[Request]Reduce RNG influence

[Zarel]

In case this isn't quite clear, the patch from akuu is also a 'parapet' system as in it avoids streaks of the RNG, but still different from the idea I proposed. akku is making a global modification, while I'd like to modify locally and independently for each unit on the battlefield. I prefer to insist because I don't find the difference to be obvious, while it is actually matterful.

True, your idea is per-unit karma, and akuu's patch is global karma, but the only real difference is that yours is more exploitable - both otherwise suffer from the same basic problem that it makes luck affect your luck.

That complexity is subjectively evaluated, in my eyes you are quite exagerating it, not that I would pretend to be objective either. The difference with the current system is only that casually an outcome is determined regardless of RNG. Appart from this special dice roll, which should not occur more than once every 2 or 3 turns in a duel from what I can tell of my experience of the game (except in the case of berzerker special fights), the anticipation of the outcomes remains the same.

...I'm sorry, but I don't understand what you're saying.

You can't easily decide to wait for getting karma, it is not something you have control over, RNG decides that for you.

Exactly. This goes back to the earlier problem: Now, not only do you have to worry about luck, you have to worry about luck affecting luck. KISS. It's important.

Let's go over why KISS is a good thing: Too much complexity means that the winner becomes whoever can remember and make use of all the different elements of the game, instead of whoever has the best strategy.

Moreover, a unit having high karma means that it failed a lot previously, it is not an unfair abuse to tab on 1 guaranteed hit for all the times you missed before. If I had to choose, as a player, I would want poor karma and no guaranteed hit, meaning that my unit got pretty lucky before, than the opposite.

No, because those failures occurred when it didn't matter. Remember, the core problem with luck isn't that people get more good luck than bad luck - the way luck works, you get mostly equal amounts of both over the course of a game. The core problem is that people get bad luck when it matters and good luck when it doesn't. A guaranteed hit is worth so much more than slightly worse dice rolls over the course of a game, because you don't get to choose your bad luck no matter what. You do get to choose where you want to spend your guaranteed hit.

I don't know if I interprete the localized luck surges correctly. From what I saw on the forum, the complaints are generally that people see 'crazy' things such as having one's unit take many hits despite on good defense terrain, or miss many times in a row with a good cth. I don't think I have ever seen someone ranting on the forum that they miss with the high damage dealer, and succeeds the unimportant hits, although this is a recuring argument when people analyse what should be perfectly even luck. So I'm tabbing on the fact localized luck surges means a local streak of good/bad RNG rolls while it is statistically even if you take into account the next and previous RNG rolls. And if I'm interpreting correctly, then that 'unit parapet' system exactly aims at solving this problem (I'm changing a bit the name, because I recall 'karma' to be used for other systems already, and fear a confusion)

Yes, exactly, and with your system, that one unit will continue taking that many hits or that many misses, assuming neutral karma. In fact, assuming karma, that would happen even more often, because you could have good or bad karma, respectively.

That's why my proposals are only changing localized values. A guaranteed minimum goes a long way in ensuring a string of bad luck isn't catastrophic.

That is quite true, that if you attack a unit and made 0/2 with your wose but 2/2 with the shaman it won't change. This is not the sort of things that frustrates me, or what I try to correct, because I accept that locally the wose may totally miss and either plan a failure case, or accept the consequences of the failure. What annoys me, and what I try to correct, is to do 0/2 with the wose and 0/2 with the shaman and miss again and again with the next units, which the parapet system can avoid.

Erm, this is like the fifth time you've used the word "parapet". I'm going to guess your native language isn't English. I understand the metaphor that a parapet can prevent someone from falling off a roof, and the karma system prevents someone from having too much bad luck in a row, but in English, extended metaphors are so rare that its Wikipedia article is only a stub. Extending a metaphor for more than a sentence is rare; for more than a few paragraphs is virtually unheard of. More seriously, my point here is that an American English speaker is going to have no idea what a parapet system is; could you just call it a karma system?

A unit karma system will not avoid the situation of doing 0/2 with the wose and 0/2 with the shaman and 0 with many subsequent units, as long as those units all have even or negative karma. Even if some have positive karma, the ones with negative karma will do more to even it out. That is one of the things a global karma system would do more to fix, but I'm not sure how this is an argument for a per-unit karma system.

My point of view is rather that you don't want to have high karma, it is only a small compensation for the bad luck you had. You can take advantage of it for guaranting an important hit, if you have the opportunity to at the right time, but I consider this advantage to be fair considering the precedent counter-performances of the unit. Inversely, having low karma gives the unit a temporary disadvantage, and maybe not at the best time for you, but it is a punishment for it to have been over-lucky.

It's not. Taking bad luck as it comes is something you will do regardless of whether or not you have a karma system. Being able to spend good luck?

Let me tell you the parable of the horse race.

So there's this guy, and he has three horses, they can run at 20, 30, and 40 mph, respectively. His opponent also has three horses, which run at 25, 35, and 45 mph, respectively. Now, half of you are thinking, "Oh, man, this guy's gonna lose the horse race." The other half realizes what he does: He races the 30 against the 25, the 40 against the 35, and the 20 against the 45, and wins 2 out of 3 races.

Now, there are a number of conclusions that can be drawn here, but the point is: The ability to choose when you get good luck and when you get bad luck is important.

Some of the points we are discussing on are quite subjective Zarel, so I believe it'll be as hard for me to convince as for you to convince me. Those points are : (feel free to correct me)
- such 'unit parapet' solution is easy/complex
- the main problem is 'the long series of luck streak'/'the fact you get hits when you don't need, and miss when it's important'. For this point, from a perspective where we analyse how we can make the luck system perfectly fair, I agree with you. My perspective is to try to correct my own frustrations, so that's very subjective indeed, and in that regard I want to correct only the luck streaks.

Some parts of it is subjective, but the core problems are factual and can be calculated mathematically:

1. A unit karma system will do nothing to decrease the probability of getting bad luck within a turn.

2. Any karma system will increase the average number of times a unit gets localized extremely bad luck.

3. Let D be a unit's terrain defense, let K() be the karma function, and let H[n] be a unit's previous hits.
In the old system, the unit's accuracy (barring marksman/magical) would be:
1-D
In the new system, the unit's accuracy would be:
1-D+K(H[-1])+K(H[-2])+K(H[-3])+K(H[-4])+...
Now, whether or not KISS is a good thing is subjective, but whether or not this proposal is KISS is quite objective.

4. Okay, let's talk specific situations. You have an Elvish Fighter. Opponent has a Fire Drake with 1 HP. Your Elvish Fighter attacks the drake and misses four times. Under a karma system, that situation will happen more often, not less.

5. No other game has a karma system, and many game developers would laugh in your face if you proposed one.

[Yogibear]

Let's go over why KISS is a good thing: Too much complexity means that the winner becomes whoever can remember and make use of all the different elements of the game, instead of whoever has the best strategy.

I don't understand this. Shoudn't the best strategy be the one that takes into account all the different elements?

Moreover, a unit having high karma means that it failed a lot previously, it is not an unfair abuse to tab on 1 guaranteed hit for all the times you missed before. If I had to choose, as a player, I would want poor karma and no guaranteed hit, meaning that my unit got pretty lucky before, than the opposite.

No, because those failures occurred when it didn't matter. Remember, the core problem with luck isn't that people get more good luck than bad luck - the way luck works, you get mostly equal amounts of both over the course of a game. The core problem is that people get bad luck when it matters and good luck when it doesn't. A guaranteed hit is worth so much more than slightly worse dice rolls over the course of a game, because you don't get to choose your bad luck no matter what. You do get to choose where you want to spend your guaranteed hit.

I think this is a very contestable statement. At least from my own player experience i'd say that this is rather an exception than the main reason of losing a game.

1. A unit karma system will do nothing to decrease the probability of getting bad luck within a turn.

That is something i support. Sometimes, the first serious attack decides if you get screwed or not. Unless the karma range is set very narrow, it won't solve that problem.

5. No other game has a karma system, and many game developers would laugh in your face if you proposed one.

That's redundant at best (in the sense that they won't do it because of the reasons you stated). Even if you admit that those devs will have their reasons it's not clear what those reasons are and if there are ways to overcome them.


One of the main arguments against a karma system is that it will become exploitable. In my opinion, this is based on the fact that you have information about the karma value. What if you don't (like you don't know the seed value for the rng right now)? I wouldn't possibly want to keep track of every swing a unit does just to know when it will get a guaranteed hit.

[Velensk]

I don't happen to like the idea of a karma system however I agree that not having it because other developers would laugh at it is a poor idea. However I will say Yogibear that just because you arn't willing to keep track of every hit or miss to find the actual odds of a combat doesn't mean others are not. If you are willing to accept that as a part of evaluating odds then this may not be a bad thing, but a player who has dilgently counted hits and misses should be able to much more accurately predict combat than a player who runs from memory to the point where being good at doing so and anyalysing the results would be a more relevant skill than knowing basic probability.

[Zarel]

I don't understand this. Shoudn't the best strategy be the one that takes into account all the different elements?

Hmm, perhaps I misspoke. Regardless, KISS is usually important.

I think this is a very contestable statement. At least from my own player experience i'd say that this is rather an exception than the main reason of losing a game.

Yes, because the main reason of losing a game is bad strategy, because Wesnoth is a balanced strategy game. For the most part, there's nothing wrong with luck as is in Wesnoth. You can't point to a Wesnoth game with karma and say "Look! This isn't that bad of a game" if Wesnoth without karma would be just as good, if not better.

The core point is that karma won't improve anything.

Go play a game, and record each hit and each miss, and the probability for each. You'll have nearly exactly as much good luck as bad luck. Now, what is a karma system supposed to enforce? Exactly that. So on the surface, how does it even change anything? It really doesn't.

That's redundant at best (in the sense that they won't do it because of the reasons you stated). Even if you admit that those devs will have their reasons it's not clear what those reasons are and if there are ways to overcome them.

As a game developer, I have a rule: If you're doing something that no one else in your genre is doing, you should have a clear idea of 1. why they aren't doing it, and 2. why those reasons don't apply to you.

To be better than others, first, you have to not be worse than them. That is the reasoning for mentioning other developers. Only when you've mastered that, can you move on to more important things like learning from their mistakes.

One of the main arguments against a karma system is that it will become exploitable. In my opinion, this is based on the fact that you have information about the karma value. What if you don't (like you don't know the seed value for the rng right now)? I wouldn't possibly want to keep track of every swing a unit does just to know when it will get a guaranteed hit.

So, to fix the exploitation, your solution is simply not to tell the player their hit probabilities? Are we still trying to make a strategy game anymore? Knowing your hit probabilities is essential to formulating strategies. And having hit probabilities that are easy to calculate is also essential to formulating strategies (which is yet another reason why I object to any karma system at all).

I can think of another game in which keeping track of tiny things like that can give you an advantage. It's called poker, and the advantage is called card counting. Incidentally, poker can handle it since it's at least partially a game of luck. But to introduce such a mechanic in the interests of reducing the influence of luck is ridiculous.

In my opinion, the exploitability isn't even the biggest problem with karma. The biggest problems are, in order, 1. it doesn't fix anything, 2. it makes probabilities depend on luck (in addition to making them needlessly complex).

I mean, I've said it too much, but it really is that important: isn't it obvious that making luck depend on luck is a bad thing?

[Max]

chances are good that such a karma system has been used before - there are thousands of single player games where nobody ever cared to question the mechanics behind the scenes.

One of the main arguments against a karma system is that it will become exploitable. In my opinion, this is based on the fact that you have information about the karma value. What if you don't (like you don't know the seed value for the rng right now)? I wouldn't possibly want to keep track of every swing a unit does just to know when it will get a guaranteed hit.

So, to fix the exploitation, your solution is simply not to tell the player their hit probabilities? Are we still trying to make a strategy game anymore? Knowing your hit probabilities is essential to formulating strategies. And having hit probabilities that are easy to calculate is also essential to formulating strategies (which is yet another reason why I object to any karma system at all).

hiding the current karma status of a unit would indeed be a problem. karma would become an essential part of micro management. knowing that you could do a better job by remembering hits/misses history - that would spoil all the fun - at least for me (i hate some card games for that very reason).

[hiro hito]

I think that things should be more simple:

I think that "karma" is a possible way if we consider it like a "luck tank". And this luck tank is empty every turns of fight.

An example:

- I do my first attack: I miss all my hits.
My luck tank wins some ChanceToHit (let's say the invert of my calculated stats/2). This ChanceToHit will be active at my second attack. When I do my second attack, my ChanceToHit will be add to my NormalChanceToHit. After my second attack, my luck tank is empty. Etc,etc...

That way, managing luck/risk is always a part of the game, but the unlucky player has a chance to counter attack with a fair amount of luck. Thus we can see if indeed he has managed his own risk.

[Yoyobuae]

So far all of the "karma" systems have been look-back.

Would there be a way to make a look-forward one? You'll say: "You can't predict the future", but there is a way.

The trick is buffering. Keep a number of RNG results in a memory buffer. These can be analyzed and tweaked to try to filter out cases of really bad/good luck. Whenever one is used in battle, it gets discarted and RNG generates a new value, the analysis is repeated.

Note that the players don't really have any information about what's inside the buffer and thus cannot make predictions. Or, can they?

[pauxlo]

I'm not sure whether the proposed unit karma system for each unit would keep track of how much the concerning unit (1) hits more or less then expected, or (2) gets hit more or less then expected or (3) both in some combined manner or (4) both individually.

In case (1), you could exploit when you have a "unlucky" (high karma) unit and go in to get a sure hit when needed (Leader assassination or so). Or can track the karma of the enemies units, and attack the one with low karma with your 1-HP unit near to leveling.

In case (2) however, you either have a lucky unit (a lot attacked, still surviving) -> it has bad karma, and so will be killed soon. Retreating does not really help, since then you can't ever again using the unit (it karma stays low until it gets hit).
Or you have an high karma unit - but this unit is usually already dead, so it can't use its karma.

(4) combines both problems ... if you had some "high-attack-karma low-defense-karma" units, you could do kamikaze attacks on only one free spot of an important enemy unit (with normal karma).

For (3) it would depend on the kind of combination. (I think chance to hit could be influenced by the difference of attacker karma and defender karma).

But, try it out, for this we are in the experimental corner

[gabba]

My speculative mind can hardly resist such a speculative thread . So here's my take on this.

I don't like the idea of guaranteed dice rolls from time to time, I think it'd just be confusing, and has the potential to be as annoying as the RNG for the player who's at the receiving end of the guaranteed hit, or has to try and take the other player's karma into account when calculating odds. Nightmare. A die roll is a die roll dammit!

I'm a big wargame player, and I never saw as much complaining about luck in games such as ASL/Advanced Squad Leader (boardgame) or Panzer General and Steel Panthers (computer). I think the root of Wesnoth players' frustration with the RNG lies with:

1. the player's inability to tweak the situation in their favor more than 80%
2. the possibility that even a massive attack with the best they've got (i.e. 3 powerful units attacking for 2 turns against a single dwarf in a mountain chokepoint) can very well have no effect whatsoever

In ASL instead of hit-or-miss, you have graduated effects based on how well you rolled. This is the fire table that's used (download the pdf, it's in english). You cross-reference the combined firepower of all attacking units (at the top) with the chance-to-hit modifier (at the left). Results are not just hit or miss: from best to worse, they rank from killing off the whole squad (KIA), damaging it, submitting it to a morale check, or just a pin check (i.e. do they keep their heads down under fire), or finally nothing at all.
Weak attacks still have a pretty good chance of having no effect, but with large amounts of firepower you're almost guaranteed the weakest effects. There are also situations where units are automatically eliminated, such as a demoralized unit caught fleeing in open ground with enemies in sight in every direction.
This makes situations where you use every ressource you got and fail to achieve anything (and then lose the game because it was a critical moment) much less likely, unlike Wesnoth where they come up a bit too often.
But to come back to the guaranteed hit/karma topic, I think the (much smaller) risk of catastrophic failure still has to be there. It's exciting, it gives a realistic dimension (you can't simulate everything) that makes the game more lively, and it gives you stories to tell.

In Wesnoth terms, here are some solutions I've thought of to allow the player to tweak the situation in his favor:

1. have a fatigue system that imposes gradual penalties to units as they get attacked a lot (for instance they become easier to hit and eventually slowed). Fatigue could come back at a constant rate, or recover like health.
2. a morale system could also be interesting; it would be affected by distance from leader, number of nearby friendly/enemy units, success in battle (both individual and overall), combat experience and what not.
3. have a ladder of effects depending on how well you roll (ideas in no particular order): damage, tire, slow, pin (i.e. unit loses some movement/attacks), ...
4. drop the "hit chance is only determined by terrain (or magic, or marksman, or whatever attribute is gonna be added next )" principle, and allow the player to stack the odds by using various factors in his favor. For instance the following could each increase the hit chance by 2-5%, to a max of 99%:
   • attack vs tired unit, vs wounded unit, vs slowed unit, vs surrounded unit
   • attacking from high to low ground or from one of your favored terrains
   • some units could have a "demoralize aura" that makes nearby enemy units easier to hit
5. whether the unit moved last turn could also affect things: for instance (quick idea) a unit considered "in motion" would be harder to hit with ranged attacks, but would have worse defense in melee, since it's not fighting on "prepared ground".
6. true ranged attacks (so you can stack even more effects on the same hex) would probably help with this too.
7. these ideas are mostly attack-oriented, but some defense-increasing modifiers would be useful too (in moderate quantities)

All those derail Wesnoth from its simplicity principles, but they would definitely help the player feel more in control, i.e. the better your strategy, the less the RNG will matter, even though a big setback would still happen from time to time.

BTW don't expect me to develop this much more this summer, but who knows, maybe later on I'll be tempted to implement some of these ideas.

[krotop]

Remember, the core problem with luck isn't that people get more good luck than bad luck

By 'more good luck than bad luck' I'm assuming you say 'better RNG numbers than their opponent'. I don't know for sure if you're attributing me that point of view, but in case there's a confusion I never said a player had overall higher rolls, or even thought so.

The core problem is that people get bad luck when it matters and good luck when it doesn't

I understood that point and explained already why I disagreed with it. I don't think anything productive will come from repeating the same arguments to support our thesis, unless a clarification is needed.

A unit karma system will not avoid the situation of doing 0/2 with the wose and 0/2 with the shaman and 0 with many subsequent units, as long as those units all have even or negative karma.

It can be right depending on how large the karma limit is and how low the karmas of the attacking units and defending units are. This is not a complete removal, but it does reduce the probability of it to happen, and depending on the karma limit, it can reduce it a lot. I'll use your example of the elvish fighter battling a 1HP burner with a 60% cth.

With a karma range set from -30 to 30, which corresponds to situations with less than 5% chance to happen, the kill is guaranteed if the attack-karma at the beginning of the fight is at -6 or above, the survival chance is the same as standard wesnoth if the karma is between -7 and -27 (2.56%), and higher if the attack is at -28 or lower (6.40%).
Admittedly these figures take the attack-karma only into account, and I purposedly did not consider the karma of the defender to avoid citating too many cases. But those are enough to show that the system does not increase the chances to "have extremely bad luck", which I understand as "doing 0/4 successful hit on the burner", and on the contrary diminishes these chances, without erasing them.

I believe this to also answer the points 2 & 4 of your sum up.

1. A unit karma system will do nothing to decrease the probability of getting bad luck within a turn.

That is something i support. Sometimes, the first serious attack decides if you get screwed or not. Unless the karma range is set very narrow, it won't solve that problem.

If you don't want the karma system to be intrusive in the game, thus setting it very large, then yes it won't change anything to the current gameplay. If someone is more scared about bad luck streaks then you are giving the solution Yogibear: that person can set a smaller karma range, that comes of course with accepting a more frequent intervention of that less-luck system to correct instances of bad rolls. That balance between intrusion and bad luck correction either is to be set from playesting or let to the players' convenience.

3. Let D be a unit's terrain defense...
[...]
Now, whether or not KISS is a good thing is subjective, but whether or not this proposal is KISS is quite objective.

It is not exactly how a formula of that system would look like, but yes, it is probably not KISS, because it removes the simplicity of independency between events (what you call 'luck affecting luck').

One of the main arguments against a karma system is that it will become exploitable. In my opinion, this is based on the fact that you have information about the karma value. What if you don't (like you don't know the seed value for the rng right now)? I wouldn't possibly want to keep track of every swing a unit does just to know when it will get a guaranteed hit.

Personally I would be very fine with hiding what are the operations to correct the chance to hit. My ideal would be to just make my risk assessment as usual except with a guarantee that the RNG won't go too wild for or against me. If I'm proposing to show the info, it's because I'm assuming that other players will try to understand those operations to take advantage of it, and then use it to have an edge over the people who can't exploit those operations. I believe that during a game you tend to memorise particularly lucky or unlucky events and even without precise figures you intuitively know that if your mage made 1/3 the previous turn it will most likely have a better chance for his next attack, and possibly with experience of that mode your idea on how better this chance is will become more precise. That is why I expected to show some rough gauge of karma to give everyone a quick idea on how to exploit it, for people keen into calculation, experienced players, or newcomers to have the same chances to use the system.

What I'm tabbing on is that the exploitation of a high or low karma value for you or your opponent will remain fair considering the way you earned that karma.
For example, the exploitations mentioned by pauxlo (it's (4) by the way in my head) don't shoke me. You get 1 single more hit and it's enough to kill the enemy leader, and it did cost you to have the unit to miss a lot previously, nothing unfair for me. For the surviving unit, it was supposed to be dead according to statistics and instead you 'only' get a living unit that will likely take 1 single extra hit next time, nothing unfair for me. It can indeed happen that you have 'killer opportunities', but how frequent are they? It appears absolutely exceptional to find a unit with both so low karma and so low HP that it can be killed in one hit, and a unit with so low HP and ready to level at disposal to finish it off with a guaranteed hit on his *first* strike.

[eyerouge]

Design Goals
To make this discussion/thread into something meaningful it should probably have clear design goals stated. As it is now it's the wild west and any ideas, no matter how far inbetween and no matter what kind of changes they require from the game, are being discussed in here. While the thought was a nice initiative from Yogi, I believe we'd benefit by more structure.

For example, if people wish to post 20 times about pros and cons with i.e. a Karma system, then a thread called "Karma RNG System" should be started up and linked to from here. All would benefit form such a structure, because threads like these tend to become very clogged up by parallel and multiple discussions.

An even more pressing matter is that the call for a a new random RNG-system should have some design goals associated with it. With other words, criteria that has to be met by people that suggest whatever system they drool over.

Example example design goals for a less random RNG:

• Add no extra in-game variables that are shown to the player or which the player can keep track of in any way.
• Add no extra in-game info of any kind. No new placeholders, no new nothing that is shown to the player as info in any way.
• Keep a 100% unit/era/rules compatibility with normal RNG Wesnoth (as it's defined in fork version of BfW)
• Keep it intuitive & KISS.

Again, this is an example of something which could be the beginning of a sane design goal list that could work as a good starting point for people that want to contribute to the solving of a the specific problem at hand (which reminds me - define that as well - as it is now all have their own interpretations, which of course only makes the confusion in here as to what solution is the correct one the larger...) I am however not proposing that my list should be used. I even imagine that different threads could use different lists as they probably relate/see/mend different stuff related to the RNG.

Perfect solution that already exists...?
Is saurons mod, as described in the Wiki page which probably few read anyway .

I (currently) don't see any problem whatsoever with his solution, nor why anyone would seek a more complicated way of remedying what he seemingly already has fixed with his solution (I dunno is sauron is a guy, but whatever, not the issue).

I'd love seeing valid logical arguments against that way of solving the problem that demonstrates that the solution has more cons than pros going for it. To me it seems as a hard task, and I honestly believe the system should be given some serious consideration by whoever is actually implementing it in the end.

If you want to
a) keep RNG
b) make it less influential
c) at the same time offer the deterministic players a hard and huge bone and d) keep compatibility 100% for a zillion good reaons, then that would be the obvious choice.

Personally I think it's meaningless discussing more complicated ways of handling the task at hand since the solution seems to already exist and the offers this far all have had their own different but serious drawbacks. Something saurons lack. If one on the other hand wants to do more with the RNG system than what was originally asked in the thread start and if one wants to add additional features/mechanics into the game then there could of course be a million of other ways to deal with the issue and at the same time introduce x into the system.

[hiro hito]

Perfect solution that already exists...?
Is saurons mod, as described in the Wiki page which probably few read anyway .

I (currently) don't see any problem whatsoever with his solution, nor why anyone would seek a more complicated way of remedying what he seemingly already has fixed with his solution (I dunno is sauron is a guy, but whatever, not the issue).

I'd love seeing valid logical arguments against that way of solving the problem that demonstrates that the solution has more cons than pros going for it. To me it seems as a hard task, and I honestly believe the system should be given some serious consideration by whoever is actually implementing it in the end.

I would like to test his mod but I am playing with Mac OS X version.
maybe I am wrong but I think his mod is only available for Windows...

It would be great if this mod will be like an add-on... Every body could play it with no technical problem...

[eyerouge]

I would like to test his mod but I am playing with Mac OS X version.
maybe I am wrong but I think his mod is only available for Windows..

It would probably run on any plattform that supports BfW since it's code is likely not bound to a plattform specific libraries etc. (That said, it still doesn't mean a pre-compiled version for all plattforms exist. I don't know the status of it for OSX, but you would probably compile at easily as you would regular mainline BfW...)


.

It would be great if this mod will be like an add-on... Every body could play it with no technical problem...

I think the thread is about features built into the experimental BfW (eWes?). As such this isn't supposed to be an add-on. It's supposed to be in-game and native, if implemented.

Edit: Mod thread >> http://www.wesnoth.org/forum/viewtopic.php?t=26803

[Yogibear]

An even more pressing matter is that the call for a a new random RNG-system should have some design goals associated with it.

Thank you for bringing this up, i'd like to follow up on that.

Generally, i want this fork to be as compatible as possible with mainline wesnoth. Ideally, you would be able to play on the regular mainline mp server with Experimental Wesnoth. Therefore, if possible, i would like to have new features able to being switched on and off and thus guaranteeing compatibility if wanted.

I know that this is a goal i might not be able to keep forever. Some things are really tricky to implement without breaking compatibility and for a number of reasons i wouldn't want to create a parallel universe (codewise).

While the thought was a nice initiative from Yogi, I believe we'd benefit by more structure.

I love structure . I think splitting this into some side discussions is a good idea. On the other hand, i wouldn't want to suppress any thoughts, just because we have a good solution already. If you have new thoughts, just keep them flowing, they might lead to some interesting aspects that haven't been seen before.

All the design goals that eyerouge listed are good ideas in the way that they try to minimize work on all sides. But again, before discarding solutions, i'd like to have them develop without having all the show-stoppers in mind already. When the discussion has come to an end and we still feel this is a good idea, then we can try and adapt it in a way that it becomes feasible.

Regarding the RNG-topic, i have started to refactor the RNG-code completely, as it is a little messy at the moment. The goal in short is this:
- Provide a (C++) framework that is easily expandable with new algorithms
- Giving the player the possibility to choose which RNG-algorithm he wants
- Keeping compatibility with mainline by implementing traditional RNG as one of the algorithm options

I got the framework in place almost, i hope i will be able to finish it next week (together with a guideline documentation on how to add your own algorithm).

Sauron's mod is ready to be used indeed and i plan to incorporate it as the code is already there. I had a short contact with Sauron and he is happy to see his mod come to life in Wesnoth Experimental. If i am not mistaken, it still has a problem: Attack prediction doesn't work right with it yet. Not so much a problem for the player maybe, but could be a big one for the AI, as it relies on this functionality.
Other algorithm implementations might follow, depending on what you give to me .



One design question to you: Where to choose the RNG-algorithm?

I can see two promising places: mp lobby game creation dialog and preferences dialog. The former would target multiplayer human vs human games specifically, the latter would also give the possibility to have the RNG-algorithm apply to campaigns. Any ideas on that?

[eyerouge]

the latter would also give the possibility to have the RNG-algorithm apply to campaigns. Any ideas on that?

I've never played a full campaign to date, so I'm probably not the one to speak. My guess is that some campaigns rely on and were designed around BfW RNG and that eWes RNG-modes could very well have funny or not so funny effects on some of the campaigns and how they play out. This would all depend on what RNG-modes were implemented ofc.

Because of that, and in order to keep compatibility with add-ons, campaigns etc from/for BfW, the default RNG for campaigns should be default BfW RNG. There could of course be an option in the preferences to change that, but it would also be a good idea to have a pop-up or similar notifying the player what he does and how the campaigns can become affected by changing that mode.

Now somebody will suggest that some new WML or info tag is added to campaigns so each can have a correct RNG-mode associated with it. As good as that may sound, it isn't a good idea if and only if it breaks compatibility. Else it would be a pretty tight solution and would also offer each campaign maker new power over his/her creations.