Putting the 'role' back in role-playing games since 2002.
Donate to Codex
Good Old Games
  • Welcome to rpgcodex.net, a site dedicated to discussing computer based role-playing games in a free and open fashion. We're less strict than other forums, but please refer to the rules.

    "This message is awaiting moderator approval": All new users must pass through our moderation queue before they will be able to post normally. Until your account has "passed" your posts will only be visible to yourself (and moderators) until they are approved. Give us a week to get around to approving / deleting / ignoring your mundane opinion on crap before hassling us about it. Once you have passed the moderation period (think of it as a test), you will be able to post normally, just like all the other retards.

Vapourware Zodiac Legion - X-COM and dungeons

Darth Roxor

Royal Dongsmith
Staff Member
Joined
May 29, 2008
Messages
1,878,376
Location
Djibouti
Huh. I thought Invisiburu was running on its own engine.

#themoreyouknow
 

Galdred

Studio Draconis
Patron
Developer
Joined
May 6, 2011
Messages
4,338
Location
Middle Empire
Steve gets a Kidney but I don't even get a tag.
Huh. I thought Invisiburu was running on its own engine.

#themoreyouknow
They said it was a heavily modified one. That is one of the minor perks of moai. You don't have to write: "made with the personal version of Unity moai" on your splash screen. :)
It is buried somewhere in their credits.
 

Galdred

Studio Draconis
Patron
Developer
Joined
May 6, 2011
Messages
4,338
Location
Middle Empire
Steve gets a Kidney but I don't even get a tag.
Fog of War is on, but I need a way to prevent seeing through walls and I need smoother hex borders.

A1NME0O.png


I used another tileset to make the transition to black tiles better in the middle of the screen, but it would involve automatically checking the neighboring tiles for each visible tile, unless I find a cleverer way to do it. The "standard" result, without tile substitution is what you see on the right edge.
 

Reinar

Scholar
Joined
Oct 20, 2015
Messages
145
The highlighted hexes are strange looking and the characters look like they're not in the right angle by a small amount in comparison with the terrain, but otherwise nice job.
 

Galdred

Studio Draconis
Patron
Developer
Joined
May 6, 2011
Messages
4,338
Location
Middle Empire
Steve gets a Kidney but I don't even get a tag.
The highlighted hexes are strange looking and the characters look like they're not in the right angle by a small amount in comparison with the terrain, but otherwise nice job.
Would you recommend a subtler highlight for movement range? Or maybe just coloring the edges?

Your game looks pretty awesome from what I've seen so far (and I hope you played some Battle Brothers for good influence too :P )
Regarding this game and Battle Brothers:
I did play Battle Brothers, but the focus of Zodiac Legion is pretty different:
Battle Brothers plays more like Mount and Blade with Turn Based Tactics.
Zodiac Legion draws more from Eador when it comes to arranging things on the strategic layer, and Mordheim for the tactical part itself (with move points and action points), but with the objective based missions of XCOM2 (or Descent: Road to Legend for the ones who played the board game).

On the strategic layer, most things will be rather abstract (especially on the military side. ie no separate mini battle system ala Birthright). Military campaigns will be abstracted as sending X points of Might vs the opposing Y, and they will open missions for the heroes to be able to improve the results (they might assassinate the enemy wizards before a battle, or lead twenty good men and raid the camp , kill and impersonate or mind control the enemy scouts, poison the castle water supplies...).

On the tactical map, we want to have the tactical part play like special force ops in a fantasy setting, with the team going in, completing the mission, and going out before getting overwhelmed by reinforcements (or before the opponent completes his own objective), so this leads to
a mission structure more similar to XCOM 2.
The rationale for not going TU is to avoid crippling melee builds so that they can still move and hit things (ie charge), without requiring a special feat/skill or whatever.
We plan to keep the number of AP small to avoid letting one side get too many action before the other one can act, so Mordheim would be a pretty accurate comparison (although we have no plan to draw inspiration from their UI).
 

Sturmgewehr44

Novice
Joined
May 8, 2017
Messages
36
Location
Welsh Scenic Byway
I just discovered this game through your post in the "Fort Triumph" topic, but I'm excited to see more about this and I'll be watching this closely -- at this point I have no real feedback, so I'll just say keep up the good work, Galdred.
 

Galdred

Studio Draconis
Patron
Developer
Joined
May 6, 2011
Messages
4,338
Location
Middle Empire
Steve gets a Kidney but I don't even get a tag.
On the tactical map, we want to have the tactical part play like special force ops in a fantasy setting, with the team going in, completing the mission, and going out before getting overwhelmed by reinforcements (or before the opponent completes his own objective), so this leads to
a mission structure more similar to XCOM 2.

Sounds good as long as you keep in mind the flaws of XCOM2's mission structure and how it interacts with other systems. X2 vanilla's XP system incentivized players to prolong missions as long as possible and farm every loot, every corpse etc, even though the missions were meant to be guerilla/stealth stuff. LW2 did the opposite and incentivized stealthing every mission, leading to a samey experience.
That is a very good point.
In order to avoid incentivizing prolonging missions, I had decided to make all XP mission based, and not opponent based. It is also easier as we don't have to attribute XP for support and similar things now.
Only the mission outcome and its difficulty are taken into account.
Regarding stealthing vs brute force, Stealth will be handled off screen mostly (something like you choose the stealthy option on a text menu. A skill check is rolled, if you succeed, you start the mission in a better position, and with less guards, or slower reinforcements).
These options usually will only be available to some characters (not the guys in plate armor, mostly), or some missions.
But stealth might change a bit, even though I am pretty sure I don't want it to turn the game into a sub par Invisible Inc.
 

Galdred

Studio Draconis
Patron
Developer
Joined
May 6, 2011
Messages
4,338
Location
Middle Empire
Steve gets a Kidney but I don't even get a tag.
I am almost done with ranged attack. I am not too sure about what to do when you click to attack a target that is out of range: move to max range and engage? move to optimal range and engage? Do nothing and mock the player (well, probably not)?
Currently, they move to melee range and engage, which might not be the most sensible behavior.
 

Galdred

Studio Draconis
Patron
Developer
Joined
May 6, 2011
Messages
4,338
Location
Middle Empire
Steve gets a Kidney but I don't even get a tag.
btw, I added a totally unrelated blog entry about getting things done, especially when you work without a foreman:

DUNGEONS AND MONKEYS
Date dim. 23 avril 2017 Tags development / lifehack
THE MONKEY AND THE PANIC MONSTER

Some time ago, I had read a very interesting article about Procrastination on waitbutwhy, where Tim Urban described procrastination as an instant gratification monkey wrestling your decision-making process from your rational self.



I cannot recommend it enough, unless for some reason you are a mutant totally immune to procrastination (but then, it might still give you some insight about the way normal human beings work).

Note that procrastination is not just about picking playful things over work, it can "just" be a problem of selecting easy work over important hard work.

As many others, I got used to doing my assignments at the last minute in school.For a limited period of time, I managed to get acceptable results after entering panic mode. But it became worse and worse as time passed. I ended up writing my report for a 3 month research internship on the day it was due, sleeping 4 hours in 3 nights. My thesis was handed in 2 days late, and pretty messy.

Part of it was due to hopes that I would enter panic mode early enough to get things done, but that proved unreliable to say the least. Panic is not something you can really "manually" trigger. My first job was in an open space environment, and procrastination was not such a big problem, at least when browsing the web was not possible. But when I started working on my game, it became quite difficult to suppress the instant gratification monkey. There were so many reason to delay getting started on the code: I started writing a 50 page design document no one would ever read, after having read several dozens of fantasy series to get a better feel for world building. I also spent an awful lot of time trying different engines, to make sure I would have enough information to choose the most appropriate one. I just had to replay most of the classic games in this field to get a better feel for it (at least, I was able to avoid also playing the classics in other fields to "broaden my horizons"). The problem with working alone is that a lot depends on your motivation, which in turn is strongly driven by your recent results. This is very risky when you need to get things done, especially on such a long-term task as developing a game. I used to dread taking a holiday break whenever I was on a good working streak, but kids force you to take a lot of these.

I tried to enforce self-imposed deadlines, but they suffer from several issues:

  • The complete lack of consequence for missing one of these deadlines turns skipping them into another bad habit: how are self imposed deadlines useful when you don't even always respect external deadlines?
  • Deadlines can be "scary" if there is a lot of things to do in a very short time, which may encourage you to write them off as missed already. Or if they are too far away, you can easily convince yourself that you will enter overdrive mode "later" and spend a good part of your working time browsing web sites, or testing games that are really relevant to the one you are writing. Pretend there is still a lot of time to do it.
  • Rushing work is not something you should do on a regular basis (at least, I like to have my code readable and commented so that future me is able to take over from there without spending days trying to figure out what present me wanted to do).
It did not really work with me at all, as I wrote them off as either "too close to be acheivable", or "too far to start worrying about it".

I had a weird schizophrenic feeling where I was at the same time convinced that I would be perfectly able to complete the task if I started working on it soon enough, but I was also quite sure I would fail to dedicated enough time to get things done in a timely manner.

I used several methods both to make sure I was spending time on important tasks or features. Breaking task into small chunks. Using an Internet blocker. Putting a lot of post-its everywhere with the stuff I had to do. But these all failed to change things in the long run.

THE TOMATO OF WILLPOWER
I used several time tracking softwares to monitor the use of each application, but it was hard to measure the amount of work done. It ended up only giving negative feedback(ie tracking the time spent not really working), instead of providing a goal to strive for.

Then the pomodoro technique completely changed the way I looked at my work.

The technique itself is pretty simple: you break down your work into 25 minutes intervals, with 5 minutes rest between each, called pomodoro (poms). Each pomodoro is dedicated to a single task (it can be a generic one, and you can bunch smaller tasks together in a pomodoro). No interruption is allowed. You just tell people that you will get back to them, and you take note of whatever it is you wanted to do if you feel the urge to do something else. If you have to interrupt the pomdoro, it is voided, and you need to start again. Obviously, you can dedicate several poms to the same task.

These pomodoros can also be extended to other tasks: Many also use them for chores.



When I first tried this method, a few years ago, I was not very convinced: I was working from home in Shanghai, and my 2-year-old daughter required a lot of attention from me, so the distractions were very frequent and hard to ignore. But now that the girls are older, I have found the results pretty amazing, and I cannot recommend it enough if you are in a similar situation.

EFFORT ORIENTED APPROACH
As long as you don't allow yourself to do anything else than whatever was planned for the Pomodoro, you can easily keep track of the time really spent working on the given task. Unlike deadlines or result-oriented methods. Deadlines had stopped working for me. So taking an effort-oriented approach instead of a result-oriented one has yielded much better results. You can misjudge your ability to complete a given task, but the time spent on something is a metric you cannot really argue with. I have found that having an objective metrics to measure your efforts really helped.

Here is a chart to guide you:

Poms level
0-4
Abysmal Not even trying
5-8 : Poor day Boring work?
9-12 : OK Your standard average day
13-15 : Very good 13-15 avg is doable, but hard
16+ : Stellar Very hard to sustain over a long period
The amount you can do will of course depend on your fatigue, on the boredom level of the task, and many other parameters. It also assumes that you don't have a job where you are supposed to be interrupted by others. I usually settle for a goal of 65-70 pomodoros/week on average.

REVERSE PROCRASTINATION
Whenever you feel the urge to do something unrelated to the task at hand, like checking whatever other people think on the Internet, it is easy to postpone it for a duration of 25 minutes, especially if you have to cancel "scoring" for all the work you have done in the pomodoro up to this point. More often than not, the urge to check whatever came to your mind during your pom will have disappeared after, and even if it doesn't, you won't feel as bad checking it. In a way, this works like reverse procrastination: you tell yourself you will do the unimportant stuff later, and end up not doing it. I also found it makes it easier to tell other people not to disturb you while you work. But everyone reacts to stimuli differently, so if you prefer working 20, 30 or 55 minutes Poms, or prefer to skip the break and chain 2 together, you should follow what works best for you (I personally usually work 2 pomodoros together and skip the break in the middle to stay in flow). These are only there to give you guidelines. And even if you are a "light" procrastinator who only does things in the wrong order, (ie unimportant work before important one), you can still benefit from this technique by dedicating several pomodoros a day to the task you know to be the most important in the long run. If you are a perfectionist, it can help you force yourself to dedicate part of your working time to other things that needs to be done, instead of improving things that are already good enough.

ADVENTURING WITH THE MONKEY
I tried to gamify my working hours a bit. I first tried running ProgressQuest when I was working, but it turned out not being interactive enough for me. It also didn't really help me keep track of anything. I also tried Pomodorium, but it was a bit too involved for me, with a mini Role-Playing Game combat system.



Then I found Habitica, which works as a task/habits manager, with a game layer on top of it: You get XP for doing stuff you set as good, completing items on your todo list, and you lose hitpoints for doing bad stuff, or not completing your daily tasks.



As you can see, it doubles as a tracker for all the stuff you are supposed to do or avoid doing. It has 3 categories: habits, dailies, and Todo. The habits is there for habits you want to build or suppress. Depending on the sign, you either lose hitpoints and mana or get gold and XP for following them. Dailies work differently: they are things you need to do everyday, and skipping any of them results in a loss of hitpoints and mana at the end of the day. And Todo is just your usual todo list. I find it much tidier than post its, and the daily/habit tracker can help with other parts of your life, like doing some exercice, flossing teeth, going to bed early... The gold you get by doing that allows you to purchase in game weapons or armor, or you can add any custom rewards, like going to the movie theater, or whatever.

The game elements are rather light, you just have powers you can use against boss opponents, and otherwise, completing items on your todo list, checking your daily tasks, and following good habits all damage the boss. But even if you are not into RPG, there are challenges that can keep you motivated.

It works wonderfully, because it let you set deadlines every day with things that you can perform (I have a fixed amount of pomodoros to do daily), thus reinforcing the effort oriented approach, and there are consequences to missing these daily deadlines (loss of hitpoints), and rewards to completing them (XP, gold, and mana), so it also tricks the monkey into doing things that make sense in the long run, as he likes to go adventuring too.

monkey_quest.png

The monkey likes killing monsters sometimes

It also adds some peer pressure, because missing daily tasks when facing a boss monster results in everyone in the party losing hitpoints. The best part is the monkey also gets to play!

My 6 year old daughter has been enthusiastic as she saw me using it, and wanted me to create her an account. She once asked me to skip the bedtime story to study maths, because there was no reward associated to listening to stories, while problems are worth a lot of points. So I am quite happy with using it with here too. It even has a pokemon skin:



Of course, it is quite easy to game the system, as you are more or less setting your own rewards and evaluating yourself. Just don't do it. And you also need to care one way and another for this kind of rewards, so it might not work very well for people who are averse to Role-Playing Game type progression and don't care about prettying their character.

DISCLAIMER
Procrastination takes different forms, and feeds on different feelings for everyone, so what I wrote there might not be directly applicable to you. When I first read about procrastination, I was convinced that all of this was irrelevant: the tips given to suppress it did not work on me, and my first attempt to use the pomodoro technique was quite underwhelming. But trying to find new ways to build good habits should pay off in the end, even when these methods don't help directly.

WHERE TO GO FROM THERE?
Tedd talk from Tim Urban about procrastination

Reference page about the pomodoro technique

Pomodairo, My Pomodoro timer

lifehacker on time tracking applications

Habitica, the self improvement RPG

College Info Geek video review of Habitica
 

Reinar

Scholar
Joined
Oct 20, 2015
Messages
145
The highlighted hexes are strange looking and the characters look like they're not in the right angle by a small amount in comparison with the terrain, but otherwise nice job.
Would you recommend a subtler highlight for movement range? Or maybe just coloring the edges?
I don't remember what I was thinking about it back then, but yeah, subtler (more transparent) highlight would be better.

I am almost done with ranged attack. I am not too sure about what to do when you click to attack a target that is out of range: move to max range and engage? move to optimal range and engage? Do nothing and mock the player (well, probably not)?
Currently, they move to melee range and engage, which might not be the most sensible behavior.
Just let the player know he's out of range and let him decide.
 

Darth Roxor

Royal Dongsmith
Staff Member
Joined
May 29, 2008
Messages
1,878,376
Location
Djibouti
Just let the player know he's out of range and let him decide.

Yeah, this. I find that more often than not, something taking control from the player and doing things "automatically" leads to many unwanted situations and missteps.
 

Goldhawk

Goldhawk Interactive
Developer
Joined
Feb 2, 2015
Messages
63
This game looks pretty neat - I like the unit sprites in particular. It's kinda nice to see something go through the shitty pixel art stage and then come out the other side into something that actually looks professional, so well done with that.

BTW, I saw you mentioning a struggle with publicity etc in the Fort Triumph thread - feel free to drop me a PM if you feel you need some advice on publicity or a Kickstarter etc, I've had some experience doing that sorta stuff.
 

Galdred

Studio Draconis
Patron
Developer
Joined
May 6, 2011
Messages
4,338
Location
Middle Empire
Steve gets a Kidney but I don't even get a tag.
This game looks pretty neat - I like the unit sprites in particular. It's kinda nice to see something go through the shitty pixel art stage and then come out the other side into something that actually looks professional, so well done with that.

BTW, I saw you mentioning a struggle with publicity etc in the Fort Triumph thread - feel free to drop me a PM if you feel you need some advice on publicity or a Kickstarter etc, I've had some experience doing that sorta stuff.
Thank you! That is very nice to hear from you. I will pm you soon.
 

Galdred

Studio Draconis
Patron
Developer
Joined
May 6, 2011
Messages
4,338
Location
Middle Empire
Steve gets a Kidney but I don't even get a tag.
I got a question on the Steam Turn-Based Tactics group about modding and customization, so here is the answer:

Customization

Graphic customization
- Each armor type will match several piece of armor sprites (basically, light/medium/heavy, normal/magic, with several colors to choose from).
So the player will be able to match pieces with each other within certain limits (no leather for plate armor and so on).
- The helms will have to remain in line with a given zodiac sign, so there will only be 3 to choose for a given sign.

Gameplay customization
To be honest, his is not set in stone yet, as there are not many perks in currently, and we are still iterating on the core gameplay, so it will probably change a lot during testing. The idea would be to have around 50% passive skill or stat increases, and 50 % active skills (or active skill upgrade/sidegrade) on level up.
Some will be random, and other will be decided by the player.

Random skills will probably be given at the lower levels, because it would be a bit pointless to spend time skilling up newbies that will be sent to the meat grinder, and it makes it easier to plan the character build around given random perks than to have it compatible with all options.

Given that the player is leading a small military order, hero characters will usually be supplemented by regular soldiers or spies. These will basically act as level 0 units that can be recruited to the player roster of elite soldiers after a mission.

Character creation
In order to make it viable to sacrifice characters to complete mission objectives, there will be no character creation: Your recruits will have a random perk in the beginning (cf gameplay customization).
You will play the commander of the Zodiac order, but you can die. In this case, you will switch to the next one in charge, like in Crusader Kings.
You lose when the frotress of the order falls to the enemy, not when the commander dies on the battlefield.

Stronghold customization
Unlike in XCOM, there won't be a strict grid to let you arrange your base to maximize defense (X-COM) or adjacency bonus (XCOM). Instead, you will have several facilities to build or upgrade, and several spots in the citadel, inside the city walls, or outside.
It will determine how easy it will be for the enemy to raid each of your facilities.
The stronghold itself can be sieged and assaulted, so some of the inner building slots might be reserved to bolster its defenses.

Building can then be upgraded by either using upgrade schematics to magically enhance them, or assigning characters or special rewards to them (like a 24/24 worker golem).

Modding
The game is made with moai, so most of the game logic uses lua (except for a few calculations that are made in C++, like line of sights and pathing). It means the scripts can be modified by anyone, but it may be a bit complex (I will try to keep data separated to make it easier).
The maps are made with tiled, an open source map editor (I plan to move from maps assembled from preset parts made with tiled in the future, like X-COM or XCOM2).
So these will be moddable. The only problem would be the documentation, and the need to be proficient with lua.
For the graphics, it will be much trickier, as I use Texture packer, and I cannot really ship it with the game.
The graphics <-> game data file will be editable, but the character sprite sheets themselves would require texture packer.
That said, I doubt anyone will be willing to animate each character frame by frame for 20 to 40 frames/sec, so I don't think texture packer would be a big issue.
So I would say that modding missions would be easy.
Changing scripts would be much harder, and adding new pieces of equipment would require even more work.
 
Last edited:

Galdred

Studio Draconis
Patron
Developer
Joined
May 6, 2011
Messages
4,338
Location
Middle Empire
Steve gets a Kidney but I don't even get a tag.
btw, I added a totally unrelated blog entry about getting things done, especially when you work without a foreman:

Try it on hardcore mode: No games, no Internet, no TV, for 7 days.
But the thing is, I intend to keep gaming. I think one of the problems with game designers nowadays is that they don't play other games enough. So not ever doing it before having completed my objectives for the day works better (I just need to stop before 2 AM...).
No gaming outside of week-end also works well.
I don't watch TV already, and no internet would make it pretty hard to market the game, or find solutions to problems already solved by others. :cool:
 
Self-Ejected

Davaris

Self-Ejected
Developer
Joined
Mar 7, 2005
Messages
6,547
Location
Idiocracy
btw, I added a totally unrelated blog entry about getting things done, especially when you work without a foreman:

Try it on hardcore mode: No games, no Internet, no TV, for 7 days.
But the thing is, I intend to keep gaming. I think one of the problems with game designers nowadays is that they don't play other games enough.


So not ever doing it before having completed my objectives for the day works better (I just need to stop before 2 AM...).
No gaming outside of week-end also works well.
I don't watch TV already, and no internet would make it pretty hard to market the game, or find solutions to problems already solved by others. :cool:

That's a interesting theory. If you watch the Tim Cain interview, he said he was in sensory deprivation mode, for around six months. Possibly more. They had no life outside of Fallout.

Marketing is a problem, that's something I'd need Pomodoros to do. :P

Getting help can be hard, depending on the engine you use. The engine I use is really easy though, and the help is inside it.

I thought I'd mention it because my connection went out for a week. It was like coming off a drug + I got a lot of work done.
 

Galdred

Studio Draconis
Patron
Developer
Joined
May 6, 2011
Messages
4,338
Location
Middle Empire
Steve gets a Kidney but I don't even get a tag.
btw, I added a totally unrelated blog entry about getting things done, especially when you work without a foreman:

Try it on hardcore mode: No games, no Internet, no TV, for 7 days.
But the thing is, I intend to keep gaming. I think one of the problems with game designers nowadays is that they don't play other games enough.


So not ever doing it before having completed my objectives for the day works better (I just need to stop before 2 AM...).
No gaming outside of week-end also works well.
I don't watch TV already, and no internet would make it pretty hard to market the game, or find solutions to problems already solved by others. :cool:

That's a interesting theory. If you watch the Tim Cain interview, he said he was in sensory deprivation mode, for around six months. Possibly more. They had no life outside of Fallout.

Marketing is a problem, that's something I'd need Pomodoros to do. :P

Getting help can be hard, depending on the engine you use. The engine I use is really easy though, and the help is inside it.

I thought I'd mention it because my connection went out for a week. It was like coming off a drug + I got a lot of work done.
I don't think crunch mode is very healthy :)
With kids, no life would not be an option anyway. That said, late in the development, it is true that the value of playing other games diminishes, but heh, maybe the lessons learned will serve later, and I have little concentration left after 12-13 pomodoros anyway :)
I agree, I could not do community management/marketing or accounting without using pomodoros.
The moai comunity is very small, but helpful, so getting by asking other devs directly does work, but hoping that the solution could be googled is usually futile. But most of the things I look for are not related to the engine anyway. It is mostly algorithms (for Line of Sight, procedural generation, AI...).
I used to work in a company without Internet before, but I had to take walks when my concentration was dropping too much.
 
Self-Ejected

Davaris

Self-Ejected
Developer
Joined
Mar 7, 2005
Messages
6,547
Location
Idiocracy
But most of the things I look for are not related to the engine anyway. It is mostly algorithms (for Line of Sight, procedural generation, AI...).
I used to work in a company without Internet before, but I had to take walks when my concentration was dropping too much.

An ISP that offers a service where you can book times you want Internet access, would work nicely with that. Unfortunately I've heard of no such service. Taking walks would be better than aimless surfing or news reading, as your brain can rest and wander, instead of being given more noise to process.
 
Last edited:

Galdred

Studio Draconis
Patron
Developer
Joined
May 6, 2011
Messages
4,338
Location
Middle Empire
Steve gets a Kidney but I don't even get a tag.
But most of the things I look for are not related to the engine anyway. It is mostly algorithms (for Line of Sight, procedural generation, AI...).
I used to work in a company without Internet before, but I had to take walks when my concentration was dropping too much.

An ISP that offers a service where you can book times you want Internet access, would work nicely with that. Unfortunately I've heard of no such service. Taking walks would be better than aimless surfing or news reading, as your brain can rest and wander, instead of being given more noise to process.
You can use stayfocusd, or a similar app to limit browsing. You can whitelist the sites you want, so it is pretty flexible. I think blacklisting everything, and whitelisting sites like stackoverflow and gamedev.net works better than the other way around. But I feel much less urge to browse news now that I use the Pomodoro technique anyway, as it usually fades away by the time I am done with the pom.
 

Galdred

Studio Draconis
Patron
Developer
Joined
May 6, 2011
Messages
4,338
Location
Middle Empire
Steve gets a Kidney but I don't even get a tag.
Ouch, it has been quite a long time. I am currently working on teaching the AI how to dtermine sensible areas. I know I could just do it by hand, but:
1) where would be the fun in that?
2) it is not scalable, so it would prevent me from adding procedural generation in the future (ok, I would only proc gens dungeons from rooms, but still).

I am using something close to what was done in Kohan 2 iirc:
Step 0: "remove" small obstacles like pillars, trees, to get large areas
Step 1: get a depth map of the level:
bdPZSxK.png

The depth is the distance to the closest obstacle.

Step 2: make a 1st rough pass at deducting an area map from it:
Areas or basically : groups of tiles for which the depth = the starting depth or is strictly deacreasing (so it can be 6-6-6-5-4-3-2, but not 6-6-6-5-5-5 and not 6-6-6-5-4-3-2-3).
Of course, the rules are a bit more complex.

xI6KXoa.png


Step3: correct the areas by swaping border areas that create non regular shapes

Step4: try to make the areas "more or less convex": ie, not like area 57 here:
It cannot be strictly convex, because our walls are bizarrely shaped. So it must accept that we consider our rooms convex :)

h2f8a3h.png


In order to do this, I use the following rule:
If any border hex that is not in the area is next to 4 area tiles or more, or to 3 tiles that are not contiguous, then it doesn't pass the test.

Zj0MdoH.png

Then check whether any area hex crosses the line in white on the image above (they have the same shape as our walls). If one does, it fails the test, otherwise, it passes.
Then the AI determines how many tiles are beyond the white border. If it is "too many", it tries several splits of the area (from the first tile that did not pass the semi convexity test), and determines which one works best.



So that's it basically (it took me an awful time to determine the various rules and parameters to use).
Why do this at all?
Because it is much more convenient to dispatch AI squad 1 to room 32 that is on the path between the heroes and the objectives, than precisely determine on which tile we want to send NPC 21 when said tiles are 5 turns away at running speed.
It also makes it easier for the AI to understand which areas it controls, and which ones are under player control (count the total levels in each room for each side, and then "diffuse" these influence values to neighboring rooms).
 

Nathaniel3W

Rockwell Studios
Patron
Developer
Joined
Feb 5, 2015
Messages
1,225
Location
Washington, DC
Strap Yourselves In Codex Year of the Donut Codex+ Now Streaming!
Oh crap! I totally just spammed my own game in somebody else's thread. Sorry!

Seriously, I thought this was a different thread. I'll edit my previous post and go put it where I meant to...

You're doing a great job Galdred keep it up! I'm totally envious of your neural-net-driven AI. Maybe when I have an actual game ready I'll come back and ask you for tips on making my AI smarter.
 

As an Amazon Associate, rpgcodex.net earns from qualifying purchases.
Back
Top Bottom