the least - and as for savedukenukem.com, it does appear to be a genuine site devoted to cobbling together enough cash to resurrect the game, and not a marketing gimmick at all.

And then we come to the DNF project itself, originally announced back in 1997. Just to put that date into context, I was still at college in '97, and playing my N64 when I should have been studying. It's unthinkable that a game should take so long to produce. And while dedicated fans have maintained that DNF's protracted gestation was a sign that the final result would be spectacular, I'd argue that the opposite is more likely to be true.

Just look at the way the games industry works: it's all about the Next Big Thing. What's stunning one year is only mildly impressive the next. Resident Evil 4 came out and everyone was blown away by it. Four years later,  Resi 5 comes along and punters are less impressed. Why? Because in the intervening period Gears Of War appeared, and Capcom failed to learn from its innovative example - which, ironically, Epic borrowed (and refined) from Resi 4.

And so it goes, each game expanding and improving upon the achievements of the last.

DNF's production began before Half Life, before Grand Theft Auto, a few months before Goldeneye 007, even. Since then, hairstyles, attitudes and our expectations of games have all changed. The FPS has itself changed, and has taken on elements of other genres as it has evolved - levelling from RPGs, sandbox environments, superior AI.

I'm not suggesting that 3D Realms worked on DNF in a creative vacuum - far from it - but at the same time, how could they possibly work on the game for twelve years and still keep it current as the industry evolved around it? The answer, it seemed, was to repeatedly go back and rebuild the project, over and over again.

As the DNF production dragged on, the 3D Realms team were forced into a thankless game of catch-up - it began life on the now-archaic Quake II engine, before switching to an early version of the Unreal engine, and later to a custom-coded, in-house engine. In 2002, DNF director George Broussard reportedly said that 95 per cent of previous level design work had been scrapped. This week, a former 3D Realms employee, Mark Skelton, claimed that real progress had only been made in the last two to three years: "In two years of being there, we were able to take this convoluted mess and make it into a badass game," he wrote.

The stream of screenshots and production work that has leaked out of 3D Realms since its closure is evidence of its troubled history - while some of the character design work and still renders are impressive, much of the footage is not, and some of it looks suspiciously last-generation.

The rumour that 3D Realms closed its doors when Take Two, the game's publishers, refused to hand over any more funds also suggests that things still weren't right with DNF - and the fact that leaked development schedules appeared to show parts of the game still only 60% complete must have set alarm bells ringing somewhere. Duke had simply run out of time.

It's more than likely, of course, that DNF will see the light of day eventually; the IP is simply too valuable for Take Two, its holders, to ignore. And when the game does finally appear - no doubt finished by another, rather more efficient studio - I'll wager that it'll be a huge disappointment. Not only can it not measure up to the weight of twelve years of colossal expectation, the act of releasing Duke Nukem Forever so that fans can actually play it will also wreck its mythical status - like Elvis, the Loch Less Monster or Bigfoot, the project's infamy has become bigger than the game itself.

Ryan writes his gaming column every week at Den Of Geek. Last week's is here.