You can never go back. That�s what they say, and when faced with the constant movement of the PC world, it�s hard to imagine working any other way. It�s certainly true that the majority of modern applications are light-years ahead of their predecessors, and the thought of permanently eschewing Windows XP for the original Windows would be enough to put the wind up even the most hardcore PC user. How far have we come in twenty-plus years? Unless you only ever work in one window at a time, do all your work in Write and consider looking at a clock the height of excitement, it�s hard to understate the changes that have taken place.

But let�s try some different names; Zork, The Secret of Monkey Island, Ultima, and Space Invaders. These are the kind of names that not only call for a certain amount of hushed reverence, but are often still believed to be the pinnacle of their individual craft. Questionable as such claims may be, many a programmer has sat down in an attempt to recapture the sense of magic from those early games, but with little success. Just take that old classic Pacman. Everybody knows that it�s a little head in a maze, eating pills and running away from the evil ghosts. Fewer are aware of the point in a level when the red ghost will start speeding up, or how many times the game�s in-built pressure valve will blow and scatter the enemies to the far corners of the screen to give a panic-stricken player a chance to get their breath back. Even going back this far, there�s an element of craft in the computing world that goes beyond a nostalgic glimpse at an old product; the level of detail that goes into some of them is simply jawdropping. Regardless of how and why the past returns to the present, it�s certainly worth putting the future on temporary hold.

The games we played
�ScummVM�s purpose is to allow various classic 2D adventure games to run on a bunch of modern systems,� explains James 'Ender' Brown. Ender is one of the leading developers on the project, which is undoubtedly the most famous of its kind.

It all started with a simple problem. Lucasfilm Entertainment Corporation, later renamed LucasArts, was originally founded as George Lucas� games division. Its earlier adventure games, including Loom, The Secret of Monkey Island, Day of the Tentacle, and Sam and Max Hit The Road are deservedly beloved around the world, but there�s a catch. Games from that era are very hard to get running on modern systems. It can be an incompatibility with Windows XP, it can be soundcards that were never intended to run under DOS, it can be custom memory managers written in a day when people had to juggle conventional memory and its extended kin, or even the fact that not everybody uses Windows. What do you do if you want to play Day of the Tentacle under Linux, on your PDA or even � pause for dramatic shudder � on a Mac?

Put simply, ScummVM doesn�t just remake the games themselves, but takes their original data files and wraps them in an environment capable of playing them. You can increase the low-resolution images, pipe the sound through any card that your operating system supports, and play the old classics from start to finish. It�s a fully open-source project, with plenty of ports for various computer systems, and you can find it at www.scummvm.org. Who�s interested? Almost everybody. As Ender says, �We�ve had feedback, bug reports and donations from tens of thousands of people. We also have a huge international flavour, with thousands of users in almost every country from England to America, Australia, Germany, Italy, France and Japan.�

Sadly, one important member has never signed up as a fan � LucasArts itself. The project really shot into the headlines after LucasArts came to the mistaken conclusion that ScummVM was illegally distributing its games archive, and called in the lawyers. In practice, you have to get the gamefiles elsewhere � searching for existing copies on eBay. Luckily, as is often the way, this animosity would seem to have stemmed from a snap decision, and for now, it appears that the company has decided to ignore the project. Ender is quick to point out that �We have received some words of support and acknowledgement from various people who actually worked on these games, for which we are grateful�, and this attitude is mirrored in other similar projects, such as the Ultima VII game engine Exult, which carries the blessing of series creator Richard �Lord British� Garriott.

Rise of the not-emulators
Projects like ScummVM and Exult offer plenty of immediate attraction, as a result of their directly appealing subject matter, but the trend is not restricted to classic games. The Linux application Wine (www. winehq.com) attempts to do something similar for the entire Windows platform. It�s not an emulator � the very name stands for �Wine Is Not an Emulator�. It�s not a straight virtual machine like VMWare or Win4Lin � which will happily run almost anything, but still requires you to have a licence for the operating system that you want to run it on. Instead, it�s a compatibility layer. It sits between Linux and the Windows application, re-implementing the Windows API and converting its instructions into penguin-friendly equivalents.

One of the most interesting things about Wine is just how long it has been around. The original version came out in 1993 as a way of running Windows 3.1 applications under Linux, and it has been in development ever since. The main branch is the most famous, coming with most modern distributions and enabling many applications to run straight from the desktop as if they were native Linux software. More focused spin-offs have also been released, notably Crossover Office (www.codeweavers.com) for the Microsoft Office suite and WineX (www.transgaming.com) for games. Unlike ScummVM, and despite acting as a none-too-subtle way of attacking Microsoft on its own turf, it has oddly received little attention from the lawyers. The closest it got to a tussle came in mid-2003, when Microsoft leaned on the editor of FoxTalk magazine to prevent him speaking about using its software on the emulator. This finally resulted in the semi-explanation, that while running the software itself was innocuous enough, creating applications to run with it was iffy enough to call in the lawyers. Finally, much as LucasArts went silent, the issue faded from view, with an embarrassed cough from the powers that be.

Ghost of a shell
While everybody has at least one trusted application that they wouldn�t be without, it�s not very common to see entire communities rise to seize the reigns. When a company dies, or its product lines are discontinued, most programs find themselves on a slipslide ride to transient eBay fame or dusty obscurity. There are some exceptions, and the growth of the Internet has made it exceptionally easy for like-minded enthusiasts to get together.

The Blender project is one of the biggest recent successes. Blender was, and remains, a 3D modelling and animation tool, and was created by Dutch company Not A Number. The company folded in October 2001, taking Blender with it. Frankly, almost nobody outside its community had ever heard of it � and it was certainly never spoken of in the same breath as 3D Studio, Maya and friends. Nevertheless, the community set itself the task of buying the software and all rights associated with it � raising the necessary $100,000 in just seven weeks � and promptly open-sourcing it to the world via the GPL. Since then, the Blender community has grown steadily, earning worldwide attention, plenty of support sites and many a stunning rendered image. Importantly however, it�s not simply individual applications that foster this level of attention. Operating systems can easily turn into tooth-and-claw matches, and it�s hard to think of many more contentious classic systems than OS/2. This was originally fated to be the IBM-PC�s principal operating system, back in the days when IBM was automatically inserted into the term. It was the fighting between it and Microsoft that ultimately paved the way for Windows to seize control. OS/2�s last real attempt to make up for lost ground came in 1996 in the shape of OS/2 Warp 4, with advanced features such as speech recognition, a working Java Virtual Machine and OpenGL support available right out of the box. These days, OS/2 is definitely a niche market, but one that its supporters will vehemently defend.

�For the people who use it, OS/2 simply does what they want in the way they want to do it,� explains Sid Gale, Chairman of the OS/2 UK Users Group (www. warp.in-uk.net). �It�s not a gamers� platform, and it lacks advanced multimedia facilities (like video editing), but for office tasks, email, web browsing and as a server platform, it�s excellent. OS/2 licences probably number in the many hundreds of thousands (Sun has estimated 20 million), many of them in corporate use in the finance sector, and there is still OS/2 development going on. IBM still maintains it (a fixpack was released a few weeks ago), although a subscription is needed for support.� It�s not hard to find admirers if you know where to look. There�s a regular e-zine available at www.os2 ezine.com, bursting archives of software, including the OS/2 Supersite at www.os2ss.com, and even companies that specialise in OS/2 development � such as Innotek (www.innotek.de), last seen porting across the open source Open Office.org suite. As we pull inexorably into the new millennium, it would be great to see more individuals and companies engaging in more of this form of retrospective gazing.

Like all good archaeology, there are many reasons to do so. Why restrict yourself to simply looking at what did happen, instead of what could have happened. For instance, what would computers have been like if we�d never moved towards windowed layouts? We�re so used to the desktop metaphor for controlling our computers that anything else � everything from 3D environments to old fashioned command-line interfaces � now feels uncomfortable and over-complicated. You never know when you might find the perfect answer lying out there in the mists of time, and perhaps still going strong in an unexplored corner of Internet obscurity.
Richard Cobbett