The Present and the Future for GPL

Just recently our community has been excited by the prospect of Papyrus creating GPL2. Most among us, if the posts in the Racesim Central forum are any reasonable indication, seemed to positively salivate at the thought of a sibling for the original sim, while a few exceptions were less than happy at the idea of the new arrival, and expressions by those who were neutral were conspicuous only by their absence. Doubtless, there must have been as many preconceptions as to what GPL2 would actually be when, and if, it arrived, as there are members in our community. Regardless of which of these myriad suppositions might have proven to be closest to the reality had it been realized, the one thing that is certain beyond a shadow of a doubt is that the new game would have proven to have very different implications and consequences for two distinct groups within us. It is a plain fact that GPL’ers are divided into two groups, consumers and providers (and I imply no disrespect here; it is blindingly obvious that the two groups are both mutually dependent and happy in their roles).

As a professional and experienced software developer, I can’t help but notice the original developers’ use of “objects” (or lack thereof) and also the many examples of shortcuts and economies in design that betray the limitations that the development team were happy to impose on future development (it is quite possible that they anticipated no future development, hence the lack of technical encouragement). This is not a criticism of the developers, they are of course working within the constraints and requirements dictated by the realities of a commercial venture and the demands of shareholders, and I share the community’s admiration for their original formidable achievement; but these design features provide ample evidence that GPL has become something that the original developers did not envisage. And these same limitations have presented challenges to those that have sought to extend the game. As a simple example of original economy that does not facilitate further development, the same graphic element is used in the original cars for the interior cockpit panels for both the right and left side; expediency and economy at the expense of historical accuracy and future convenience.

But today, hurdles notwithstanding, we have a game that seems to be almost daily extending its boundaries and claiming new territory. This is also no doubt why we as a community seem to be gaining new members just as quickly (a remarkable thing for a game that was originally released in 1998!). The game is vibrant and alive in a way that most one year old games would envy.

Now I have seen more than once the view expressed that whatever Papyrus produced, GPL2 could be no match for GPL in its current highly developed state. This argument runs something like this: “look at our tracks; look at our cars; these are photo-realistic, how can Papyrus beat that?” Those that would attribute a lack of modesty to me would most likely assume that (since I number among the editors that have helped GPL to evolve) I would share this view – I do not. The simple fact is that to produce a game today costs something approaching a figure that some smaller principalities would be happy to have as their gross national product. You can purchase an awful lot of talent when you are willing to take that much money out of your wallet, and there is not the slightest possibility that the combined efforts of the enthusiastic add-on editors can come close to competing. (Although I concede it is possible to take a very good game and produce a sequel that’s not worth the silicon it’s printed on – I won’t mention any names here but an example that comes to mind has the number ‘3’ in it’s title – Papyrus have a demonstrated history of consistently improving on their games, and the Indycar and Nascar series are ample evidence of this).

So, we have a very high investment in effort on the part of the add-on editors to overcome the original design limitations and to acquire knowledge and develop techniques to enable them to produce products of all kinds – graphical improvements, software utilities, new tracks and cars et cetera – that either advance or add new features to GPL. And we had, until very recently, the possibility that Papyrus might release a new version of our beloved classic. It is these two factors that conspired to insinuate that GPL2 was a very different proposition for a provider than it was for a consumer.

As a consumer, the simple choice would be to either stick with the original, or to switch to GPL2. Certainly some motivation for switching would no doubt be realized as external pressure to do so – the relative scarcity of leagues that continue to support the old GPL, for example. A further motivation would be present if my prediction that GPL2 far surpassed the original in most, and possibly all, departments proved to be true. But even so, the choice would be up to the individual.

For a provider – the GPL add-on editor – the situation would be vastly different. Most providers have invested a considerable amount of time, not only to actually produce their offerings, but also an initial outlay (and in some cases this is ongoing) in order to learn how to produce an item that can interact with the original GPL. And overnight, had GPL2 been produced, this investment would have become worthless as it is likely that neither the skills and knowledge nor the add-ons themselves would be successfully transported to the new GPL. Don’t kid yourself here; time moves on, techniques and practices evolve, and it is highly unlikely that GPL2 would be created in the same way as GPL and it therefore follows that the new game would not support the products we have already built, nor value the skills and knowledge we have so far acquired.

If you doubt that GPL2 would differ from the original to such a degree that it would not be able to use our current crop of add-ons, and further that the skills and knowledge required to produce these add-ons would become irrelevant, consider this: game makers today have realized – somewhat belatedly – that to design their games in such a way as to encourage or even invite further development by the game playing community itself is in their own best interest. The benefits are manifold; it enhances the commitment to the game by the original playing community; it extends the life of the game; it extends the appeal of the game to a wider audience (as new features of the game are introduced); it increases the exposure of both the game and the game company, and more. Further, advertising the extensibility of the game increases the attractiveness and thereby increases the original sales, even before any third-party work has commenced. CodeMasters’ Operation Flashpoint, for example, advertised right on the box the fact that the game came with so many missions and that these could be supplemented by those found on the web that have been created by players themselves. Quake of course, and also Thief, allow users to create their own levels, and in the case of the latter complete “chapters” of the game can be built. MythII allows maps and whole campaigns to be created. Taking this extensibility feature even futher, Activisions’ Screaming Demons Over Europe (SDOE) exposed a completely object-oriented model to the third-party developer, allowing, for example, an aircraft builder to not only design an instrument panel, but because a particular dial is an object (with properties and methods) it can be programmatically tied to the feature of the “engine” that it reports on. In fact, SDOE is a prime example of a game that sold modestly but spawned a large and productive community that extended the game far beyond it’s original limitations. Presenting extensibility to the game playing community equates to more dollars for the original producer, and for this reason GPL2 would almost certainly have been built in such a way that it would have been wholly incompatible with the creations we are now building.

And so it was with mixed feelings that I considered the possibility of a new GPL. In truth my concern was more academic than concrete; the vast majority of software projects that are actually begun are stillborn because at some point during the life cycle, and for a wide variety of reasons, a project is canned. And here we were talking about a rumour that Papyrus might be thinking about doing (not even might be working on) a sequel. Given that this was the argument for expecting GPL2, and that it squared off against very legitimate and compelling arguments for not producing GPL2 (from Papyrus’ point of view), I had never considered the possibility of a sequel to be very likely.

But that’s enough of GPL2. It’s over. So, if we are not going to be getting a sequel, what does the future hold for the original game?

Graphically, we might have reached something close to the limits imposed by the game. We have for a long time now had “photo-realistic” items – the dashboard released with Bruce Johnson’s BT24 was built by myself and was released well over a year ago now. Much of the graphical work we are seeing today is simply adding more detailed items to areas that have not previously been rendered in this way, and this is enabled by our growing ability to pay the price in FPS.

As for objects, there is still some way to go, and the future looks bright (I’m not sure if it is as bright as the future perceived by Timbuktu, but it is promising nonetheless). Gustavo’s work to essentially derive more than one car from a single 3do – consider the recently released Eagle that also offers a “McLaren” -no doubt offers potential and will reward further experimentation.

But for me, given that we are already doing stellar work with regard to textures, cars, tracks and so on, the most exciting area that seems to offer the largest expanse of untapped potential, appears to be extending the gameplay by creating software that complements the game.

There is of course today a wide range of excellent software available for GPL, ranging from indispensable utilities that have been around a long time, such as the GPLRA, to recently released programs such as the Camera Control Master (that improves upon existing GPL functionality) and GPLProxy (that introduces completely new functionality and new interest to the game). And this is to name only a few of the excellent programs that are available. However, since I am most intimately familiar with my own software and my own plans for future software for GPL, it is my own existing or future software that I will consider in trying to answer what the future might hold for the game. It follows naturally that this is both a biased and a limited image of the future, but even though much other development will take place and enrich our gaming experience, the software which I am about to consider nonetheless serves to highlight some interesting opportunities.

Of course, we currently have GPLAIM but the full potential of this software has not been realized. You should not infer from my preceding statement that I allude to a lack of imagination on the part of the GPLAIM user. It is only the case that GPLAIM is underutilized today because all the complementary software that will help it reach its full promise has not yet been created. Some of this software, however, has recently become available. The CarSet Manager (available from within the latest released version of GPLAIM or as the standalone GPLCSM) now offers a very convenient way of switching different cars in and out of the game. While this was technically possible in the past, it was somewhat cumbersome to do. Now it is much more practical for an editor to create a few very different versions of the same car and provide them in a single package and it is likewise much easier for the end user to swap these cars in and out. As you read this, Bruce Johnson is working on the Lotus and his package will include the 49, the 39 and the 33.

But how does this impact on GPLAIM? Well… a package could be created that includes cars that span a period of time such that it would be natural to complement the cars with appropriate driver files. And if a series of these cars were produced, and the corresponding driver files were also available, the GPL player could use the combined convenience of GPLAIM and GPLCSM to easily progress from one period in time to the next (1967 > 1968 > 1969…). Actually, some imaginative driver and AI files have already been created: Shane Addison has put together a package of files that a player can use progressively as the player’s driving skills increase; Paul Moss (PlaYtiMe) has recreated the 1955 season, complete with driver.ini, gpl_ai.ini and setup files. (Both sets of files and others are currently available on the GPL:W! site). And the application of this strategy is not limited to different eras but also different racing theatres. For example, Steven Skane is the GPLAIM Carset Editor Co-ordinator, and in this role he is currently – with the help of others – working on creating a complete carset for the Tasman Series which will include both cars and drivers.

So far, we’ve discussed using existing software (GPLAIM and GPLCSM) to change eras or series using various cars and drivers, as they become available. But cars and drivers are only a part of the picture. The other major element is the collection of tracks available. Cars have progressed to the point where a single car may now come with a vast array of options, notable examples being the recently released cars from Jon Ferraez and the Brabham of Bruce M. Johnson, and providing the cars with a very wide range of options is, of course, encouraged by the availability of GPLAIM which is designed to make swapping alternate files in and out a very easy process for the player.

But tracks are not offered in this way. One track will be the same from one computer to the next (assuming the player has not added any of the (quite limited in number) optional and separately available files such as high resolution skies). A little while ago a thread appeared by Magnus Thome in the Legends Central forum where he expressed a desire for a trackset manager of some kind that would have a visual element and enable him to easily switch between sets of tracks. This thread inspired Don Hodgdon to ask “Do you think that, sometime in the future, we could also manipulate track graphics in the same fashion (as GPLAIM does)? Don had anticipated my thoughts as I was already considering the possibilities, envisaging how a program might look and work that was similar to the Carset Manager in GPLAIM. For example, a track might come with many different variations of skies, sunny or moody; or might include or exclude a particular chicane or a particular pedestrian bridge, depending upon which time period is selected; or different numbers of spectators – in Geoff Crammond’s original F1GP and I think also in GP2, there were never crowds on practice day, and it’s always bothered me that there are full crowds in GPL when I go out to practice (which, judging by my lap times, is not often enough!).

There is a lot of potential for such a Trackset Manager. When you consider the carsets that are currently being worked upon by Steven Skane that are a combination of cars and driver.ini’s (some of which represent particular era’s in motor racing history), and also Bruce Johnson’s soon-to-be-released Lotus package that will include a few different cars from different periods, the possibilities become apparent. The significant point here is that such a Trackset Manager could be used in concert with GPLAIM and GPLCSM to swap in different versions of tracks, cars and driver.ini’s as a coherent and logical set.

And so, in addition to using the Trackset Manager to customise a particular track, by selecting maps, programs and in-game details such as the number of spectators, or trees from one particular editor in preference to those produced by another, the Trackset Manager could be used with GPLAIM and GPLCSM to effectively create an entirely new era (complete with the appropriate AI) – 1972, for example! Of course, you could stay in 1967 but shift the focus of GPL to an entirely different arena, such as the Tasman Series or the South African championship.

It follows that such a set of programs is a natural candidate to work with a championship manager of some kind. As a matter of fact, the only reason GPLAIM was actually built was because I happened to be working on the programmatic manipulation of the AI, as part of a much larger – and much more ambitious – project. A little over a year ago, I created a forum thread to determine the interest that might exist in a racing management game I proposed to build that would fully integrate with GPL. This game would have all the trappings of such a management game including such elements as staff hiring (drivers, mechanics, engineers and so on), sponsorship, research, engine contracts and more. Generally, there was much support for the idea (and some good suggestions forthcoming!).

The essential premise was that although I could not affect the human player’s actual skill level to reflect the performance of the human player in the in-between management sequences, I could affect the human player’s relative performance by programmatically adjusting the AI ability up or down according to the human player’s management performance. It’s not quite as simple as this, but the basic idea is that if the sum total of all the things that a player can do on the management side between a race would equate to an improvement in the team, then the AI would be adjusted down to some extent (to some degree commensurate with level of management success) to reflect this, such that when the player actually races in the game the player’s driving has “improved” relative to the AI drivers. (AI drivers on the player’s own team would not, of course, be reduced in ability as those particular silicon-based drivers, by association with the team, derive the same benefits as the carbon-based driver). And of course the compensation system would operate in reverse if the player’s management performance proved to be poor.

This management game could be used to span several seasons and there, of course, is the tie in to GPLAIM, GPLCSM and the proposed Track Manager. In essence, this collection of software could all be used to transform GPL such that each individual player could decide exactly what they wish to see in the game, not only in terms of the cars, drivers, and tracks but also the gameplay and such elements as the time period. Of course, a GPL’er would not need to employ all the software I mentioned or even to use it exclusively. For example, another developer (Onnel) is currently working on a championship program that is “aimed at those whose primary interest is in just racing and the ability to watch their progress (and the progress of drivers around them) over time”. Onnel’s program sounds very interesting and could be used in place of the management game I propose, but still used with GPLAIM, GPLCSM and the Track Manager to increase the gameplay scope presented to the GPL player.

In conclusion, although we won’t be seeing GPL2, I think the possibilities for further extending GPL are very exciting. This is obviously a cheerful and optimistic viewpoint. So let’s balance things out a bit, shall we? When GPL is abandoned by most of us, the end will come very suddenly. One day, we’ll all swear by the best racing simulation the world has ever witnessed. But when a better game comes along, we’ll all jump ship quicker than you can say “here’s a better game”. It happened to SDOE, which until the release of IL2 was by a wide margin the best flight simulator you could own. But overnight the SDOE community became the IL2 community and all the investment by the editing community was virtually worthless. A sobering thought for a GPL editor…