Digital Amusement Game Development
  
|
I shattered not one, but two long standing goals. Then I spliced those into other goals for the map. The first goal, make -the best- water in the game. I did that. Making good water has been my bane for years. Even weeks ago, I struggled with temple2_alpha's water, still only coming up barely acceptable. This time, not only does the waterflow look great, but I added raindrops to the surface, the underwater dirty, and made the whole thing look like water at night. The next goal was the rain. (21:05:58) [Dragonne{TRIAD}] I gotta say, your rain shader screenshots are crazy (21:06:08) [FSK405|wily_duck] should see them in motion (21:06:21) [FSK405|wily_duck] gonna blow some socks off (21:06:22) [Dragonne{TRIAD}] how do you overcome the volumetric issue? (21:08:39) [FSK405|wily_duck] well, it's a matter of perspective, the basis is cones, motion, and some really clever shaders making use of sort In order to get the rain looking thick in the foreground, and taper off in density into the background, I had to quickly master some previously unknown concepts - sort and alphafunc. Alphafunc was easy, this allowed the depth density, and helped fps. The hard part was sort. It's a whole new dimension in shader behavior. Using it was easy, however, getting the rain, drops, water, layers of rain, terrain, pretty much anything rendered, to render in the way I wished it to behave was a huge challenge. The idea behind sort is that instead of letting quake3 tell you what order it's going to render certain behaviors of shaders in (like opaque now, transparent after, etc), I had to manually assign values to all shaders involved as to when I wanted the engine to draw them - essentially grouping rendered layers. Even though this was the first time I used sort, I had to be extra smart about it's usage. Done wrong - players would see stuff like the river floating in front of the rain, even though the rain is definitely between them and the river. Or the rain being unrealistically dense. Or just unrealistic blending of rain to the terrain/background. The rain was very delicate in it's execution. I broke the water shader more than once (and rain too), trying to perfect the rendering order needed. Sometimes I had to mess with blendfunc on a shader to tweak it, but this changed other shader behaviors. As you can see, making rain was a huge pain in the ass. :) On the last day, I had no gameplay in the map to speak of, just a fancy tech demo. Knowing I needed a blue base, some Dr. Wily bases from Megaman came to mind. Threw in a tunnel for good measure. The base turned out nicely. Unfortunately I lost huge amounts of time when I thoughtlessly spattered target_speakers all over...which silently killed my other entities. There was a timely cleanup process involved. Over the week, I had multiple instances of seriously lost time because something totally failed. There's no lightning in the map, yet I spent a whole night just on lightning. I could go on for a while longer with a post-mortem, but I'm going to save some of that for a bit later. Much of this experience will go into the next mapmaker, which I intend to begin work on soon. As for the speedmap contest, ut4_raid is by no means winning, although I never expected to win. The map is an incredibly huge success for me, so I'm quite content already. :D -wily duck
(18:23:52) [FSK405|wily_duck`hypermode] spose this is as good of a time as any to add spawns
(18:24:10) [Delirium] :P
(18:24:16) [@Dragonne{TRIAD}] hahahaha
But I did it. In my opinion, there in no map in Urban Terror, nor any that I've personally played to date, that put you the player into a torrential downpouring rainstorm. Some people are not going to like this map, simply because the environment is so nasty. I've built a gaming experience that I never have had. After one week, ut4_raid has been submitted to the Urban Terror Speedmapping Contest! It is a phenomenal map with a very rich atmosphere. Could I have done more? Yea. There's a couple artifacts that need to be addressed, but nothing really serious whatsoever. Game play appears to be in-tact, visuals are good, all the pieces in place. Overall a HUGE success. Now, who's going to take top prize? One week til we find out... A quick look at the map. Lots of rain, but also, places to be that aren't raining, like under the bridge:  The base I whipped up today. It turned out pretty good:  Watch the submission forum for maps from everyone that participated. I'm digging the stories they're coming up with LOL!!
Current state:  Behind the bridge:  Underwater, looking up at bridge:  -wily duck :V
  Five days to go. That's not a lot of time. The other mappers are coming up with some great stuff too. An overall success so far. That, and I've learned quite a bit in the past 4 days, using methods and techniques that previously had no idea how to do. Necessity, that's how it gets done! Honestly I don't care if I win, would be cool, but the shear amount of new abiilities have already awarded vast prizes. Those images didn't happen by themselves. :V
The map looks better than it did before, skybox works now, the waterfall looks really good, overall a lot of fun. I think I'll post one more pre-alpha before calling it an alpha and handing the project off. This way both Nightcrawl gets a solid start on the hand-off, and the community gets a playable map. Want to play the current version? Download ut4_temple2_prealpha2 from the pond! One week until the speedmapping contest. I'm getting ALL OF MY DUCKS IN A ROW. That's ridiculous lol. What I have done though, is grabbed up the q3map2 shader manual (obsidian and ydnar's version), printed, put each page into plastic protection sheets, tabbed all the chapters, subtabbed the commonly referenced subjects. Also grabbed other important info too, like the radiant shortcuts doc, etc. Basically gave my mapping binder a significant overhaul. Doesn't end there, got a fresh directory install of ioUrT, applications and tools ready, physical stationary ready (graph paper, markerboard), getting url resources ready (bad things inevitably happen), getting a backup environment ready (second computer in case my primary fails), current devmap snapshot backup system ready for quick rollbacks and quickly migrating to a new computer, reading over documents included in gtkradiant's files, reading through the script/*.def files for refreshers' sake, recalling and stepping through developmental failures and backtracking to avoid similar issues, etc etc etc!!! Regardless if I happen to win or not (no prize is on the line), this will all lay the foundation for the next mapmaker release. If all goes well, we could be seeing a v03 by the end of November. Lots more preparation to do, wonder how the other guys are doing?! :D
Making good out of bad, ran into some files in my quest for answers. They're posted in the pond as another means of locating them on this vast internet. The latest q3map2_fs by 27, his latest test builds including specular, and ShminkyBoy's realworld_light shader. I'm being tempted more and more to make an updated release of the mapmaker. The fire burns.
Finally archived the main page. It hasn't been rotated in over a year and a half! I thought working on laberinto wasn't that long ago...how about year? Fast-paced year, doing big things! Back to gamedev. :) I've been taking notes on what makes a good, easy mode A.I. opponent. Game developer also had an article on this recently. For starters, I find that if an opponent sees an opportunity to score, but intentionally avoids taking it due to consideration of current standing, this works in most generic situations. Next, how focused on the goal is the opponent? They can still be fast and skilled, but instead of going slower or taking wild shots, they become more adventurous, creating very short term goals that can be excersized with the A.I.'s full potential, but does not impede on the progress of the human player. They can set up experiments that would work, but they intentially fail them before completion if the concensus reveals that the adventure/experiment would yield a significant advantage - or succeed in the test, and doodle around for a few moments (look busy!), or do nothing and give a status/tip on the situation. This would allow very inexperienced players to both stand a good chance of winning 50-65% of the time, while allowing the experienced players to face the A.I. head-on in a "serious" mode. When choosing a difficulty level, sometimes you don't want to pick between 'dumb' and 'smart' (artificial), but 'just playin' vs 'I will crush you' (natural). For even more variety, include both types of A.I.
I've had some success so far. When I left off, it appeared the method was done backwards, but still working. This means accomplishing the shader is very possible...I'll also have to look around to see if I'm re-inventing the wheel. Made more updates on temple2 last night. I'm treating the map as a beta, although it is an alpha. After some playtesting of my own, I've identified the new positioning of both flags to add significant CTF gameplay. One bomb point was also modified. The red and blue spawns will move the flags accordingly. I think this decision will prove to make the map as fun as possible. Nearly all bugs that were in alpha1 have been resolved. There's a light bleeding problem I can't seem to beat, and the skybox still refuses to render. Onward to victory!
A huge, HUGE breakthrough announcement. Bigger than that. I've been asked by Woekele about making Supersoldier an official Urban Terror map. I have no further details at the moment other than that we're going to discuss this situation further. I know that this is a huge undertaking. I also know that this is a milestone of game development success! Yay! There's a long list of other things that this means, but I only wish to focus on the fact that just the offer is enough to allow me to feel greatness. See me babble in glee! Yea! Here's another episode of Gaming Futures. I thought about id Software; the difference between Doom 1-2, Quake 1-3, against Doom 3 and Quake 4. The latter have not seen nearly the same success as the former. The commonality: storyline. It is becoming apparent to me that the need to add storyline is ruining games that are expected to tell the story simply through playing the game. I then applied this to other games, and a pattern of degredation began surfacing. Most games begin failing, maybe not by sales figures, but by replay value, when storyline is introduced as a main part of the design. Let's take Metroid for example. Through many games, the whole story had maybe 2 paragraphs of written text. The story spoke for itself, there's no need for further interruption. With Metroid Prime, the story came from tablets on the wall, or computer consoles, etc, but you never had to stop to read them. This was both an excellent way of blending a deep storyline into a written-storyline-less game, and the end of Metroid as a game that allowed the gameplay to speak for itself. The focus turned to Zelda-style gameplay in Prime 2, and more interruptive chaff. Who asked Sonic to start talking? Mario? Let's take it back to 1980 and realize all that needs to be said can be said in one line. It may take 2-3 years yet before the realization hits that forcefully injecting storylines into games (that demand focus on only the gameplay element) is evaporating as a fad. The return of retro gaming will help nail this point down. More to come on temple2 and supersoldier! -wily duck
I intend to have a gameplay in a functional state upon release. Meaning although it is an alpha, at least it'll have and hold replay value for some time. I've been knocking out a long list of issues that have quick fixes in order to make the release. The hardest was the water shader by far. Many are fairly trivial but important. There are a few speedbumps: - the map was built in and for urt2/3, so I'm continuing building it in urt3 - port it to urt4 (sounds complicated but I already cleared the sticky issues on Saturday) - have to make sure the new urt4 lighting is correct. - came up with an awesome idea to fix my water problem today. Might take a few tries to get it right. - bombmode? - odds n ends - final stand-alone testing - Phew got one of those out of the way. I'd made a long list of issues to handle, half are done. Technically I could release my pre-alpha pakfile. :) Can you feel the tease? Hang tight, a release of a map many years in the making (and shelving) is very near.
When will this be??? If all goes well - today. If not, by next weekend. There's shader work to be fixed. The map borrows a lot of textures from urt3 and q3base. Most of the textures/models/sounds will need to be swapped out...have fun with that Nightcrawl. :) Onward! A map shall be released soon! :V
I got an email from a player named nightcrawl. He's interested in continuing the work on temple. Here's what I wrote back: Good to hear from you nightcrawl. Yea I'm willing to share the .map in it's current version. I've spent a LOT of time with this map, major revamps, crazy brush work, sweat-blood-n-tears shader work. Actually my best shader I ever made started here, the water shader which I later incorporated into superman_b4. The original author gave me the .map file as something of a v1.2. This meant that I could not 'just' fix the flag problem for ctf, no, because there were all-new areas and changes and other portions that blew away the ability to simply recompile. Thus, I took the new material and ran with it. - There's now a vast underground/underwater passage (my creation), a restaurant area (original authors' creation), upper grassy area (mine), surface water tunnel, some other changes I don't exactly remember. - I experimented with a 'rope swing' from the waterfall to the high edge, which never really was successful in my opinion (a func_pendulum brush with a ladder shader). - The water I want to look pris-f'n-tine, I spent probably well over a week of my life at 8-14 hours a day dedicated solely to trying to get this water/shader to do and behave as I commanded. The result was a shader that had good water behavior on top (this is probably 80% complete), and on the reverse cull a shader that emitted a glow which colored the lightmap of all brush surfaces under the water to create that underwater environment color, without the expense of breaking down brushes into more brushes, and a soft color change between the dry surface and wet. Shader magic. - This was the first map that used a waterfall as a means of transportation. In the underground, there's now a thick waterfall from the red spawn area to the underground which is used as a vertical path. One of the map contest maps then used the same technique later. - Spent way more time than necessary on the broken glass in the restaurant (broken glass portion is my addition). - LT1 made a stone creature model specifically for this map. - The brushwork SUCKED when I got the map. It's a wonder that the original looked as good as it did. I've stripped and replaced a lot of brushwork already. - Lots more work I can't think of at the moment. What I would like to see if I continued work on the map: - Lighting tricks/hacks to do things like synthetic bumpmaps, outside lighting motion with water reflections, varying light movement of slight cloud cover, accurate flame lighting, etc - Shrub and grass moving in the wind, soft wind sounds, extra shubs, sticks, moss, vegetative life added. - Themed for a post-temple1, see theme below. - A faster map, utilizing player skills as the availability to utilize shortcuts abound, rather than set noob-walking paths between two ends. - Things that crumble with nades. - Simple/basic one-use traps deep in the temple (maybe). - Less crates! But crates from Krat Corp (the unfinished building in supermap_b4 or new building in supersoldier_alpha02) - Helicopter pad. - Revamped restaurant. - Couple basic signs near the restaurant for direction. - A new mapping trick or two here or there to help make the map FUN, add to the replay (not gimmicky, i.e. the rope swing will probably be cut or replaced) Theme: - Temple1 was the deployment of forces to survey and clear the temple for further occupation. - The area was bought for the sake of tourism, and renovated for this purpose. - An unknown creature dwelling within the temple emerged at some point in the early occupancy of tourists. - All civilians killed, unknown assailant, forces dispatched to investigate after loss of contact with temple communications. - Squads arrive to find disarray and death amidst the moderately tourist-y temple grounds. - No solid explanation for the events which transpired, almost x-files worthy (but this is never communicated to the player, the environment should tell the tale). As for the level of modern renovation, this should be kept to a minimum. Imagine it as a retreat that only seasoned global travelers/adventurers would venture to. Around the restaurant I want to put a helicopter pad in. This should be the most renovated portion, with renovations significantly declining as one moves further into the temple. Some places should be spooky and damp. Some should be bright and lush. All places should be ancient looking, something like a South American temple (except for the highly renovated area of course). Finally there should be an unoccupied area deep within that looks like something ominous could exist beyond, this would be where said creature had once dwelled, again never spoken or made totally obvious to the player, but an area that can be reached. Maybe a few skulls and mostly devoured remains are found here. Give me a few days at least, I'll have to pull together a package for distro. This is a huge undertaking, make no mistake, if you're up to the mapping challenge. I would expect a good final release of a temple2 would not look much like it does when I hand it off. I'd also expect that a ut4_temple2 final would be a current-gen state of the art map in terms of visual fidelity and architecture, using the best the quake3 engine and UrT mod has to offer. That's what I wanted out of this map. The original map was state of the art when it was released. I'd considered and posted about possibly getting together a team from the community to tackle this... couple artists, couple mappers (radiant or maya), modeller, sound person, lighting person, and coordination lead (all-purpose). So yea like 8 people. Still feeling up to it? ;) That's where I leave temple. It has a design, theme, environment, and basis; still needs to get a lot of work done to release it. Elsewhere in the past three months, I've got my hands on a few interesting games: Played Gears of War. This is an interesting take one what would be an FPS, but has a primary focus on cover. Instead of running, one hits a button to automatically rush into the cover object, then firefight from there. A couple hours revealed a decent game based on the Unreal 3 engine. There was F.E.A.R., a Quake 4 based game. In fact it played much like Quake 4, plus gravity manipulation. One can walk up and upside-down these sticky paths, or use portals in weird ways (including one that I have yet to figure out). Maybe I'll pick this up on the PC at some point, Xbox360 controller still can't beat the mouse/keyboard for FPS'n. Recently I stumbled across a "game", or actually a physics similation. It's called Phun. Imagine Windows Paint, with a physics engine. This is in fact, fun. A few minutes learning the UI, you can start manipulating objects, and let math do the rest. Try it. I'd taken up Pokemon Puzzle League for a bit a few months ago, but gave up being unable to defeat 'mew-two' or whatever on super hard mode (beat hard, unlock very hard, unlock super hard). It's a matter of time and patience, only one of which I had. Love me some Tetris Attack. On the arcade emulator, there's a number of japanese shooters I've never seen, sprout into existence. A few have been immensely fun, somewhat on par with Raiden (even got to play Raiden Fighters, yay!). Another was a robot fighting game, kinda TMNT-style, except you bust arms and equipment off the enemies and wear them as power-ups. Fun! The last one I put down quickly is actually the basis of a recently released Wiiware game. Some spiral-ball-busting puzzle game. It's fun for a while then it's time to run away! Ew... another puzzle game... I'll try to update the pond more often now. Check back when you can! And leave a shout on the left, always nice to read those. -wily duck Next page |
