/lv/ - Libre Vidya

I really like how polished it feels but it is a little too random for me.

It's great if you have someone to play with. Do not ever play against AI, it fucking cheats like a motherfucker

>>877

Apparently there are a couple mods to get rid of the random shit. One of them makes hits always land and instead multiplies the damage by hit percentage and rounds up. This can unbalance units with low damage but lots of successive hits like cuttle fish though, so there's this other deterministic mod that seeks to address that, but I don't really understand how that one works.

I'm actually becoming quite interested in making a comprehensive mod for Wesnoth that ensures all attacks hit but proper damage is done so the game is still as balanced as ever. The questions I just need answered are whether or not it's possible to make the game allow fractional damage amounts, and whether or not you can mod stuff like the UI elements to display the right numbers when the attack prompt shows up (for example I would like to change a unit attack with 14-4 at 70% chance of hitting to simply say 39.2 ).

Then the only other real obstacle is how to keep special effects like Slow and Poison balanced. The probably of them working may have to stay randomized but at least attacks would do expected damage.

I need to find out if all this stuff can be modified with an add-on.

>>882

*probability of them working

>>882

That sounds really good. Have you tried sending the devs an email?

>>885

I talked to a few in their IRC channel now and it seems that it sadly isn't possible to do fractional damage–all of Wesnoth's damage is calculated by integers. Perhaps scaling health by a factor of 10 is enough to deal with the imbalance generated by rounding damage values however.

Having the status effect of a special attack applied with a probability of happening after the initial damage is done definitely seems possible.

Only other obstacle now is I apparently have to learn two different new programming languages to mod Wesnoth now (LUA and WML) and I'm not sure where to start (got some experience in other languages though so I'm not clueless). Once that's done this really shouldn't be too difficult a mod to put together.

One other thing I'm considering with this mod is having two different modes. One mode would retain the number of attacks that each unit has individually. Another mode would change every unit to use a single attack since, without hit probability, having multiple attacks no longer serves a purpose. Due to damage rounding this would probably be a bit more balanced too. The second mode may take a fair bit of work if I have to tweak every unit individually without the aid of a simple universal script.

So apparently multiplying everything by a factor of 10 isn't what really matters (though it can help a bit). What really helps the most is going to have to be combining the sum of fractional attacks together before the rounding operation is done.

Here's a Cuttle Fish example examining how its tentacle move changes with different damage approaches. Tentacle attacks 10 times for 3 damage each. Let's say a Cuttle Fish attacks a unit on a tile with 60% defense (40% probability of hitting) and 30% Resistance to impact damage (70% damage):

Expected damage of classical Wesnoth system provided the RNG behaves itself perfectly = RoundUp(3 * 0.7)(10 * 0.4) = 12

Damage with attacks summed together after defense and resistance effects = RoundUp[RoundUp(3 * 0.7) * 0.4] * 10 = 20

Damage with attacks summed together after resistance applied but before defense = RoundUp[RoundUp(3 * 0.7) * 10 * 0.4] = 12

Damage with attacks summed together after resistance applied but before defense and damage scaled by 10 before final rounding operation = RoundUp[RoundUp(3 * 0.7) * 10 * 0.4 * 10] = 120

>>901

So basically what really needs to be done to balance damage in this modification is to reduce all damage exchanges to a single instance. I'm trying to decide if I want to turn every unit attack into a single hit or perhaps leave them in but only make the final hit display damage.

I've just tried it out, it's very fun.

>>882

Did you manage to make the mod yet?

>>921

I've familiarized myself with WML and Lua now; it's just a matter of learning which of Wesnoth's internal functions I need to call up within Lua in order to work with them. This is turning out to be unfortunately the most difficult part. Most Wesnoth mods involve scenarios and campaigns so sadly the documentation for this sort of thing isn't as straightforward as it should be.

>>922

You could ask the guy who did the no randomness mod, Dugi, I think?

He's still around on the forums.

>>924

Yeah I PMed him about this project and he doesn't really care about his mod or even me using his code. That mod was just a cynical effort to get people to keep quiet about randomness from the beginning; he actually likes the randomness of the game.

>>925

>he actually likes the randomness of the game.

The entire forum community over there have accepted it. I guess since they have been working and playing with it for YEARS, they arent likely to change.

But having got into this game like a year ago, I think the randomness is kind of holding it back. Its not even like fire emblem or something, where you can arm or train units to have better hit rates.

And considering how time consuming some of these campaigns and battles can be, it seems a bit unnecessary in terms of difficulty.

The concept of a free community built SRPG is awesome but the random gameplay is kind of holding it back in term of widespread appeal, while the veterans who have gotten so used to it are unwilling to budge to welcome new blood.

>>902

Ugh, damnit. I just realized a problem this might cause. In defensive exchanges where a unit should normally kill the other before they themselves get killed this will create some sort of imbalance.

I'm really not sure if it will be possible, but maybe… I could write some sort of conditional that calculates damage for each unit concurrently and then somehow transfers the correct damage when a unit is going to die?

>>927

For instance,

Let's say there are two land units in shallow water with only 10% defense. Let's say the attacking unit has five attacks that do 5 damage each and the defending unit has 11 HP left. While the defending until does enough damage to kill the other unit in a single hit. Normally this would all but guarantee a kill for the defender before the attacker can manage to kill them, but if all the damage is summed together in a single exchange this mod is going to imbalance the game in favor of attacking.

I think what I'm going to need to do is call up the Wesnoth function for unit health and build that into a sequential loop that tests whether or not a unit's health is depleted, and if it is then the damage summing is halted and the surviving unit gets dealt the damage they would be expected to receive in a multiple-attack exchange.

>>928

So how far have you gotten with it?

>>937

Not too far, sorry. I've been a bit busy/lazy.

Man you guys are pussies.

Would RISK be a good board game if there was no random elements? Would any board game with dice or cards (hint: fucking almost all of them) be worth playing with no random elements?

Fuck no.

Harden the fuck up.

>>1006

Risk mitigates the influence of bad luck through huge armies and number of trials. Wesnoth does not have this luxury because individual units are so valuable and their health pools permit quite limited numbers of random trials before death.

And incidentally, there is a Risk clone that shows what Risk can be like with less randomness called Warlight and it's undeniably a better game for it.

>>1009

Warlight: 60% chance to kill enemy army - that is like same shit you face in Westnoth. What a terrible counter example.

Go play Catan with no dice.

>>1010

It isn't, because in Warlight you do one of those separate 60% trials for every single army on a territory. This results in engagements happening as you would expect from them more often than not. In Warlight the general rule of thumb is to only attack territories when you have roughly 250% the armies as the defending territory. If you follow this rule you'll almost never be upset by bad luck. Wesnoth on the other hand does far fewer trials before a single unit dies, and unlike Warlight every individual unit in Wesnoth is very important, so you can often get screwed hard by some bad luck.

They've been asking for more people to help with development recently:

>Wesnoth, as a project, is understaffed. At this time, there are fewer than half a dozen developers working on each new version of the game, and even fewer of them are able to work on the engine itself. We do not collectively have the time or skills to fix bugs as quickly as we should, or implement features as rapidly as we would like. The game itself suffers from an aging codebase and old software. As the gaming industry marches on and even the simplest games become more complex, Wesnoth has begun to feel outdated. Our internal organization is in need of improvement. The long length of the 1.11.x development cycle was caused less by us working on fixes or features than by an inability to successfully do these things in a timely fashion.

>To put it bluntly, this ship is sinking.

From what I've seen from the libre world there isn't much interest in keeping it going. It's been around for over a decade and the project is feature complete for most people. Updating the graphics with higher resolution sprite art that would look good on a 4k screen would a be massive effort considering the quality and quantity that's already there. Already on a 23" 1920x1080 screen the characters are smaller than a dime at native resolution. It still looks fantastic at 800x600 though.

>>1043

I don't understand why they can't just scale the sprites? It looks just fine at 1280x1024 for me.

>>1044

>>1043

Wesnoth has the best sprite artists in all of libre game dev, a side effect of this is that they also have the strictest artistic guidelines of any libre game. Just look at their character portraits, that is fully professional quality art there, and its a very hard style to emulate for most free artists. They require pixel perfect art, and they are fundamentally opposed to sprite scaling

Even at their small sprite sizes, their art requirements are a pretty huge art burden, if you made the sprites 4 times the height, that would make each sprite contain 16 times more pixels per animation frame, multiple directional attacks, multiple types of attacks, unit progressions and all units in every faction. A sprite upgrade is just not going to happen.

I have played the game extensively and spent many long nights following the internal development process. I love Wesnoth, but I can confidently say that the only situation in which I'd work on Wesnoth would be if they made a sequel in a 3d engine.

>>1101

The nearest-neighbor one is the exact same thing only bigger. The rest are only a departure because of the disgusting filter you used.

>>1006

Agreed. War, even on the individual scale, is decided more by your ability to provide for contingencies than your ability to plan and execute cunning tactics. Strict determinism is for puzzle games. Wesnoth's big problem in single-player isn't its randomness, but the difficulty (especially for novice mappers) of balancing multi-map campaigns against snowballing resource accumulation.

>>1074

Polygonal 3D would be a hardware-crushing abomination for a 2D turn-based game like this. Just draw all assets at extremely high resolution (or in some sort of natural media, and scan,) then allow people to download whatever sprite set they want. Either way, like the flawless pixel-art style has to be ditched to go higher.

If there's anything major the game really needs, it's a ground-up rewrite of the graphical subsystem to be more stable, efficient, and scalable. RAM requirements are absurdly excessive on cellphones, and wildly fluctuating CPU requirements can cause even PCs to chug or glitch at times if you're doing something else in the background. Think of something like StarCraft, which did hundreds of 8-bit sprites with multiple levels of alpha at 640x480, all buttery smooth in software on a Pentium with 16MB RAM.

>>1087

>>1044

You folks DO know the game has had a nearest neighbor zoom function (default keys "+" "-" & "0" in addition to a conspicuous magnifying glass icon next to the minimap) since the very first version, right? I guess they could add more GUI sizes, scalers, and some sort of automatic HiDPI detection to zoom by default.

>>856

I played the fuck out of this game. Its great.

>>1813

>War, even on the individual scale, is decided more by your ability to provide for contingencies than your ability to plan and execute cunning tactics. Strict determinism is for puzzle games. Wesnoth's big problem in single-player isn't its randomness, but the difficulty (especially for novice mappers) of balancing multi-map campaigns against snowballing resource accumulation.

I would argue that intelligence and understanding the tactics of your enemy and how to outsmart them are more important. The real problem with this argument though is that we're talking about a game here. War in real life may be decided by random factors but there's no reason to play it up so much in an actual game. It doesn't make it more fun; it makes it frustrating and makes it particularly unrewarding to put a lot of thought into planning ahead. Instead the planning becomes more a matter of playing in as safe and passive a manner as possible. Imagine if you were playing a game of chess and your queen could randomly fail in capturing a pawn and instead be defeated. It just isn't needed to make a competent game of strategy.