Board ArchivesSite FeaturesActiveworlds SupportHistoric Archives |
35850 // User Search
35850 // User SearchImportant Message Very IMPORTANTJan 15, 2001, 1:06am
Important Message Very IMPORTANTJan 16, 2001, 8:27pm
Multiple API support?Jan 8, 2001, 2:49pm
I'm relatively certain the intention is to add OpenGL support as an
option... I don't think there was ever any consideration of changing to OpenGL and removing Direct3D. At least, there shouldn't be... that would be totally absurd. Aside from increasing the number of well-supported cards, OpenGL should (assuming it's implemented correctly) provide a significant boost in performance on most 3D cards (I'm not just talking a few FPS... if used correctly, the differences between OpenGL and D3D can be quite dramatic). [View Quote] Letter to Activeworlds.com, Inc.Jan 6, 2001, 5:34pm
That's just the beginning. We're just starting to see games appearing that
use etremely impressive new technological innovations, like NWN's 'build a D&D campaign to play online' thing. Granted, that still uses an external level editor and most-likely does some sort of compiling (of course, I don't know for sure - and if it doesn't I'll be impressed). Regardless, if you look around these days, you'll see games with extremely impressive graphics - and that's just the beginning. Take Hitman's IK engine. It's got bugs, but the dyanmic skeletal physics are impressive. I'm afraid it's much too late for Worlds, Inc's little test project, though. The rewrites required to bring AW up to par as far as performance and ease of use would be incredible. At this point, more gains would be made by beginning the project anew - with a new set of programmers, and a completely new organization. The bulk of AW code is most-likely a horrid mess these days. Not to mention that the future of such a project, especially if you want to bring in things like 'dynamic vis' (portal rendering, etc.) will not come from software based on the referencing of premade RWX-format objects. It cannot be done - at least not effectively. For any sort of true interractivity, as well as the dynamic implementation of a vis system and a full lighting/shadow system (..um... even this has to be saved a bit for higher end systems; a true all-geometry shadow system would be intensive), the editor of the geometry itself needs to be IN THE SOFTWARE. Eep, I've heard you talk about putting object editing functions into the client, and that's a good idea. But it is still not going to cut it, IMO. The server needs to store the geometry, not just which objects are at what coords. It needs to store the actual geometry of the whole world in a database. Why, you ask? Because once the server and the client know the geometry, two things can happen: 1) Editing is COMPLETELY unlimited, since you're BUILDING THE GEOMETRY. 2) Since there's an easily-available database of geometry on not just the server, but also the client, you can EASILY implement dynamic vis, dynamic shadowing, COMPLETELY DYNAMIC GEOMETRY (as seen in Red Faction [ http://www.redfaction.com/ ]), and ... well, at this point the posibilities become very impressive (aside: once the client has the ability to build and modify geometry on the fly, we enter new territory. At this point, it's very plausible to implement kinetic physics engines, dynamic (not by building interraction, more along the lines of destroyable) terrain, etc. Treat the ground, if there is any, like solid mass, and give builders (or players?) the ability to remove chunks of it by editing (or destroying?). It's still in our best interest to section off pieces of geometry, much like the "brushes" in Quake-based level editors, as there still needs to be some sort of concept of an object for the application of actions and so forth (movement for doors, etc. etc.). I would also recommend the ability to import 3D models from certain formats (3ds, etc.) into the geometry. Basically, this would rip the geometry of the object and throw it into your world, wherever you specify. Texturing would most-likely be handled in some manner to be determined by the (obviously skilled, if they already built the rest) programmers, and a little trial and error. A more user-friendly version of the texturing seen in quake-style editors would be a good place to begin, unless one actually wishes to set the UV coords of each vertex by hand (granted, not as difficult a task while you're looking at the geometry and editing it). Build your rendering engine from scratch, and use OpenGL. Direct3D is the easy way out - so you can support it too. But concentrate on OpenGL, you'll be glad you did when the performance improvements show themselves. You see, this is how modern games generally get to such excellent performance. The compiling process usually involves writing the geometry into an easily-readable database, as well as performing lighting calculations and vis calculations. These days, though, compiling would not be required, as the needed information about geometry could be written on the fly, and an acceptable completely dynamic vis engine could be written. It can't happen in the current system, where the client is told by the server where certain objects should be, the client downloads the object, and then renders it. At least, not efficiently. Here's the kicker ::leans in:: Remember Quake modding? Sure you do. Every new game that knows what it's doing allows you to recode the game rules. So you want the option for gaming? Just put in a simple mod-scripting system. World owners can leave it alone (no game whatsoever - just a regular building/exploring world, like we see in AW these days), or they can modify the rules of the world (how much gravity? how much friction? how fast do people run? can they jump? how high? can they fly? how fast? you get the idea), OR they can write in rules for gaming (inventory? Weapons? What kinds of weapons? Can you destroy geometry? Use IK? HUD info? NPCs? Entities? (that's a critical one... if you want to make up your own actions (aside from the defaults, such as opening websites, making things move, etc) for 'objects' (brushes) in the world, you have to code these actions, then link the brushes to entities). etc.). It's probably a good idea to have a module-based mod system. This way, people can download modules (such as an inventory module, or something) and plop it right in. Tribes used a module system for mods. It also is probably the best idea to leave this as an uncompiled scripting language, just like in Tribes. There. There's your future-product. It does everything all at once. You can edit the geometry, you can recode the world so that it can by anything from a simple public building world to a racing game, it will run fast, and you build the world and drive the car (or kill terrorists, whatever your things is) using exactly the same software. You'll never have to buy another game again. Quite frankly, I don't think we're going to see it anytime soon. Why's that? Simple. The people with enough resources to make it don't have the vision. You have to remember that most (okay, not all) game companies are as short-sighted as AWCI, they simply have more resources. Those with the vision don't have the resources. Not in any way. And the companies with the resources, but no vision? They're quite happy without vision, thank you very much. If I ever get tired of quantum physics and biochemistry, that's going to be my project. ::points:: Who's with me? ::chuckle:: -- 35850 [View Quote] Letter to Activeworlds.com, Inc.Jan 7, 2001, 3:11pm
Quite true. Similar effects could be reached simply by giving the world
server all the information about every object... which could in turn be sent to the client... which would be able to put that information to good use as far as vision detection and the like. Still, I think a critical feature involves the dynamic modification of geometry. Building an effective geometry editor (in the style of 3D Studio MAX, one of the worst hackjobs ever) isn't that difficult. IMO, the dumbest thing about AW is the inefficient way it loads objects into view (and has been since the first versions). Another good reason to convert to a geo database instead of an object database... no horrible popup. Of course, I've always been more concerned with problems I could see. =) [View Quote] "Restricted Commands"Jan 11, 2001, 2:44pm
*ahem* Idea.
It's been agreed on that the light command needs to be restricted, thus making it completely useless in all public building worlds. So, here's a thought. Why not make yet another 'extended RWX command' such as #! collision? Except, in this case, it would of course be #! light, or something along those lines, with similar parameters as seen in the action field version. A small collection of simple, premade light objects could be added to the public world paths (they could be tiny invisible spheres, or whatever... doesn't matter). omni5m.rwx, spot2.rwx, and so forth. This way, lighting could still be used in public worlds such as AW, but not too an excessive degree, as you could only use certain lighting options. Hell, this'd save a lot on data limits even in non-public worlds, since instead of adding the light command to the action field of every street light object, you could just put it on the street light object. [View Quote] An 'Open Source' Worldbuilding ExperimentJan 12, 2001, 4:18pm
Well, I'm bored, I think it's time for a little experiment.
So, I have a small world, right? Well, I'm thinking, "Hey! Why don't I backup my stuff, clear it out, and open it up for free building and see what happens?" Then I'm thinking, "Hey! I won't bother to put in a registry either!" Then, I thought about that for a while, and I thought, "Hey! Why don't I allow people to upload objects too?" There's a couple catches, of course. First, I can't guarantee how long it'll be around. Second, the current OP is rather limited - it includes mostly AW's base objects (minus a few, since they never released a new ZIP with all the new objects, and I've just been adding them as I need 'em) plus some that I added quite some time ago, which mostly consist of new basic structural building equipment, and a random hodge-podge of other objects (I've just never really taken the time to invest that much effort into object creation). Third, by sending me objects, you're pretty much giving permission for them to be used by anyone, as this will use no password protection. So, by adding objects to the OP, you are, in essence, making them free public domain - so don't expect to be able to up anything you're afraid someone might 'steal.' Third, the world is small, like I said. It's a PS8XE, which means it goes out to 20 NSWE At first, I'm just going to leave it completely free - that is to say, I won't be doing much in the way of cleanup, unless it's really distracting or damaging to the experience (like if someone comes in and fills the whole place with thousands of random walls). We'll see how that goes. I'm assuming that some version of moderation will be needed, though, so if things start to look bad, I'll be looking for a couple people to help keep the place clean. Censorship will not be practiced - I do not believe in it. The only enforcements made will be in the spirit of keeping the world from looking like a trash heap (And, well, I suppose you could consider that a form of censorship - but I hope you understand what I'm getting at). So, here are the details: -The world is called "proxima" (no, I don't know why - I didn't originally start the world, it was sort of 'handed' to me). -As we speak, I've just installed the 3.1 server, just so you know. -At the moment, objects, textures, avatars, sounds can be sent in a couple ways: you can e-mail them to objects at gamepants.com, you can put them on your own webspace and email me that URL at the same address, or you can ICQ them to me (which means you'll have to ask for UIN). Sorry, can't FTP them yourself right now. -Everyone has build rights, bot rights, special objects. -I'm not planning on currently using an actual ground object -I won't start playing with the lights, background, bg color, etc. until people start to make suggestions -If anyone could provide me with some ZIPs containing ALL of the objects, textures, sounds, etc. from AW, I'd be very greatful. =) -I'll be taking some object-making requests. I have the skills for the most part, just never the motivation. -I'll be setting up a message board for the project in a moment. The URL will be posted in Proxima. That's all I can think of. Let there be chaos. -35850 Important Message Very IMPORTANTJan 15, 2001, 1:06am
Important Message Very IMPORTANTJan 16, 2001, 8:27pm
i miss 3.1Jan 16, 2001, 11:38am
Well, all I'm going to say is that, while the closed beta is a good idea, I
don't believe that the best group of testers was really picked... a number of the testers seem to have little to no idea what they're actually doing, and seem to be reporting just about anything without actually thinking, then correcting themselves repeatedly by posting more messages, and so forth. [View Quote] Random Thought of the Week #1Feb 5, 2001, 10:08pm
Here's the deal. We all know there are plenty of people with the
programming talent to take on this task. Hell, I could do it. The problem is that it would be a huge investment, with virtually no return for quite some time, if ever. The only reason AW was a reasonable idea to begin with is because someone could put up a little money to start. Building a newer, better AW-like thing wouldn't be the kind of task you take on as a hobby - at least, not if you want to do it right. And it must be done right, or else it should not be done at all. Half-assed attempts like the one AWCI is giving aren't even worth a second glance. So, why do we complain? Because the people with the resources will not allow them to be used correctly. The people who have the money and whatnot simply do not have the vision. You can show me as many pieces of current software as you want, I can make a judgement right now. They all suck, either because they were built by some blind fuck of a company, or because the maker was doing it alone in his spare time. Everyone's too scared to take a leap. Oh no, I might lose some quantity of money I put into the thing, then whatever will I do? You'll die, like everyone else. . . . Either that, or they just can't come up with 'leaping' sort of ideas. [View Quote] |