[OLD] Old OXCE discussion thread

[Dioxine]

I'd suggest leaving Warboy's Apoc-style Training Room functionality intact, and making an alternate functionality apart from it when/if such a thing is needed, in part because I think many people would be interested in Apoc-style training rooms without any extras.
Before saying more on that issue I need to test it in the first place - I have no slightest idea how big the skill increases are, especially since XCom:UFO uses a vastly different timescale compared to Apoc (I wouldn't mind a link to the relevant code!)

[Warboy1982]

https://www.ufopaedia.org/index.php?title=Training_(Apocalypse)#Combat_Training

https://github.com/Yankes/OpenXcom/blob/OpenXcomExtended/src/Savegame/Soldier.cpp#L643

the code is practically verbatim.

[Dioxine]

Thanks, it is really simple to understand. Any tweaking would probably entail:
1. possibility to customize the speed of training (circa +20 on first month all stats on a rookie might be just too much for longer campaign mods);
2. possibility to customize caps maybe (get statTrainingCaps, defaulted to statCaps if undefined) - if a modder wants the peak skills to be accessible only through combat experience. Still, the effectiveness of the training facility is really low once you're close to the maximum.
2. possibility to customize training facilities (any training facility has a list of stats that are trained there, defaulted to all if undefined)

[robin]

https://www.ufopaedia.org/index.php?title=Training_(Apocalypse)#Combat_Training

https://github.com/Yankes/OpenXcom/blob/OpenXcomExtended/src/Savegame/Soldier.cpp#L643

the code is practically verbatim.

(Maybe you're already taking this into account in your code) Apocalypse cityscape is "organized" in weeks rather than months. Everything is (or seems, to me) synched to that shorter time-frame, compared to UFO: one day in Apoc is (or seems) a more significant (and, potentially, events plentiful) time span than a day in UFO.
So having training gain rate set the same as Apocalypse's, might be too much for UFO.

It should be tested. I understand though that this is not the priority right now.

[SDEDEN]

So I noticed the Training Facilities mod was updated in the OXCE mods pack. Uh, what was changed? Was it just the file structure?


Anyway, I'm no coder and I just play and mod XCOM for fun, but I've been using the Training Facilities with Extended 2.0 and it is quite potent. After about two months soldiers who have stayed in training that whole time are pushing high 50's to 60's strength. Another two months and they've got 60 health, 70 time units, 70/80 accuracy, and 70 strength. Melee and throwing accuracy are getting into 60/70's too. Pretty much everything but bravery and reactions is uber.

Now I've never played apocalypse, but I think some kind of cap on what the training facility does would be appropriate.

Like either providing a fixed, flat increase to each stat or raising all stats to a hard limit. Like you know, in the first option soldiers get 15 time units, 20 energy, 10 health, 10 reactions, 10 accuracy, ect...

Or cap health at 40, strength at 35, reactions at 50 (so that those really low types can be average), and such.

[Yankes]

Only file structure. Overall I have plans to add more control over stats of training facilities.

[SDEDEN]

Only file structure. Overall I have plans to add more control over stats of training facilities.

Well I look forward to it! I used UFO Extender back in the day and was very happy with it. I started using it with Open XCOM mainly because of the AWACS mod and I loved how it combined with the Hyper-Wave Decoder not being 100% reliable.

Having to take a proactive approach to find and attack UFO's makes the game play a lot better than just sitting back passively and waiting for them to fly over you.

[Yankes]

New bug fix version with update to recent nightly.
Its fix reload bug and some UI issues with crafts.

[Solarius Scorch]

A thousand thanks, Yankes.

[Yankes]

New version 2.1:
Craft can now have up to 4 different wepon type per slot.
Oprion for overriding shape of alien base.
Option for setting chance of special effect of item.
Moving can trigger all close proxy not only one.
Corpse explosion now can have bonus based on unit stats.
Grenades now can explode at once without destroying each others.

[Solarius Scorch]

Splendid Yankes, great news again.

Now, I have another thing in mind that would be very handy for the Piratez mod (not so much for the vanilla X-Com, but maybe a bit). You see, it's a bit awkward to see 10 "enemies" being civilian passengers running around the ship, so you need to kill/stun them all to win.

The solution I propose is something I call "surrender mechanics". It is simple: whenever all remaining enemies are panicked, you win the battle automatically (just like when they're all mind-controlled).

---===---

And one more, completely unrelated, thing. In the Xcomutil, there was a "fix" for how the clips are recovered: for every clip there was a percentage chance of how likely it is to be recovered based on how much of it was depleted - so a full clip was always recovered, an empty clip was always recovered, a half-empty clip had 50% chance of being recovered, and a 10% full clip had 10% chance for being recovered, and so on. I really liked that system, since in the current one you hesitate whether to fire your special weapon (which you only have one of) because you know that even if you only fire one bullet, the entire clip will be wasted. Which is not very much fun. So I thought if we could, eh, have it the Xcomutil way.

[Arthanor]

haha! That's an interesting idea. I just suggested reducing the size of some clips to alleviate the problem of "fire once, lose whole clip" in Piratez. That would be an alternative that's interesting. With OpenXCom counting bullets, it is much less of an issue for your main guns, but it is still really annoying for those guns you only carry 1-2 of and only fire 1-2 times.

Maybe it could be: Count bullets, get remainder of Bullets/Clipsize, give an additional clip back remainder/Clipsize*100% of the time? That's an easy step to add at the end of the current bullet counting instead of entirely replacing it. (One day I'll learn to code in C++! If only for these super easy things!)

That way if you fire 25% of the shots from two clips, you get one back for sure, with a 50% chance of the other. It is slightly different than 75% of keeping each (which gives you 56% of keeping both, 38% of keeping one, 6% of losing both). The results average up the same (average amount lost being either 0.5, or 0.38+2*0.6=0.5). It just makes the odds a bit more reliable (you know you keep one, you are more likely to lose one, and you won't lose two). Of course, if you only have one clip then it's exactly the same

[Yankes]

I think that Arthanor way is better. Only last half empty clip will be recover with random chance. Only thing I dislike in that that you could have miracle and use two bullet clip for 4 shoots in 4 battles. Or have 100 bullet clip and lost it after only one shoot. This more psychical problems than gameplay problem because in long run clip usage will be average.
Simple current system with fixed clip size will always create glitch with weapon that have limited ammo.
How about system where we count bullets not clips?

[Solarius Scorch]

I think that Arthanor way is better. Only last half empty clip will be recover with random chance. Only thing I dislike in that that you could have miracle and use two bullet clip for 4 shoots in 4 battles. Or have 100 bullet clip and lost it after only one shoot. This more psychical problems than gameplay problem because in long run clip usage will be average.

Yeah, neither system is perfect. With this, we could at least escape the "fire once, lose everything" problem, but I don't know if it's best.

How about system where we count bullets not clips?

What happens with the "loose" bullets? You get one semi-full clip?

[Arthanor]

I thought I had read that OpenXCom already counts bullets. I think it does something like this after the battle:

Code: [Select]

1- Count all bullets left in all clips of a given type, call that n1
2- Divide n1 by the capacity of the clip, call that n2
3- Round n2 down to the lower integer, call that n3
4- Give back n3 clips to the geoscape.
An example of this would be having 4 heavy plasma clips and you fired 28 shots. At the end of the battle you have n1 = 112 shots left (4*35-28). Divide 112 by 35, n2 = 3.2, so you are given back n3 = 3 heavy plasma clips back (ignoring whatever was recovered and any extra you brought). That's 3*35 = 105 shots. The remaining 7 shots are lost since we only track full clips. You didn't lost all your clips, but you lost something.

Unfortunately, this method does not help if you used only one clip of a given kind and the enemy has none (a situation that happens relatively often with special weapons in mods like Piratez or the FMP). In that case, n2 is always smaller than one, so n3 is always 0, so firing one single shot always takes away your whole magazine.

Over time, you lose out on a ridiculous number of shots, way, way more than you fired. For that reason, I tend to limit my use of special weapons (ex.: my 2 flamethrower soldiers use exactly 1 tank's worth of shots between them if possible, certainly never more. So I lose 1 tank/battle. If they fired a tiny bit more, I'd lose 2. If they fire less than a full tank's worth, some shots are wasted and I lost a tank any ways).

My suggestion is to add a "step 3.5", which would be:

Code: [Select]

3.5- Take n4 = n2-n3. Pick a number between 0 and 1. If that number is smaller than n4, add one to n3.
In the above example, n4 = n2-n3 = 0.2. So you would have a 20% chance to get an extra clip back. At first, this sounds like 20% chance of cheating (you fired all those shots for free!), but it is only "compensation" for all those extra shots you lose when you roll above n4.

This solution will never have a miracle (it is always at most 1 clip away from the exact answer) and works well if you had only one clip and only fired a few shots. Most of the time you fired for free, but once in a while you lose the whole clip, meaning that on average you lose exactly the number of shots you should.