aliens

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - NancyGold

Pages: 1 ... 7 8 [9] 10 11 ... 13
121
Offtopic / Re: XCOM Inspired Fantasy Game
« on: July 10, 2020, 08:37:39 pm »
Things one learns: voxels actually do look like cubes, if you render them at resolution higher than a pixel. For some reason I believed the will look like pixels, but that is the case only when one projects them as pixels, when rendering like trinagles. And one needs to render them at higher res to precisely edit them. But triangle-style rendering for voxels has a lot of quirks:
1. One has to sort voxels by distance from camera.
2. The one draws them upscaled based of Z.
3. If several voxels project to the same screen pixel, one just blends them.
That was very inefficient but it worked for small scale models.

Alternatively one can convert voxels to triangles. And that actually works well for sparse scenes. But the conversion process is very expensive, and one cant modify the model after conversion - polygons are surprisingly hard to mutate. Although if one doesn't do large modifications, then space can be broke into parts, which are much faster to convert  to render into triangles, after being modified. I personally just raytrace them like spheres, since raytracing allows sampling very large, potentially infinite constructs. One guy even made a complete video game base of infinite fractals: https://www.youtube.com/watch?v=9U0XVdvQwAI

There fractals are mutates them with time and it acts like complex a course to navigate. Think Marble Madness + Monkey Ball on speeds.

Anyway, I have already mentioned that XCOM Apocalypse had all objects internally made out of voxels, so projectiles move with great precision
https://www.terrygreer.com/xcomapocalyse.html
Quote
Tiles were complex data structures. Each one also had a reference to line of sight definition – this was a simple 4x4x4 grid – a sort 3d texture – where each cell was either solid or empty. There were a relatively small number of these solid line of sight definitions, but enough to approximate the 3D shape of any shape of tile created. This meant that the world was effectively broken into a 3d voxel grid where weapon fire could be accurately ray-traced. It was a ball ache to set up initially, as every tile had to have a line of sight definition assigned, but really made life easy later on as collision in game then became automatically generated from the map editor.

It was a genius approach, especially in the years before decent 3d raytracing – and one of the key identifying features of an xcom game. Players always loved being able to snipe enemy forces from right across the map through a couple of windows or blown open walls. It’s also a technique which I think useful to reinvent for use in conjunction with true 3D worlds and can think of lots of good mechanics you could use them for.

Obviously I plan to integrate that voxel engine with Spell of Mastery. Just because it opens so many possibility on both visual effects and gameplay sides. But it is unlikely I will use complex projectile raycasting, since I want player being easily able to guess where unit can shot. The game does use simple raycasting already and it was real pain to make remotely intuitive. Yet I need robust large scale voxel modifications to produce stuff like tiles breaking apart or just for site map transformation.


122
Where are Mulder and Scully?  :D

123
Offtopic / Re: XCOM Inspired Fantasy Game
« on: July 09, 2020, 11:04:35 pm »
Rewriting the code to C/C++ without any optimizations instantly allowed working with 256x256x256 cubes. Although I had to fix a nasty bug with floating points -0.0 behaving differently in edge cases due to -fast-math compiler keyword. I'm sure I can boost that to 512x512x512 cubes with basic optimizations and multithreading. Still not enough to draw a full XCOM map where every pixel corresponds to actual voxel. The target size 1024x1024x1024 will require actual GPU rendering.

Modern GPUs apparently can execute C/C++ with some limitations. A huge leap from the basic DirectX assembly shaders I remember being used in games like Morrowind. But shaders are still not very portable.

Then again, if I implement optimization structure, like octree, I can reduce Z dimension, since XCOM has just 4 height maps and each tile is 16 pixels high, only around 1280x1280x80 voxels are needed to draw full XCOM map.

124
Offtopic / Re: XCOM Inspired Fantasy Game
« on: July 08, 2020, 09:36:56 pm »
I still keep working on the voxel utility. Basically I'm redoing Photoshop, but with voxels. The editing functionality will be very similar and one could as well edit 2d images with, yet there will be a way to apply filters in 3d. Hope it could also be useful to OpenXCOM modders, since it is much easier to edit and paint voxel than to edit 2d sprites (if you have the editor lol). If I manage to optimize it to show 1000x1000x1000 cubes, one could even map the whole XCOM site into voxel space. And then do stuff like bullets making actual holes in objects.

There is the bare minimum functionality to make it useful for my purposes (i.e. any voxel animation makes layers mandatory), and I will have to transfer the core rendering code into C/C++, since Symta still has no unboxed floats and I need every single voxel out of it. Probably should even use GPU (but I have none :-\ ). UI will remain in Symta.


125
Offtopic / Re: XCOM Inspired Fantasy Game
« on: July 08, 2020, 08:45:22 pm »
A nice video explaining that to have uniform non squares or hexes cells one will need to use hyperbolic space. Dunno if there are many tactics games in such space. But the author of that video even made a Marble Madness style game completely with 3d fractals.


126
Fan-Stuff / Re: Graphics from Ufo:TTS and Source Files
« on: July 06, 2020, 08:21:26 pm »
The site doesn't exist anymore :(

Has anyone saved the rar?

127
Programming / Re: Auto-Battle
« on: July 06, 2020, 08:01:00 pm »
It didn’t hand control of your troops off to the AI to finish the battle. Instead, it reduced the participants on both sides to a couple of abstracted numbers, and then rolled some dice.
Given that aliens always have numerical and firepower superiority, that wont ever work in a fair playground. So yeah, it has to make assumptions. Moreover, players always has a chance to retreat, taking away some stuff. For me it was always Chryssalid present = instant retreat.

Some people might be tempted to say that an auto-battle function is cheating, or it defeats the whole point of the game, but I have a few reasons for thinking that OpenXcom would benefit from having one. In no particular order:
Yeah. It makes a totally different game, similar to Civilization, or more precisely to Emperor of the Fading Suns, which had stack autocombat, with different type units getting different initiative (i.e. artilery goes before tanks).

The whole point of XCOM is tactical combat, while the strategic world map is here just to glue the tactical missions into a coherent story.

In my XCOM inspired game I do have autocombat, but it is used only for AI vs AI battles, since properly simulating battle scape would take some time and annoy human players.

128
Offtopic / Re: Finally (re)Found
« on: July 05, 2020, 04:43:50 pm »
https://youtu.be/XFZl83i4-JA?t=56

That photo was made in modern Russia. Look at the American clone car plate and the well fed faces (Soviet citizens had far less healthy faces). Yet an "ageing" photo filter was applied to it for some reason. Sinatra's music was impossible to obtain in USSR. And the fact you have such recording would have been enough to net you a nice gulag sentence or just execution on the spot, without any court hearing.


129
Offtopic / Re: XCOM Inspired Fantasy Game
« on: July 05, 2020, 04:42:00 pm »
Currently I'm working on a little voxel utility which I plan use for few dynamic effects, like say tiles breaking apart in explosion. These cannot be drawn manually, especially not in a singe dev project.


130
Offtopic / Re: XCOM Inspired Fantasy Game
« on: June 30, 2020, 02:46:21 pm »
Had to rethink explosions. The original XCOM (which I use a starting point for everything), had some very obscure explosion mechanics. You never knew if that explosion will clear the wall or damage the UFO.

Now I want to make it simpler: all fragile items in some range will get destroyed, while fragile tiles will get cleared. In addition throwing a bomb on top of a castle wall will turn that wall into rubble, which will fall out if there is free space near it. That way cannon shots would gradually turn castle walls into rubble. In addition the explosion is sending units around it flying a few tiles back.

Now compared to XCOM I don't have squaddies with firearms determining their capabilities, but instead monster types doing that. Although there are stuff like grenade flasks and health potions, weapons humanoids have mostly determine their attack and damage. But I find these annoying to use, since player will have to manually resupply them after every battle, and that was a big part of the original XCOM - ordering enough grenades. So in my case goblin gets just 5 grenades, it can throw during the day, and such ammunition auto-resupplies each word turn, just like health.

I havent yet added the general rock-paper-scissors damage/armor system. It probably should be added, because it will help further differentiating different units and items. I.e. if a dragon uses fire as a weapon, it must have some inborn resistance to fire. As of now there are only undeads, which are resistant to non-blessed attacks. Yet they also do more damage to non-blessed units and have higher to-hit chance against them. So resistance system is not that simple, since on has to decide how and where it will apply.

Still a lot of work to do. And I haven't even decided on the story, beside that the player should start the game without a city and arrive at a neutral port city. The story is not really that important for a strategy game, as opposed to the victory condition, which can even be made random. IIRC, The Sims had such system, where victory condition was based of the randomly generated character's personality. And such approach actually nicely suits strategy games, especially multiplayer ones, where player has to guess what goals pursue other players. Same way it will fit nicely with a narrative based AI. And if you think about it, the original XCOM already had asymmetric game sides, where the player and the AI controlled aliens had different goals. On the other hand, Apocalypse didn't had a narrative AI, but instead had a rubber-band AI, where the quicker the player advances, the harder the AI hits.

131
Offtopic / Re: XCOM Inspired Fantasy Game
« on: June 26, 2020, 09:23:07 pm »
Proceeding with implementing the city plundering. When the city has not enough mercs to defend itself, but still has some population (expressed as wealth/taxpayers), city streets will be replaced with actual civilians. These are very weak units, but would still be a nuisance. I also think implementing mechanics of punishing players for killing them. Similar to how in Vagrant Story killing civilians in some missions led to these missions being lost.

Maybe something like metropolia sending a ship with knights to end the carnage. Not children sprites, because I'm too lazy to draw them and it will create some controversies. Still in real life wars and propaganda children victims play some non small role. And children were frequently used for  service roles, cannon fodder and last resort reserve. And it is still a thing even in modern world: https://en.wikipedia.org/wiki/Child_slavery#Child_soldiers

Anyway, I already had a male sprite, and so I had to finish the female sprite. Animations are not very polished but should do for now. Proper approach would be hiring a professional artist, instead of doing animations myself, but I have no access to crowd funding or any funding at all, and I think professional animation would require more frames of animation, and that will also require implementing actual resource management.

Currently all sprites are loaded into memory and never freed after that. Some sprites are also generated dynamically, so it would require some refactoring. In addition, resource system implies special archive format for fast loading from HDD, since many people still dont have SSDs. In fact, the recent Command & Conquer remaster was criticized for not accommodating for SSDs.


132
Offtopic / Re: Footage of real XCom ufo interception.
« on: June 26, 2020, 09:01:49 pm »
https://www.politico.com/news/2020/06/23/senators-ufo-government-reports-336021
Quote
Senators want the public to see the government's UFO reports

So now it is not about the UFO existence, but about their nature. Are these some Russian ICBM testing use to hint US that Putin can nuke everyone any moment, or maybe a very advanced spy planes of some country or the actual aliens probing some stuff on the planet?

133
Offtopic / Re: XCOM Inspired Fantasy Game
« on: June 23, 2020, 10:27:59 pm »
Government is more a feedback loop, I saw enough examples as bad government can corrode whole society and opposite can happen too. Look on current USA, this can go in any way, one person making one correct decision could define next 100 years of it. No society is unified, all have some subgroups, and depending on circumstances everything could go upside down.
But government still needs a strong and integrated nation to regenerate itself reliably over time. Russia had several revolutions, but each time it moved to the same core, surviving both the complete change of its elites in 1917 and the total economic collapse of the USSR. No matter the Tsar or Stalin, Russia will remain Russia. Just like if you turn the planet Earth into dust, that dust will accrete back to create the planet (although likely very damaged), unless it gets velocities larger enough to escape the synergistic gravitation of all other dust particles. On the contrary, Roman Empire got disintegrated, and its language went extinct. So government is more of the function of what there is under it.

If you're a single person on an uninhabited island, then you're both the king and the people there, but surprisingly not much you can do. I.e. if Bill Gates gets onto an uninhabited island, he wont create Microsoft. If he gets on an island with some uneducated aborigines, there wont be a Microsoft either. To create Microsoft, Bill Gates needs a large, healthy and educated nation with a lot of resources and developed technologies. I.e. Bill Gates is the function of the nation, which itself is a function of the environment it has, which is, well, first of all shaped by that force called gravity.

Anyway I still keep thinking about the proper way to implement leaders/officers in the game, what makes people loyal and breaks loyalty. I.e. why for example Putin's officers wont just kill him, and wont start a war between each other. What truly stops these generals? How does the loyalty arise?

In addition there is a question about how to make AI challenging. There are a few possibilities.
1. Non-cheating AI. That AI just doesn't cheat and start with the same resources as player. Hard to implement and not very exciting.
2. Rubber band AI. Cheats by having its income to be a multiple of player's income. So the more money player makes, the more money the AI makes. Such AI will always be on match with the player, no matter how quickly player acts to capture resources. The AI will be annoying, since no matter what player does, it will still come back, and will be treading on the tail. The game will likely turn into some strange meta, where player will try acting with low income to avoid turning AI into powerful opponent. Player basically plays against itself.
3. AI gets stronger with time to the point of being impossible to beat (original XCOM AI). Such AI  will pose ominous challenge for inexperience players, yet will be boring for advanced players, who can managed to get far ahead of time. And the speed with which AI gets stronger is something hard to balance. It quickly gets from easy to impossible. I.e. that is just a game with a time limit, where player can move time back with bonuses. Original Civilization also had similar AI, varying its speed with difficulty, but modern remake uses a non-cheating AI, which can still beat novice players: https://techcrunch.com/2016/12/06/aragos-ai-can-now-beat-some-human-players-at-complex-civ-strategy-games/
4. Stupid AI. Starts with a lot of resources, but doesn't care to use them, instead attacking player with small manageable forces. Typical for most video games. I.e. if all monsters on the Doom map will suddenly go after the player, there will be no chance to survive. But monsters are too stupid to cooperate, so player can kill them one by one. Sad story. Later games like Quake Team Arena tried to make non-cheating bots, which were far less exciting than bullying the clueless Doom monsters.
5. Trigger-based AI. The AI reacts to player actions with unpredictable cheating, possibly random. I.e. if player moves near AI city, the otherwise docile AI will respond with full force of dishonestly earned resources. Pioneered in Doom games, with trigger sectors, unleashing monsters.

Ideally single player AI should react to player's advances and try to disturb them. So I will likely go with a trigger based AI influenced by the difficulty level and random dice roll. Such AI can have almost unlimited resources and act unpredictable, or according to some narrative. Although I still plan to bind AI resources to the actual income, just provide with larger starting resources. I.e. no gold will be materialized out of nothing.

134
Offtopic / Line of Defense Tactics
« on: June 18, 2020, 10:24:28 pm »
One of Derek Smart's last games is actually a real time XCOM clone. A few nice ideas, but really unfinished.

135
Offtopic / Re: What happened to Blizzard?
« on: June 14, 2020, 01:00:47 pm »
There is also the source code for the remaster:
https://github.com/electronicarts/CnC_Remastered_Collection

Apparently the game was originally written in C++ back in 1994.
Watcom compiler was used.

Guess now people can easily port it to any platform.

Pages: 1 ... 7 8 [9] 10 11 ... 13