Castle Paradox

Newbie_Power · Joined: 04 Sep 2006 Posts: 1762

You don't need to do checks during or before a battle. You can do all of the checks to see if a Blue Mage is alive in an after-battle script, before giving the Blue Mage the spell. Even if the tag is set to be on, you can check to see if the Blue Mage is alive or in the party before teaching the spell, and if he is dead or not in the party, you can turn the tag back off and not teach him the spell.

Which leads me to a different method I came up with...

1) Have four different attacks, corresponding to their own tag (I would reserve tags with some range of number, like 800-999 so that I can keep organized). Use the "On Hit, set tag# = on" in the attack editor. Let's use tags 800-803 as an example.

2) After the battle, if a tag is on, like 803 have it check the corresponding hero slot you chose so see if a Blue Mage is both alive and in that slot (in this case, slot 4). If so, give the Blue Mage that spell.

3) If the Blue Mage isn't in that slot, turn the tag off and continue the loop.

I would use four different for loops that all count by four, that way it will be easier to script this beast.

msw188 · Joined: 02 Jul 2003 Posts: 1041

To Newbie Power:

I don't think you are quite grasping the problem that my method was devised to solve. In my system, it is not important that the learner was targetted by the attack he is trying to learn; my learner would learn by observing, NOT by being hit. Otherwise, there is no way to learn any kind of healing or support kinds of skills. Your system would not have the learner learn the skill unless they were actually hit by the skill, which was not my goal.

Given that, it might seem that no slot-usage is necessary and one could simply check after battle whether or not the learner is alive before awarding the skill. My problem was the following scenario (and I'm sure I posted this somewhere up there...): suppose that the learner is dead when a skill is used. The skill's tag is turned on. Then, before the battle ends, the player manages to revive the learner. Then after the battle, the learner checks out as alive and the tag is on. But the learner should not learn the skill if he was dead when it was used. This problem is solved by my slot & tag system, I believe.

Your system could also get into trouble, I would think, with skills that target the whole party. What tag do these attacks turn on? A fifth, so that you are checking in loops of five? Furthermore, tags are something we have a rather limited number of, as compared to enemies and attacks. I believe that my system is cleaner and better, AS LONG AS you want the learner to learn the skill by observing it, regardless of who it targets.

Also, your claim

Newbie_Power · Joined: 04 Sep 2006 Posts: 1762

Moogle1 · Posted: Sun Apr 15, 2007 8:53 pm Post subject:

I think you're thinking of the Morpher class? Either way, Gau is a singular character in the Final Fantasy world; the closest thing to it is probably the lame Tamer class from FF5 or the stupidly powerful Morpher class from FFTA.

One problem with the solution you're proposing is that the enemy will attack the Blue Mage disproportionately often if you have one copy of the normal attack and another of the BM attack. A workaround (one that is much simpler, too) is to just create four copies of each attack, one for each slot, then give the monster all four. You can check after battle if the hero's slot's tag is set. This severely limits the monster's other attack options, unfortunately.

Also unfortunate is that the solution won't work at all for spread attacks. Your best bet may just be to allow the BM to learn the spell if he is alive at the end of battle. You can make up some story fudge about how the character is able to osmose the spell while unconscious as long as he revives soon enough afterwards, I don't know.
_________________

Slime Salad :: Independent Game Design

Newbie_Power · Joined: 04 Sep 2006 Posts: 1762

msw188 · Joined: 02 Jul 2003 Posts: 1041

Okay, either I'm misunderstanding you, Moogle, or you're misunderstanding me. In my system, only one attack slot is used (unless there is more than one learner), and there really only exists one copy of the skill itself. However, there are four OTHER attacks that are set to target by slot. These attacks do NOTHING other than set the tag, and then chain to the actual skill. In effect, all they are doing is testing whether or not the learner is alive in his given slot. The skill attack is NOT set to target the previous target, but instead is set to randomly choose its target, as if it is a new attack. Thus the learner should not be targetted by the actual attack any more often than anyone else, only one tag is needed per attack (unless there is more than one learner), and the skill is never used if the learner is dead. That, by the way, is the main flaw - the enemy cannot use the skill if the learner is dead. But this will only be an issue until the skill is learned (see my post above).

So, as long as there is only one learner, my system uses five attacks per skill (the skill itself, plus the four slot checks); five copies of each enemy (one for each slot that the learner may be in, plus one for if the learner is not present, all assuming that you cannot change an enemies attacks using plotscripting); one attack slot per enemy (either the slot check that will chain to the actual skill, or just the skill itself); one tag for each attack (all four slot checks can safely target the same tag); and is learned by observation, but ONLY if the learner is alive WHEN THE SKILL IS USED.

Also, even though the topic began as "Blue Magic", I don't think we should allow Final Fantasy's definitions to dictate the limits of our systems. It is up to the game author how many skills can be learned by observation, and whether or not they are restricted only to magic. I'm trying to come up with as flexible a system as possible, and I'm pretty happy with what we've come up with so far. I may give a first try at making a contribution to the wiki with this.
_________________
My first completed OHR game, Tales of the New World:
http://castleparadox.com/gamelist-display.php?game=161

This website link is for my funk/rock band, Euphonic Brew:
www.euphonicbrew.com

Newbie_Power · Joined: 04 Sep 2006 Posts: 1762

Ok, I stopped caring whether there is one attack that chains, or four attacks on one enemy...

Your observation is simplified into two things in my system: Getting hit by the attack, or watching an enemy do a support spell (only because there isn't reflect yet).

For attacks that hit, it must actually hit. Miss or otherwise, then it doesn't set the tag on.

For attacks that a Blue Mage has to observe and not get hit by, it just checks to see if the Blue Mage is in the party before setting the tag.

Then the rest is done afterward, without much need to make five different enemies and switching them out before a battle starts.

I decided not to care about what Blue Mage attacks are defined. Just keep it different than other magic so it will actually be interesting and fun to use. That's the point of Blue Magic is: for semi-different strategy and formula in magic spells, not because Final Fantasy does it and you should too.

Instead of assigning attacks to tags, we need to find a way to use variables instead. That would be a better answer than both of our ideas.
_________________

TheGiz> Am I the only one who likes to imagine that Elijah Wood's character in Back to the Future 2, the kid at the Wild Gunman machine in the Cafe 80's, is some future descendant of the AVGN?

msw188 · Joined: 02 Jul 2003 Posts: 1041

It would be so lovely to be able to affect variables in battle. Even just one global variable reserved for battle targetability would be beyond mortal beauty. We can fake it using the extra stat right now, but that extra stat is visible on the status screen, a possible cause of confusion for the player. Plus, the extra stat is often used for many other interesting things already.

Newbie_Power · Joined: 04 Sep 2006 Posts: 1762

msw188 · Joined: 02 Jul 2003 Posts: 1041

Okay, I guess we are still arguing about something. Our ideas are NOT the same. How do your observable support spells check whether or not the learner is in the party, and/or alive and thusly able to observe when the spell is cast? I'll agree that your method would be easier for single target attack skills (although it would take up four of the enemy's attack slots), but I honestly do not see how it would work for any other skills without a chain similar to what I am doing. And then, I believe that it is easier to use the same method for all attacks, if only for ease of the script outside of the battle.

Also, I got the gist of what you meant about variables, but it is impossible to get by on them alone because they cannot be affected by battle. Naturally there will be some serious variable usage to run the loops necessary to check everything efficiently.
_________________
My first completed OHR game, Tales of the New World:
http://castleparadox.com/gamelist-display.php?game=161

This website link is for my funk/rock band, Euphonic Brew:
www.euphonicbrew.com

Newbie_Power · Joined: 04 Sep 2006 Posts: 1762

Moogle1 · Posted: Mon Apr 16, 2007 6:11 pm Post subject:

Ah, I see what you're talking about. That's actually a clever method, and it works for multitarget attacks.
_________________

Slime Salad :: Independent Game Design

Newbie_Power · Joined: 04 Sep 2006 Posts: 1762

Uncommon · Posted: Mon Apr 16, 2007 10:35 pm Post subject:

It is, of course, much less of a hassle to the player and the creator if you just use the Final Fantasy 6 method of just seeing a learnable attack happen (be they support or attack) instead of the Final Fantasy 7 method of having to be hit by them, given what we currently have to work with and the methods suggested already, especially since FF7 had its own methods of forcing the attack on the learner, something we, once again, do not have the capabilities for. The target-specific checking is just too much trouble with too little accuracy and I can't actually think of a practical application of that targetting method in any situation to begin with.

There really shouldn't be so much argument over this. Just have all of the spells use the same method as the support spell plan Newbie outlined above. The attack sets a tag, then an after-battle script checks to see if the Blue Mage is in the party and alive. If so, he learns the spell. If he's out of the party or didn't survive through the battle, the tag turns off. Simple.

Now, why doesn't he get the spell if he saw it before he died? Well, it's the same concept of a dead character not getting any EXP, whether he died at the beginning of the battle or right before it ended. When the battle engine has advanced to the point where we can cast aside this convention, we'll be able to look at better ways of incorporating blue magic, but as it stands, the method I've suggested is the easiest and most practical.

msw188 · Joined: 02 Jul 2003 Posts: 1041

Wow. I'm sorry, but I really feel like I'm talking to a wall right now. I understand all of what you are doing, honestly I do. The problem is that your logic doesn't support the case I originally pointed out as an issue, and I feel as though I keep repeating it, but maybe I'm not making it clear that this is the real reason I prefer my method:

A dead Blue Mage should NOT be able to observe anything.

In your system, suppose the BM is dead. The enemy casts the support spell, or spread attack spell. This turns the tag on, because the battle doesn't care and turns on the tag. We will take care of checking for the BM's existence outside of battle. If the BM is dead, we can decide not to give him the spell.

But what if the BM is revived mid-battle, after the support spell has been cast? Then your after-battle script will see that the tag is on, and that the BM is alive and present, and will award the spell. This to me is a serious flaw, logically inconsistent and counter-intuitive to the player. I do not see how your system can possibly correct this.

A second reason I prefer my method would be the following. As a player, I would expect my BM to learn all of his magic the same way. It seems counter-intuitive once again for some magic to be learned by observation, while other magic needs to be felt. But there is no way for a BM to 'feel' support or healing magic from the enemy without status effects like reflect, or confuse, or possibly some others that we don't have. This leads me to feel that all magic should be learned by observation, which my method does. But you have to be alive to observe. That is the main problem which my method was designed to solve.

As a side note, I don't think checking for death after battle is as much of an issue. As long as the BM was alive when the spell was cast, I would give him the spell afterwards. My script after the battle then would actually be quite simple, and it is the pre-battle script and in-battle stuff that gets somewhat complicated in my system.
_________________
My first completed OHR game, Tales of the New World:
http://castleparadox.com/gamelist-display.php?game=161

This website link is for my funk/rock band, Euphonic Brew:
www.euphonicbrew.com