Digital Amusement Game Development
  
|
Blue flag is in this room below: 
So needless to say fixing the CTF portion of this map is NOT going to be easy, as there is many other things to handle as well. The blue flag is located in the new dining room, by the dig site.
I know I'll be refering back this this, so I hope it helps you too.
I've tried my hand at tweaking ut_temple to fix it up for ctf. The author released the .map file, but a few things are really impeding progress. The map was changed since the final release, and there are shaders/textures that are refered to that I have yet to find. So it will take some time to get that all together. I've been asked about the progress of ut_superman, the answer is progress is moving very slowly currently. Don't fear that it's an abandoned project, because it's far from that. Just a lot of things to think about now before I build half the map and have to redo all of it cause of nasty gameplay bugs etc. My friend and I are scheming up a very ingenious game for the GBA that we hope to start work on in the beginning of the new year (after he's out of school). More info on that later. For those that don't know or forgot I/we can do GBA programming. A tip for people that has helped my UrT game a bit. If you've noticed, quake3 automatically uses 60hz monitor refresh for any screen resolution. Want better performance? Get Refresh Lock. Their homepage is here. With this tool you can force any screenmode to whatever refresh rate you want. Let me tell you, running 125fps at 120hz vs 60hz makes a big difference... I know Id Software (and most developers) don't look kindly on this, but since I have the Doom3 alpha, last night I started going through it, pulling it apart to see how it worked. I learned a lot. I realized what makes Doom3 very different from Half-Life2. For starters, Doom3 uses almost 4 textures for each surface texture. Why? You can't see it in the screenshots, and probably won't notice unless looking closely - lighting reflections off of metal and shiny objects move dynamically with you. The amount of work spent on bump-mapping and shadering per texture is incredible. Then there are the shadows. I nearly peed myself when I found the developer command to see what makes shadows go. There's soo soo soooo much the engine has to think about for all these shadows. There's a makeshift plasma gun in the game. Watching the shadow renderer when shooting the glowing blue plasma balls is gut wrenching because of how much is going on. I really hope the algorithms are as fast as possible by now with the d3 engine, cause wow, so much going on. I don't expect much outdoor fighting unless there's another method created to handle large spacial lighting. Even though Valve has shown off their 15 minute HL2 demo and rocked the world with it, after going through d3 last night I still believe d3 is going to be more gripping. I also spent a large chunk of time going through all sound effects, especially in the sounds/nin/ directory. Again, wow. I've been told that although Trent Reznor is no longer working on the d3 project, member(s) of NIN have picked up where Trent left off to complete the project. Sounds good to me. So that's all for now. Thanksgiving is coming up so I'll definately be out for a week or two (although I'll be working on things behind the scenes, as usual). Feast well, gg!
See, I have a saying: "Having a backup is smart, but having a secondary backup is SAFE." And so I whip out the third harddrive I keep around just in case of such an emergency. The data on this drive is a little over 2 months old. I lost the .map file for koth (update! I had a copy of the .map file stored on my server, in case of an "event", so I have most of the map. The mountain stream and many final tweaks are gone though.), and a bunch of changes to what will be beta 3 of ut_superman. Besides that and some useless email, plus a few other things, I'm still in fairly good shape. I've learned quickly that I can't trust my backup procedure. What to do? I've been contimplating building a RAID 5 array for some time. I would rather do it via software and on my current hardware, but XP doesn't support RAID at all. Some vendors have ways around this, but I'm not trusting anything not guaranteed to work at this point. I went searching around for a PCI RAID controller card, and came up with 2 options by Promise. After calling Promise and grilling them for 10 minutes, I decided that their controller was the way to go, and I went with the more expensive SuperTrak SX6000. It has 6 (yes six) IDE slots, running at ATA100. It required an additional DIMM, so I maxed it out with a stick of Crucial pc133 128MB non-ecc (CL2). I also bought another 80G 7200rpm Maxtor to drop into the array, as I have 2 drives ready to go, but the third has my data and RAID 5 requires a minimum of 3 drives. I'll probably format the old drive later and stick it in the array, or put it in my FreeBSD server, not sure yet. So overall this little unexpected incident has cost me $360 so far ($245 for controller, $80 for drive, $35 for ram). I want to get back to mapping asap, but it's going to be all paper for now lest I lose this last drive before I build the array. I would be very unhappy if I lost all my data... I have data archived back to the beginning, i.e. i386 days. It's a bit more than will fit on a couple CD's, although I should drop some of it on CDs... Either way that's the ups for now, so don't expect an update here for probably better than a week. Play ut_koth on FSK405's pub and enjoy. Later!
Whew. With no time to spare, I finished koth and released it on time. This is my first officially released Urban Terror map. I came up with the idea of King of the Hill one day for no apparent reason, and thought it was so cool that I immediately went to work. I literally dropped everything else I was doing. The first factor in my motivation was that FSK405 digital said we were going to have a Funday on the upcoming Friday night, or Saturday. I pushed hard to get a working and 100% playable map by Friday night, which I did. The map was ugly, but the core of the gameplay was there. To my disappointment, the Funday was going to be put off for a week, but I took full advantage of finishing my map before the next Funday. To keep my focus 100% on my map, I up and moved all my computer equipment from my house to my friend's place and basically lived there over the next few days. By Thursay, I had plans for Friday night and Saturday morning with real-world stuff, so I knew I had until 8pm Friday night to finish and release the map. Read on and you'll know the pain of my struggle to make that deadline. Ok, I want to make this post-mortem a bit eduational so any amateur mappers out there will have some clues as to what can happen during the construction of a map. As I mentioned before, crunch time was on during the first beta release. I hadn't packaged a map in a while so I spent a couple hours re-learning how to do it, via The6thFloor tutorials (always a good source of information). By this time, I had been able to build my map, compile it with lighting, and get a basic water shader working. This is where I began learning a few new tricks. First, I took a shot at hinting the map for better vis. I created a brush that enveloped the whole map, then clipped it into equal triangles, 8 of them. I cut the brushed horizonatally about 2/3 of the way up the hill. I did this because the top of the hill was the center of the map and can see everything, but when you're standing at the base of the map you shouldn't have the other side of the hill drawn. Later, I ended up deleting all the hint brushes and redoing them somewhat the same way, but cleaner for even better vis. I added trigger_location for radio areas early on. The skybox choosen I felt best represented the map's environment. The style of the map with a hill, surrounding water, and rocky walls were all I wanted. I played with the idea of adding a waterfall to one of the sides, no time to implement it though. For terrain, I used |nv|s's terrain instructions. The idea is simple, make a normal brush, then cut it diagonally. Copy the new brushes, place the copy next to the original, copy all brushes, paste them next to the old ones, etc until you have a bunch of brushes that look kind of like wire-mesh. The reason for this is that you can then drag vertexes for each brush up and down to create terrain. In the future, I'm going to give terragen a try (a program for automatically generating terrain) or use q3 curves (can't remember the name for this). The amount of time wasted because of gtkradiant's dislike of vertice movements was astronomical. I went through every single brush by hand and moved verticies until they matched up the looks I wanted. The hard part here was not the tediousness of customizing each brush, but gtkradiant complaining of "duplicate planes" and "degenerate planes". Because I'm still learning, these two problems were the bane of my map. When a brush got marked by these errors, it wasn't good enough just to delete the brush and rebuild it (which was a pain in the ass already). No, I had to open the .map file in wordpad and physically remove the brush there!!! As if this wasn't bad enough, then gtk begins to go unstable on me. I lost about 45 minutes of work because it crashed, even though I had it set to backup automatically every 10 minutes. That took over 1.5 hours to rebuild because of the errors occuring. It came to a point near the end where I had to save and compile every single time I modified or created a brush. This was frustrating as hell. In the meantime, I began playing with neat little tweaks to the map, like making the grass texture appear to blow in the wind. Nice, but I was shooting for high fps so that definately had to get cut. I spent a great deal of time working with the water to get the best combination of color, translucency, natural movement, etc. Once I finally built a shader I liked, I had another major problem. Under water looked totally unnatural. The textures had no kind of color filter so they looked the same under water as above water. The first obvious solution was to use fog. I did not want to use fog however because I wanted max fps. After brainstorming for a while unsuccessfully, I asked around, but no one had any better ideas than "use fog". I tried it out, but since I added a func_bobbing to my water, the fog really didn't work out at all. Not to mention you can't add func_bobbing to a fog brush. I thought some more, and tried using blue point lights underwater. The result was awesome. However point lights were too focused in areas, so I did something else. I made "light rails" just above the water around the whole hill. These rails were nodraw brushes with a blue texture and a q3map_surfacelight shader to glow blue light into the water. The result was a blueish lightmap to everything under water, and a far more natural look! I tweaked q3map_sun for the appropriate shadows based on the skybox. With lighting handled, the next thing to deal with was the hill textures. The grass came down to a sharp edge, where sand started. This looked very bad. I went through q3ut3's default textures to make some choices as to what would look the best. After I choose my textures, I used photoshop to create another texture to blend between textures. The lower part of the hill I wanted to look more wet, more growth. The top could remain dry looking. I did the same for the sand in the water and the bottom grass of the hill, fading textures. Applying the textures were very tedious because I had to rotate and resize most of them. I spent a lot of time with the outer walls too. I wanted to make areas on the corners for people to do some cheeseball sniping or hk launching. They'd be in the open, and easily dispatched. I spent a great deal of time on the brushes that make up the bottom of the water, the brushes that build up to the water surface, and making sure there weren't any leaks (i.e. places you could see between brushes and see the skybox). I spent way too much time cleaning up sparklies because of gtk's schnadagans. Sparklies are when you can sort of see through the edges of a brush. I took my first shot at making a full color minimap. I had to raise the roof of the skybox pretty high so I could noclip up there and get a good screenshot. Also, the original skybox was quite a bit bigger than the map, so the minimap arrows/player locations were off. Straightening up the skybox helped a lot. Throughout the whole creation, I made sure that all my brushes started out as 6 sides of caulk. I was pretty anal about that. For the water, I made a watercaulk shader. The brushes that made up water would be invisible but with water properties. The water shader was applied to the top of those brushes so only the surface would be drawn. I made another surface shader for the river going down the mountain (and stretched the texture out in gtk). The water around the mountain I built wrong the first time. I tried to cut it up and remove the center of the water under the hill, but somehow jacked the whole thing up. I neatly redid the water brushes and saved a lot of q3 shader drawning. Now we come to the surface of the hill. I've always wondered how to get different footstep sounds, like walking through grass. Again, the6thfloor comes through. Basically you use a little .exe program to pull the texture names from your map into a text file, assign what sounds you want the surface to sound like for each texture, then use the .exe to push that info back into the .bsp file. Easy enough! Creating the hillside stream was a pain in the butt. The hardest thing was making the water flowing into the river look at least somewhat realistic. I did my best for the time I had. I wanted bots to work on my map. I never was able to make them work before. I found out why. The .aas must be created after the .bsp, but once the .bsp is changed (i.e. you recompiled your map) then the .aas is no longer accurate and won't work. With this knowledge, I was able to make bots work in the map. Lastly, I added some breakable landscape to the top of the hill for extra astetics. The new 3.0 .def file now includes breaking stone sounds, and can be specified so only grenades and hk's can break em. Fun! I wanted to add more but didn't get a chance to, ran outta time. In the end, I had only a few custom textures to include in the .pk3, my shader file gave me no grief, I made my levelshot, made the .arena file, the readme.txt file, and tested the package for integrity. I released the map just a little after 8pm Friday night, and high tailed it out! One other thing I discovered, which frustrated me for a little while cause I couldn't figure it out, was that when I had the .pk3 of my map in my q3ut3 folder, q3 would load that file first instead of the dev map file. When I recompiled my map, it looked the same as before in-game! Removing the .pk3 file from q3ut3 fixed that. I hope this information was helpful! You can download ut_koth.pk3 either here or get it from FSK405 ftp (which will be a lot faster than my cable 40k/s upload). Play on, hf and gg!!
Here's a screenshot of the almost finished ut_koth: 
FSK405 will be having a Funday this Saturday, featuring ut_koth in the mapcycle! Release of my map is set for Friday. Afterwards, I'm going to make a long post here about the development of the map.
I found this very decent site for mapping information. Check it out. http://www.claudec.com/lair_of_shaders/tutorials/skyboxes/skyboxes_claudec.shtml Next page |
