Warzone 2100: Contingency (Beta now released)

[aubergine]

I found the Cyborgs to be OK if they get enough armour upgrades, but after a while they are easily wiped with a bunch of TBMs.

[NoQ]

Shadow Wolf TJC: could you increase the timer of checkEndConditions in rules.js? Probably to 1000ms or something like that.

Cause a single call may take quite a long time on my machine (as much as 30ms in a 2-player game on a small map, so calling it every 100ms will slow the game down by 30% ... as you know, the game is completely paused when a script event is being handled). Not sure which part of it takes that long though.

P.S. I figured that out by running the mod on master (where i can get function lag displayed), which means i made some body.ini file after all. It is auto-generated, so may have some tricky little errors, but seemed to work so far. Probably you could include it in the mod? And [make me] re-generate it when you change the bodies.

[Shadow Wolf TJC]

@NoQ: Thanks for the body.ini file. Which version of Contingency did you use to auto-generate it?

I've adjusted the checkEndConditions timer for the next version as you requested. Hopefully, it'll be better.

@Giani: I fear that you're focusing too much on the Cyborgs' weaknesses over vehicles (such as their low HP) and underestimating their strengths over vehicles. That is, you're underestimating just how cheap and quick-to-produce they are compared to vehicles.

For example, a Heavy Gunner costs only $8, while a Super Heavy Gunner costs $20, a Light Cannon Bug Half-Tracks costs about $42, and a Light Cannon Viper Tracks costs about $50. So for the price of a single Light Cannon Viper Tracks, I could build 6.25 Heavy Gunners, or 2.5 Super Heavy Gunners.

Recently, I defeated an opponent using primarily Super Cyborgs armed with Mini-Rocket Pods. My opponent was focusing too much on high damage, low rate-of-fire all-rounder and anti-tank weapons and not enough on low damage, high rate-of-fire anti-personnel weapons, so I was able to defeat my opponent's forces without any problems. Sure, I was losing these Cyborgs quickly, but since they were so cheap and so quick to produce (even with only 10 derricks in my possession) that by the time that my opponent killed a horde of them, I had another horde or 2 ready for battle. Had my opponent switched to producing units and structures armed with rapid-fire weapons (such as machineguns) and/or weapons that dealt large amounts of splash damage (such as Mortars, MRAs, or the Blunderbuss), then I would've switched to producing tanks.

@aubergine: Those same TBMs would've probably wiped out large numbers of light tanks as well.

Edit: I'm thinking about wanting to try to create a Factory Module for the Cyborg Factory so that their production could be increased up to that of a vehicle factory with 2 modules. I wonder if that's possible...

[NoQ]

@NoQ: Thanks for the body.ini file. Which version of Contingency did you use to auto-generate it?

0.3.2.

[NoQ]

Updating the AI to v0.9. There were not many Contingency-specific changes, but it includes updates of the base scripts from NullBot v1.33.

[Shadow Wolf TJC]

Version 0.3.3 is out, with a few new turrets:

These Quad Minipods are twice as heavy as Twin Minipods, and have twice the rate-of-fire. The turret on the far-left is the new model for the Rocketstorm, which was made even more powerful, and more heavier, than before.

I tried to allow players to build factory modules onto Cyborg Factories so that they could spend as much power on producing Cyborgs from a single Cyborg Factory as they could on producing vehicles from a single vehicle factory with 2 modules, and so that I could require players to build modules onto Cyborg Factories in order to allow them to produce Super Cyborgs. Sadly, I couldn't figure out how, so I instead had to settle with increasing the rate at which Cyborgs are produced from Cyborg Factories to be comparable with that of vehicle factories with 2 modules (though I also had to triple their price and build points to balance that out).

I also moved the AA Flak Cannon and Sunburst AA from early T2 to late T1. Hopefully, this would give players a better chance at researching them before, or soon after, VTOLs start to pop up. This update also comes with a few new things, including a body.ini file for compatability with master versions of Warzone 2100 (thanks to NoQ), and scripting support for what I'd like to call "tier limiting".

What is tier limiting? Well, I added 2 new research items, R-T1Limiter (which is a prerequisite for all T2 technology) and R-T2Limiter (which is a prerequisite for all T3 technology), and some code to rules.js that allows it to read 2 new .js files that I included with contingency.wz: tier1limiter.js and tier2limiter.js. By default, these 2 .js files do nothing. However, by loading other mods with custom tier1limiter.js and tier2limiter.js files of their own, they can potentially tell the game to actually limit what tiers of technology are made available, by determining whether or not players would start with R-T1Limiter and/or R-T2Limiter already researched for them. I've bundled a couple optional mods, T1Limiter.wz and T2Limiter.wz, each of which contain their own modified versions of these .js files to overwrite contingency.wz's versions, with the Contingency mod.

Note: While tier-limiting is disabled by default, in the event that it's enabled via an external mod (such as with one of the tier-limiting mods that were bundled along with the Contingency mod), AI developers will need to program their AIs to perform checks to see whether or not they already have R-T1Limiter and R-T2Limiter (which cannot be researched at all in-game) researched before they even consider trying to research T2 or T3 technology. Otherwise, the game may freeze and/or crash as the AIs try (in vain) to figure out how to research these 2 research objects.

[Iluvalar]

I tried to allow players to build factory modules onto Cyborg Factories so that they could spend as much power on producing Cyborgs from a single Cyborg Factory as they could on producing vehicles from a single vehicle factory with 2 modules, and so that I could require players to build modules onto Cyborg Factories in order to allow them to produce Super Cyborgs. Sadly, I couldn't figure out how, so I instead had to settle with increasing the rate at which Cyborgs are produced from Cyborg Factories to be comparable with that of vehicle factories with 2 modules (though I also had to triple their price and build points to balance that out).

2 modules gonna be looong. I did the same but with only one module in NRS. I believe it's enough. You might also need to increase a bit the price for the "embedded CC" effect. If it still there in 3.1 ? And if the cyb fact is close enough to the start to make players able to skip the CC reasonably until that.

[Goth Zagog-Thou]

I did something pretty neat (I think it is, anyway). I ported the latest one to FlaME (not difficult, btw). It seems there are no Cyborg or Super Cyborg bodies or propulsions that it can find. You planning to add these soon, by chance?

[Shadow Wolf TJC]

@Goth: Although the Contingency mod uses many custom-made .pie files and textures, it also uses many .pie files and textures from vanilla Warzone 2100. In order for FlaME to work properly with Contingency data, you'll need to port the following:

- names.txt
- all .pie files, including from other objectData folders (though you should add those 1st, then add the Contingency .pies, overwriting the other objectData folders' .pies if necessary).
- all texture pages, including from other objectData folders (though you should add those 1st, then add the Contingency texture pages, overwriting the other objectData folders' texture pages if necessary).
- assignweapons.txt
- body.txt
- bodypropulsionimd.txt
- brain.txt
- construction.txt
- ecm.txt
- features.txt
- propulsion.txt
- repair.txt
- sensor.txt
- structures.txt (Note: You may need to make some adjustments to the .pie files used by the factories, power generators, research labs, and walls in order for them to show up properly in FlaME.)
- structureweapons.txt
- templates.txt
- weapons.txt

[Shadow Wolf TJC]

I'd like to conduct a survey on what kinds of strategies and playstyles you guys like to employ from the early-game and onwards. For example, do you guys prefer to start the match with a Machinegun Viper Wheels rush (and possibly later graduating towards, say, a Light Cannon Bug Half-Tracks rush, and/or a Cyborg rush)? How would you counter said rush? Do you guys prefer to tech-rush in an effort to improve the quality of your forces at a faster pace than your opponents? If so, then how would you protect yourself from enemy forces? What kinds of technologies, including weapon lines, do you prefer to focus on, and how would you reach them? Do you guys prefer to focus more on survivability, firepower, or both equally?

[aubergine]

While I've become much more aggressive in my game play, I'm still a turtler at heart and often revert to that if I just want to mess around in WZ to pass some time aimlessly. To that extent:

* I like to build walls to prevent enemy getting to my base, plus regular wall upgrades
* Then I build up some sensors and basic defence structures - starting out with MG towers, sometimes moving to mortar line
* I aim for Seraph missiles ASAP after that, sometimes building lancer towers/borgs on the way
* Finally, I build a fleet of MRA Cobra Hover or MRA/Seraph Python Hover and then go kill the enemy AIs

When I'm not turtling, my strategy is... well, I don't have a strategy. I just frantically do whatever seems most obvious at the time! Although usually it involves MG -> rockets. I very rarely use cannons.

Note: What I describe in this post is based on standard WZ tech tree / config. I'm still trying to wrap my head around all the new stuff in Contingency!

[Arreon]

Can you do something with the scripts which makes Nullbot avoid "giving up" and doing nothing at all?

Every time I play a game, halfway through it, it just stops everything it was doing, units included and sits there for the rest of the game.

[Shadow Wolf TJC]

Well, I decided to do a match on Vision (probably one of my favorite maps to play on) against the default NEXUS AI on Medium difficulty (I know, it doesn't compare to Nullbot ), this time focusing on winning with Flamers. Because Vision contained lots of ground that would slow down most ground units, including Cyborgs, and since the Flamers were placed further back in the tech tree than normal, I had little choice but to turtle the center and tech rush until I had Flamers, and later, hover propulsion.

I had initially planned to do a small truck rush followed by an early Machinegun Viper Wheels rush to get some units out on patrol, and to quickly cap the center oil and build Tank Traps to stall for time. While I had initially planned on doing a FRRFC build, I saw that I had enough trucks for this purpose, so just as I started work on the 2nd factory, I immediately switched to building a Command Center (effectively turning into a FRRC build) so that I could start producing some Machinegun tanks. (I resumed work on that 2nd factory soon after.) Since the Flamers line was moved back further, I needed to research rush to get to them, so I produced 3 more Research Labs for a total of 5 Research Labs.

Given that it's NEXUS, I managed to capture the center oil before NEXUS could, while my Machinegun tanks finished off his trucks, and I managed to keep his masses of Machinegun tanks, and later his Cannon and Rocket tanks, at bay with a Tank Trap barrier supported only by a couple Machinegun Towers, a couple trucks, and a Repair Facility, until I could research Flamers.

Once I had Flamers, I proceeded to start churning out Light Flamer Bug Half-Tracks and Cyborg Flamers (since NEXUS was focusing on churning out mostly Cannons and Minipods) on repeat build, and I decided to build a couple Flamer Emplacements onto each of my chokepoints. Normally, these emplacements wouldn't seem to have any use on this map due to a lack of vision-obstructing hills on the map, though since NEXUS seems to like to drive straight up to my wall of Tank Traps, they were the perfect defense. However, I continued to focus on tech-rushing for hover propulsion (and Composite Alloys, heavier flamers, and the Scorpion and Mantis bodies since the Cyborgs and light vehicles were quite fragile against masses of enemy units) since Vision was full of terrain that would slow down most ground units, including Cyborgs, so despite me securing 12 derricks for myself, I wasn't able to effectively push into my opponent's base (seeing as how such terrain would have the worst effect on shorter-ranged units like Flamers) before I could research and start building hovers (which weren't slowed down by the terrain), but boy, did I cause alot of carnage to NEXUS's countless forces.

Soon after, I stormed NEXUS's base with my horde of Flamers, and although it took them a long time to destroy NEXUS's base structures (due to them being ineffective against structures, except for Bunkers), I had won the game with no problems. However, if my opponent had turtled with some hardpoints, towers, or emplacements, or had likewise focused on tech-rushing for hover propulsion, then I would've had to wait until I could produce Lasers, which would've outranged NEXUS's Cannons or Rockets.

[NoQ]

Can you do something with the scripts which makes Nullbot avoid "giving up" and doing nothing at all?

Every time I play a game, halfway through it, it just stops everything it was doing, units included and sits there for the rest of the game.

Can you post at least one savegame where it has "given up"? I tried to fix that problem in the latest version after themac posted some savegame (after 20 pages of me begging for at least any way to reproduce this problem i have never seen), are you saying there are other cases when it happens?

[Goth Zagog-Thou]

Yep, that's pretty much what I did to the ported dataset. I was asking specifically about Cyborgs and Super-Cyborgs, cause there's no body parts for those, and thought you were still working on Cyborg stuff for Contingency. It was the initial attempt to see "how ready" Contingency was for Cam 4.

If you've no plans for Cyborgs, that's fine too, and I'll just keep Cam 4 on it's own dataset.