Exclusive interview: Cold Winter

Cold Winter is the latest in a long line of games to make a run for the, as yet unclaimed, 'best PS2 FPS' crown - a prize that has sat sparkling and tragically dormant for years.

Akin to Sleeping Beauty, slumbering in her castle and awaiting that essential kiss from a handsome prince, PS2's FPS-delivering abilities are still in repose, awaiting their own princely offering to finally arrive and awaken PS2's potential.

Could Cold Winter be it? As Cold Winter gallops ever nearer to its release, it appears confident it will be the one to have what it takes to arouse PS2's long-sleeping FPS power. But will it? What does it posses that the other suitors have lacked?

We talk to Julian Widdows, development director at Swordfish Studios, to gauge its chances.

How hard was it to create a game where everything within the levels is interactive? What problems did this create and how did you overcome these problems?
We've some experience of integrating real physics into videogames now, what with our previous game Hostile Waters having had one of the first real physics systems ever seen in a game, and Cold Winter certainly having one of the most extensive real physics systems you'll have seen on PS2, but it's fair to say that it really doesn't get any easier. The challenges are many.

Obviously with PS2, the first challenge you face is performance. Real physics is computationally expensive and for a game such as this you have a number of different systems all vying for processor time.

AI, effects, physics, weapon damage; everything requires processor time, and ensuring that the physics doesn't constantly slow the game down is a really big deal. It's fair to say that Dave (project lead) spent as much time working on level of detail and performance improvements in the physics as he did on actually integrating the solvers themselves.

The second problem you face is stability - real physics is not only costly but is also notoriously unstable, and if your physics system is always crashing it makes it very hard for the designers and artists to work productively on game levels.

This is something we experienced on Hostile Waters and continued to experience at the start of Cold Winter - stability is, really, more important than performance at the outset of the game's development cycle, as you have to have a solid platform to build on.

To this end, maintenance of the physics engine became an ongoing process throughout the development of Cold Winter - puzzles and sections of the game that relied on physics could not be properly implemented until such a time as the specific code was complete - additions to the code would adversely affect many areas that were previously working perfectly well.

This constant balancing act between adding new physics features at the risk of breaking others is one that is tough to manage and must be supported with good build practices. We ultimately changed the way we distributed game builds to the team because of early problems in this area.

The final difficulty is the tool chain. The best physics system in the world is worth nothing if the tool chain doesn't allow designers and artists to implement physics efficiently and effectively into the game. In our case, thankfully, both the Karma Kat tool and our understanding of working with game physics worked together, allowing us to create physical assets, implement them and then test them in the game with relative ease.

Is there any actual point to having everything interactive besides a desire to create a realistic and faithful representation of the real world? How does it affect the gameplay?
I've been thinking about this a lot recently, and I think the answer to this is twofold. Firstly there's a very real gameplay advantage to all the objects in the world being physical - this is clear when you play Cold Winter.

Unlike games in which your tactics are dictated by the designer's pre-positioning of objects and cover, in Cold Winter you have much more freedom of thought in that you can tip, drag and move objects around to give you a tactical advantage.

Tables can be shoved to make impromptu cover, dumpsters can be dragged around for you to hide behind, bins can be used as shields to allow you to move to better cover positions - a fully interactive environment gives the creative player options, which make each play-through different, interesting and engaging.

But the more important point, I think, is that real world game physics removes a barrier to entry - it takes away game world compromises and replaces them with real world rules. This is a big step forward for core gamers, but it's a bigger step forward for the casual gamer.

Rules that we as core gamers understand and tolerate, such as small objects like lamps not budging when you hit them, are much harder for a casual gamer to understand because outside the context of the game it makes no sense - you have to know games to understand that this is sometimes the case.

In a real physical world, however, these disconnects are removed wholesale - suddenly the world operates as you think it should, it follows the rules we've all learnt since we were born, and so the interactive experience is more intuitive, more engaging, and clearer.

So it's a bipartite improvement in gameplay - more interaction for the core gamer, with an improvement over other shooters, and easy comprehension of world rules for the casual gamer.