Jump to content

MemoryLeak

Supporters
  • Posts

    776
  • Joined

  • Last visited

Everything posted by MemoryLeak

  1. Err, not exactly. Gotham 2 was only out in the US a week before the UK, and that was only because we were pushed for time and Thanksgiving meant the US got priority (they give lots of presents then, therefore its important for sales). I suspect we'll get Halo 2 a couple of weeks later at the most, definitely before Xmas.
  2. I don't think its true, because of one reason: Who is developing the software to have launch titles this Xmas? It takes at least 2 years to make a modern AAA game. All the launch partners from the Xbox 1 have either released games recently, or are about to (PGR2, Amped 2, Halo2). If Halo 3 isn't a launch title for Xbox2 I'll eat my own shoes. Following from this, I'd be surprised if Xbox 2 is even out for Xmas 2005, as MS only said they'd want to launch before Sony, who clearly aren't interested in launching PS3 for ages yet.
  3. No split screen. (see Eurogamer for details) That should push LIVE sales up a bit...
  4. MemoryLeak

    PC gamers

    Fair point, I suppose Antialiasing does bump up the screen buffer sizes. At a guess its still mainly texture memory though.
  5. MemoryLeak

    PC gamers

    That used to be the case, when the screen buffers took up a large chunk of the video memory. These days its mainly for textures and geometry. As most people don't have 256MB cards, most games will be optimised to fit into 128MB or less. An exception to this is in some games where you can enable a high resolution texture set to take advantage of all that memory.
  6. MemoryLeak

    PC gamers

    It comes from the way circuit diagrams are laid out. Normally you'd have the power supply as 2 parallel lines, one at the top and one at the bottom for a simple circuit. Hence they look like rails...
  7. Doom3 will be out on the xbox, so I don't think consoles have been superseded yet. Remember consoles only have to run at 640*480 so can be comparatively underpowered with fill rate compared with a PC. PC graphics cards are stunning pieces of technology, but there aren't any games which take advantage of them properly. That lovely shiny water that everyone raves about in Far Cry was possible on Geforce1 level hardware... We are developing for next gen consoles at the moment, but who will be developing for the PC that far in advance? The problem is you can't release a game at the same time as a new breed of graphics cards comes out, as most of your audience won't be able to play it. You can, however, release a console game as a launch title with most of the features being used straight away.
  8. MemoryLeak

    PC gamers

    I think I saw (on TheInquirer maybe...) that it will need at least a 460W power supply. So most people will need to upgrade. Which is just bleedin' ridiculous.
  9. Again, I can't speak for everywhere, but I think the days of simple level design are gone. The low poly brush based Quake and unreal editors just aren't flexible enough to make high poly environments. If you are interested in level design, think of it as 2 different skills: the design aspect based around developing gameplay (puzzles, ease of movement through the environment, sniper nests etc), but the technical side of it means being able to model and texture a complex environment. I would highly recommend downloading the personal learning edition of Maya, as it is a fully functional version of the kind of 3D modelling/texturing/animation package that is used throughout the games industry. The other packages to look at are 3D Studio as you mentioned, and SoftImage XSI, which I believe is being used for HalfLife2 development, and will be in the Mod kit. Remember the industry moves quickly, and studios will already be developing for PS3 and Xbox2. These consoles will allow realtime rendering of what is currently done in cutscenes, hence the move toward the powerful modelling packages.
  10. We're recruiting for programmers at the moment, and to be honest a Masters in Games Programming doesn't get you anywhere (I can obviously only speak for one company, not the entire industry...) Experience is good, but honestly not essential. The key (as most people have pointed out) is a demo. The only point I'd question is whether to make a game. It depends what job you want, do you want to do technical engine type programming, gameplay programming, tools, AI or networking. In the modern games studio these are pretty seperate roles. So if you want a job as a technical programmer, make a physics demo, if you want to do game programming, make a game. 1. As JP pointed out. Finish it. 2. Polish it. Test it. If we run a demo and it crashes because you didn't check the display correctly when starting DirectX, it looks bad. 3. Supply code, and spend some time making the code well structured and commented. I can honestly say that if someone shows us a great demo and interviews well, we don't care what qualifications they've got. A CS degree is good though, and to be honest looks better than a game programming degree. Just my $0.02 and experience.
  11. go here: http://www.cs.sunysb.edu/~stripe/ and all will be well.
  12. I'm honestly not joking. The AI doesn't cheat, it has the same car dynamics as everyone, its just better at driving.
  13. Err... Yes, probably. Certainly that ballpark. The problem is how to make him move realistically through AI animation, a LONG time was spent animating Gollum, and that I see as the bottleneck.
  14. The AI cars don't (and can't) cheat. They have the same speed, acceleration, grip etc as the human cars. They're just better
  15. There's a few points in here that could almost do with threads of their own. 1. The MS-Nintendo thing sounds good, I hope its true. They certainly seem to be valid partners as each has opposing strengths and weaknesses. 2. The really interesting point though, is the standard platform, or more specifically: the console market ceasing to be high tech. JPickford is onto a winner here, and I suspect its because he knows first hand how unbelievably complicated it is to write a game nowadays. There is never enough time, enough artists, game designers or programmers to make the best of the current crop of consoles let alone the next generation. It is already possible to create a dynamic universe of planets and stars where you can zoom down to an apple in an orchard on one world and then zoom straight away to fly down canyons on another. All in real time, with flowing water, growing plants and animals, evolving over time, and following real world physics. All these things have been done individually and there is no reason no to conglomerate them all into one uber-engine which could run on an xbox. Honestly! The problem is it would probably take 10 years to write. So, as JP pointed out, we are already close to hardware being good enough. The next gen hardware will be able to repaint every pixel on a TV screen hundreds of times per-frame. The hard bit is going to be providing content for that hardware, and at that point the hardware becomes irrelevant. As for the comments on streaming, it is absolutely necessary. Think about what is really happening on a hardware level. A cpu does not care whether the data it requires is on a hard-disk, 2nd-level cache, register, dvd or punched card. It just varies how long it will have to get it. If you assume its best to have all the data in RAM at all times, where do you stop? It would be exponentially better to hold everything in cache, or even in directly accessible registers. The tradeoff is cost. Streaming architectures optimise the system's reliance on memory versus time to access it. That's all.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. Use of this website is subject to our Privacy Policy, Terms of Use, and Guidelines.