Posts for FatRatKnight

1 2
31 32 33
43 44
Posted: 10/22/2010 12:30 AM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
Wow, months of discussion in a week's worth of time? That's really intense discussion you've been going through! All I can come up with is "Absolute Nuts". Overnight, just by my lonely self. Surely your name is far superior to my own. Incidentally, I'm fine being placed randomly at this moment. Any team I'm placed in should at least consider the name "Absolute Nuts". Far better than "Team 5", I think.
Posted: 10/21/2010 11:47 PM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
Battle plans: World I - Random battle, get GAZE - 3 random battles, on path to TELEPOR - P-FROG: GAZE, win. Get TELEPOR - KINGSWRD: GAZE, win. Try to lose a mutant here. Get POWER or +5 Mana - Monster duet: Manipulate solo SKELETON. Use physical force. Not sure if we can manipulate the right bonus while forcing a solo SKELETON - STEWARD: Physical force, win. Wonder if that COLT is an alternate way to die. A reset between battles, probably, unless the RNG just happens to have everything we need. - More random battles: We need 26 Mana minimum and two mutants with POWER. World II - SEI-RYU: Turn 1: POWER, POWER. Turns 2 & 3: Beat living daylights out of SEI-RYU. Use convenient humanoid as our shield. Get +5 Mana. World III MOSQUITO: GAZE, win. Get +5 Mana Monster trio: Seems we should want a monster for this fight, to have the HP we don't. DRAGON 1 is handy. Maybe not, I should see if a worthless ally can be a replacement. GAZE a foe that isn't cooperating with keeping ourselves intact. It looks quite possible with my test movie. Get +5 Mana in any case. SABERCAT: GAZE, win. Get +5 Mana BYAK-KO: bSTONE, win. World IV ATOM ANT: bSTONE or GAZE, win. EVIL EYE: Attempt to SAW and bSTONE until something works. Probably the SAW by round 2. If someone's going to die here, we have three (two?) pieces of fodder to die with: Original MUTANT F, POWER MUTANT M, and DRAGON 1 (or dead body). MACHINE: SAW. However, we might need to avoid resetting between this and the previous battle. There's enough steps to avoid any encounters between these two, thankfully. SU-ZAKU: SAW, win. Final C- SAW, win. SAW, win. SAW, win. SAW, win. SAW, win. SAW, win. Those are my battle plans, thus far. Any questions?
Posted: 10/21/2010 6:01 PM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
I am awake again. Now for further thoughts. I stopped testing at 1.7 million frames (wanted to use my computer as a lullaby, a sleeping aid, whatever. Don't ask). I've come to the conclusion that C30B and C34B will always remain a fixed value away from each other. I could keep it going if you like. In any case, this does technically have 16384 possible starting seeds. A good number of them are similar enough that we wouldn't care anyway. As for my simple formula, it would seem that it doesn't fit everything. And like you say, this is somewhat odd, considering that two different aspects of the game do things somewhat differently. It may be darn close, but you do observe differences between the shuffler and NPC directions, and that's a pretty good proof right there that the game isn't using something consistent between different things. Is "Add" really so rare? True, our tables may be old, but you've recorded seven instances of "Add" and zero instances of "Reset to original". This would be a particularly unusual case if what you say is true, as having only 42 values instead of 85 would suggest we didn't see so many instances of "Add" back then. BURNING test. It fails. I only deal damage if the attack doesn't incapacitate. (Yeah, I don't care if my hacking makes sense, I'm having a human and some bird burst into flames.) Recall how quick it is to get STRONG potions. Now think about replacing five of those swift purchases/uses (one trip to that item shop) for a detour to another shop or a box for a piece of equipment. Keep in mind two such items still do not replace the need for STRONG -- A 74 strength HUMAN M is short that one last point of strength needed for first turn kills, meaning we can't bypass the need for one STRONG here. Then again, we'd use a MUTANT M in this case, and grab our strength from the EVIL EYE or MACHINE. Still, the shoe shop is a step, maybe two, away from our path. We buy one and eventually wear it. This is one purchase at the cost of 32 or 64 frames of walking, instead of five purchases and uses of STRONG. Still, we have inventory limits, and with 6 STRONG and 1 GETA shoe in the main inventory, we have one slot left. I'm not sure what the dialogue involving BAND is like with a full inventory cluttering us up. As for the GIANT glove, we have two choices. We're picking up the bSTONE, but the GIANT glove shop is a significant detour even with the glider handy. Compare that to the difference of buying 1 STRONG or 6 STRONGs in a single shopping trip. There should be no contest here, the STRONGs win easy. Oh, right. MUTANT M for that last point, meaning we don't need the STRONG shop. Then again, count the steps for the STRONG shop and compare what it takes to get to the GIANT glove shop. As for the GIANT glove box in the prison area, how many frames was it to TELEPOR out versus going through it normally, again? If it was something like 3 seconds and the box happens to be reachable without wasting steps, go for it. I might be underestimating the time it takes to buy and use STRONGs, but I do have that quick test up there to check again. SEI-RYU dies in 3 rounds to a pair of POWER mutants swinging a BATTLE sword and a HAMMER, and some random humanoid being sacrificed to protect our otherwise fragile lives. Not sure about the fourth member, I need to check the Monster Trio to see if said fourth member can help out there.
Posted: 10/21/2010 9:13 AM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
I'm quite tired now. I'm going to check over the previous posts later. The latest test in my previous edited post had some stats I forgot to hack more towards the expected norm, but I did another test run with the stats properly hacked to what I expect to have, and it still works just fine. I still don't think a WHIP is going to help, especially with the test's results in. POWER isn't a hard ability to grab.
Posted: 10/21/2010 8:44 AM
(Edited: 10/21/2010 9:00 AM)
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
Weapon crits Not sure what we'd crit. Would be lovely to just have the KING sword if for no other reason than sheer damage, but no such break has been found. COUNTER Haven't checked. I have a feeling it won't help, considering we need to walk by a shop with these things first, or find a different counter ability. Are you thinking about BURNING as a possibility? HYPER We already kill everything in one turn by the time this thing shows itself. Does not look like it'll help. Use equipment to boost stats? There is a shop that is one or two steps to get to, and again to get back on track. This shop has a shoe for raising strength. This pushes the favor towards MUTANT M over HUMAN M... Almost. 54 Strength becomes 64, and it's only 3 battles away for perfect SAWing then. The glove is apparently too far away for a single round of help, however. Close, but probably still wouldn't give the MUTANT M time to prepare for SU-ZAKU and ASHURA. Regardless, sticking one on a HUMAN M is a trade-off of a couple steps versus a few extra STRONG uses. With both, I'd take the MUTANT M in a heartbeat -- 74 Strength just misses the golden 75 we need, so the human would need a detour for a STRONG regardless, while the MUTANT M picks up the strength off the EVIL EYE or MACHINE. But one isn't in easy reach, however. It is true that TELEPOR out of jail has a lot of walking anyway, this is something that might need to be looked into. Confusion, paralysis or sleep? Probably not. We're already chipping through bosses at a fair clip already. I still need to test SEI-RYU and the POWER unity of BATTLE sword and HAMMER. I'm going to try it now, darn it. Edit: That went well. Any questions whether WHIP will help in this case?
Posted: 10/21/2010 7:42 AM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
Bobo the King wrote:
[...] Say... do you think we might be able to get the coveted GAZE (stone) ability this way? [...]
While it may be true that that the RNGs do "wander" a bit, problem is, C30B and C34B are in exactly the worst possible spots in relation to one another. That is to say, no matter how the RNGs wander, the two addresses will always be an exact same number apart. I just told lua to go ahead and keep resetting until (C30B + 9)%256 isn't equal to C34B anymore (the reset process also zeroes them out, so checking for not equal as well to get around this). It hasn't finished. I'll keep it running for a long while, but don't raise any hopes.
Posted: 10/21/2010 7:10 AM
(Edited: 10/21/2010 7:27 AM)
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
I feel strongly inclined against whips. However, I won't dismiss it right away, though I'm tempted to skip any tests. Although, the detour to bSTONE does take around 20 seconds, and looking at the rounds of combat against SEI-RYU, adding four rounds due to no bSTONE might be roughly breaking even. This does not take into account the fact that in place of shopping for bSTONE, we're instead shopping for a WHIP, which will take time on its own. Keep in mind we will have no good damage option. BYAK-KO has 1000 HP to kill, and without bSTONE, we have no other insta-kills, unless we can summon up a MEDUSA monster on our side, but using a non-mandatory battle for that eats up more precious time. The whip should be dealing trivial or no damage, and we might be dealing 150~250 damage per turn starting with round 2. I want to say the WHIP will never take place of bSTONE. It's 20 seconds to the bSTONE and back. An extra round takes roughly 5 seconds, give or take, judging from the SEI-RYU battle. How long to the nearest shop with a WHIP or E-WHIP and back? I really am tempted to dismiss this without a single test. If there's further push for some sort of whip, I will check, though. Edit: Actually, already went ahead and did a bit of checking. The WHIP from a shop in World I and the E-WHIP in a shop from World III each takes around 10 seconds, maybe slightly less. Unless anyone's got brilliant ideas on what to add to our shopping list, or how to use whips more effectively than what I am able to see, the remaining half of the time from skipping bSTONE will be lost on extra rounds on BYAK-KO. An extra battle is almost certainly out of the question. Any ideas on a 2 or 3-round KO? 1000 HP, immune to death, no SAW. What's the battle plan? I also feel doubt that a WHIP would help before or after this battle. I still haven't gotten around to it, but I think I can beat SEI-RYU to a pulp with brute force and POWER abuse, no whipping. Everything after BYAK-KO is pure instant death all the way through. Again, just feelings, no tests to back it up.
Posted: 10/21/2010 5:01 AM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
0x00 ~ 0x2A - Divide by 2 0x2B ~ 0x7F - Add 0x80 ~ 0xD4 - Subtract 0xD5 ~ 0xFF - Reset to "original" These are what I expect after mild testing. Tell me if it's wrong. The random number addresses tend to "wander". If C300 is zero, the following six addresses could also be zero, or perhaps they all happen to be one. Maybe the first half is zero and the later half is one. When you generated those tables, it's possible we thought that C30E was always a fixed value away from C30B after a reset. I highly recommend "Tools -> Debug -> Memory Viewer..." to better understand the RNG's rules. It is hard to explain, but to me, it looks like the RNG addresses like to be in groups of 7 before incrementing. However, every 9th group, the group size is 8 rather than 7. ... Yeah, these words aren't any good, peek into the Memory viewer and look at C300 and beyond. Reset a few times, see if any "pattern" shows up. Looking at them as single addresses will do you no good, look at all 128 of them and get an idea. I have a small function for generating a range of numbers. It might match the game's way of handling an arbitrary range, or it might be off by some amount. I do not know, but it does a good enough job at approximating that I can't detect errors myself. Said function would be: 
Language: lua
--Insert the SuperRNGTable here

--*****************************************************************************
local function GenRange(R,D)
--*****************************************************************************
-- Returns an integer from 0 to D.
-- Might be imprecise. If so, we'll come up with something different!
    return math.floor((SuperRNGTable[R]*(D)/255)+.5)
end
The value R is the RNG value we read directly from memory after it is "rolled". The value D is the upper bound of the "random" number, with an assumed lower bound of zero. If, for example, a healing effect restores 20~35 HP, subtract 20 from the upper range to get our D. If there is a number of different effects with "equal" chances (the term "equal" being used loosely, as the distribution prefers to avoid the extremes), use (number - 1) for your D. I think this function fits well with the Ability Uses Shuffler, the numbers at the top of this post would match this function with a D of 3, and a small amount of trying them out led me to link the effects with the results I've been getting. If it's not true, I'm sure further tests will prove me wrong.
Posted: 10/20/2010 7:02 PM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
http://dehacked.2y.net/microstorage.php/info/858566825/FFL_HumanSAW1.vbm http://dehacked.2y.net/microstorage.php/info/1571730213/FFL_HumanSAW2.vbm No comments to hide these links with this time. Although, while hacking for test #2, I forgot to set the defense correctly for a BAND-wearing human. Not that it matters anyway. Excellent, we now have a complete RNG table! I'll be sure to update my scripts accordingly. Not exactly in a planning mood right now, more in a testing phase. I'm going to stare at SEI-RYU for a bit, followed by Monster Trio. Any ideas on what to pick at? Two POWER mutants will be one test, which requires I bring a manly guy for the free Strength-based HAMMER. Another test will be to see if I can do stuff with only 26 Mana from spells, probably not much. I'm thinking of a test against the monster trio that involves POWER, a DRAGON 1, and twice-applied GAZE. Extra points if my pet DRAGON 1 falls in the process (probably not, they have nice HP. Too bad). Other than these two battles, I don't see anything else that clogs up the route at the moment.
Posted: 10/20/2010 5:17 PM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
Beautifully done. Very much so. I'm worried you'll run out of glitches before reaching Bowser, or at least, different ways of toying with them. But if you're sure you still have a rather full bag of tricks, keep going! Found any interesting ways for Mario to die, by the way? So, you can actually get power-up #4, outside the usual 0~3? Wonder what 5, 6, 7, ..., 254, and 255 are like. Two shiny stars being spawned by the level goal are neat! Even crazier was the fact one of them counted as a power-up and gave you a cape! Then the now-useless invincibility came from the other star. Which you began the castle with, making it so not useless. I just took a look at that page. Hmm...
When you hit a goal tape, and spawn a starman (which is never), [...]
Well, "never" happens pretty often around here.
Posted: 10/19/2010 3:00 AM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
The submission text is updated. It has more information than the ambiguous "uses JPN-exclusive glitches". Is this good enough or do you need a TASer who knows the game to give further thoughts?
Posted: 10/18/2010 6:43 PM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
I use this page for my monster needs. According to it: ALBATROSS + SEI-RYU = DRAGON 1 DRAGON 1 + MEDUSA = CLAYMAN However... DRAGON 1 + WOODMAN = AMOEBA 22 Mana. Not much...
Bobo the King wrote:
[...] Since ASHURA has 90 Mana and we'll have 94 at that point, we can even STONE him. [...]
No. oCHANGE. No stoning. No death. Only SAW.
Bobo the King wrote:
[...] Now that I look at it, we'll even have 80 Mana before the fight with the EVIL EYE, so we don't have to sacrifice our GAZE strategy [...]
No. We need 95 Mana for GAZE or bSTONE to work first thing. We'll need extra rounds otherwise.
Bobo the King wrote:
[...] Let me flip it around and ask you if there's any reason not to manipulate Mana boosts from every forced battle from SEI-RYU onward. If my strategy works, how much Mana do you think we should have when we face SEI-RYU?
26, at the lowest. Achievable in four battles with a starting Mana of 6. My plans indicate we'll pick up TELEPOR from the P-FROG bandit. Just in time to see its use. The duet of monsters and STEWARD might be inconvenient to stick a reset in between. That leaves only two points to manipulate Mana from forced encounters prior to SEI-RYU. The reason for the 26 by SEI-RYU is so we can pick up 20 more, 5 from SEI-RYU himself and the remaining 15 from three forced encounters after. 46 Mana to turn BYAK-KO into instant art. How does this impact the late game? Is our primary Mutant F simply going to fade to the role of "dieing for the greater good" after the ATOM ANT? On a side note, here's the list of bonuses with C30B == 0xF2 on the route to P-BLAST, if you want to consider a second P-BLAST: -- Shuffle ability uses Mana+5 -- Mana+5 Defense+1 AB # P-BLAST Defense+3
Bobo the King wrote:
[...] the way I look at it, our battle with the monster trio would not really change as we were basically expecting it to take two rounds anyway. [...]
I'm getting a feeling it will change. If we do go through with my plan of 26 Mana when we fight SEI-RYU, we'll have 36 Mana here. Our P-BLAST would be pitiful. Secondly, we don't have much in the way of extremely useful target-all abilities on our other members, except possibly AMOEBA's 22 Mana ACIDs or a secondary mutant's QUAKE/P-BLAST of unknown Mana (26?). We might want to GAZE two of them dead while beating the stuffing out of the third with sheer damage, for once. DRAGON 1 is a start...
Bobo the King wrote:
Fight random battles before the Bandit's Cave to get GAZE, P-BLAST, and TELEPOR on our main mutant. I can't remember how many battles that is, but I think it's around seven.
6 randoms. The seventh comes from the P-FROG bandit himself. For the TELEPOR case, the first three battles come fairly quick, while the fourth feels like a hundred steps away.
Bobo the King wrote:
•Fight one random battle at some point to get QUAKE on our secondary mutant. •Fight one random battle at some point to get POWER on our secondary mutant.
Hard to say about QUAKE, we might want to skip over it or possibly replace with P-BLAST, considering the RNG path. Almost certainly would want POWER, though, unless we decide to pick up a BEETLE for SEI-RYU, which now seems less worth it as we're looking at exceptionally few random encounters. ... Actually, I'm starting to think P-BLAST is optional -- Manipulate the duet of monsters prior to STEWARD to be only 1 SKELETON and introduce it to stabby pain. A different plan for the monster trio doesn't use mutli-target anyway. But this might add an extra round to SEI-RYU, which can be hazardous.
Bobo the King wrote:
Buy the book of STONE before facing BYAK-KO. Continue to force Mana bonuses for the remaining forced encounters until it is maxed out.
The book is necessary, so of course we're keeping that. The mana gains, on the other hand, are almost pointless if we go with a plan to have 46 Mana to petrify him. 56 Mana versus EVIL EYE? Not good for insta-deaths. It'll be too late for any of the endgame bosses if we get stat-gains exclusively from required battles. This does free up manipulation for other things, but what do we need then?
Bobo the King wrote:
Draft a Human M in Northeast Town in World 4 and immediately boost him to 68 Str. with STRONG potions. Keep at least four additional STRONG potions on hand to boost him to 76 Str. after facing the MACHINE.
Agreed. Looks like things are starting to fall in place, but testing needs to be done.
Bobo the King wrote:
It isn't clear at all to me whether drafting a Monster will be worthwhile. Much of its benefit is dictated by factors that are difficult to measure. [...]
Quite true about the difficulty. Do we use a Mutant or a monster? Or perhaps both?
Bobo the King wrote:
[...] How much time (if any) will it cost us to manipulate luck such that we get simultaneous Mana boosts on two characters? [...]
A few seconds, at most. The toughest piece of manipulation doesn't seem to need more than 10 seconds, and this doesn't seem to be a tough bit of manipulation, at a glance.
Bobo the King wrote:
[...] How quickly can we kill off a character in World 1? [...]
That's the question I would like an answer to. We might want to ask KINGSWRD if he wants to assist in an answer. Wondering how the turn order will work out, but if we can force an instant death of one of our own members just before we release our own GAZE, this would be good.
Bobo the King wrote:
P.P.S.-- I just noticed we can conscript a Mutant F in World 4 with built-in P-BLAST and TELEPOR and 41 Mana. It's highly unlikely that either spell will be useful, but we may need them for some strange reason. If so, that's probably the easiest way to get them.
Unfortunately, in all the times where we could use a 41 Mana P-BLAST or a TELEPOR, we're already well past that point when she becomes available. Eh... Currently, I'm trying to come up with a few tests. I can modify RAM, so I probably don't need to run earlier parts to get what I need. I need to try out the EVIL EYE and MACHINE, as well as the final boss rush with SAW all the way up and the lacking Agl. this implies. The plan's no good if the test rolls up in failure. Give me some time to work out a few details before I dive in there, SAWs flying.
Posted: 10/17/2010 5:35 PM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
I took the test run, waited until it reached SEI-RYU, then I hacked in POWER, equipped the BATTLE sword, and gave it a good swing at the boss after the boosted strength. 230 damage. Maybe an idea for a secondary mutant. We can defeat SEI-RYU in 2 rounds with high mana P-BLAST and a supporting POWER + BATTLE sword mutant. No need for further growth of secondary mutant. C30B needs to be 0x94 or 0x95 to have it happen in one battle. There are other values that work, but it's more battles to get. The problem with these values is that someone is stuck getting a useless BARRIER as well. At 0x94, here's the first 9 gains according to my script: Def+3 AB # POWER AB # BARRIER -- HP+13 HP+12 Mana+4 HP+14 -- The POWER for SEI-RYU never really crossed my mind, so I never really considered it. I've said it before, and I'll say it again: GAZE (stone) is impossible to get. It's not merely 12 battles down the line with a 4-mutant party. I looked. Real carefully. With how the game works, it can't generate an RNG that will hand this ability to us in any number of battles. oCHANGE can be obtained, however. C30B just needs to be 0x4F or 0x50, and will show up very early. oDAMAGE is also impossible to get, by the way. Human + STRONG + SAW. We still have a problem at MACHINE (not enough Def for SAW to work right away from reset), but ASHURA would certainly go more smoothly, and possibly GEN-BU2, depending on resets. We just need 11 STRONGs for a Human M, two trips to the item shop dialogue. True, the lack of agility may prevent said Human from acting early, but my test run did fine with a low Agl Mutant M. ... Are you thinking we drop the Mana gains of our main Mutant in favor of SAWing all later bosses instead of bSTONE at some of them? EVIL EYE might need an insta-death other than SAW, and that's going to require 95 Mana to work. Otherwise, the next highest, BYAK-KO, needs a 46 Mana bSTONE to become art. The EVIL EYE's "requirement" of 95 Mana means we probably want to get 96 Mana anyway to produce art from some of the later bosses. But if we can bypass the need for the usual insta-death for EVIL EYE, that cuts an awful lot of Mana gains that we need, with the presence of 75 Str SAW for every boss after MACHINE. 95 down to 46 probably saves 10 battles by itself. This makes planning for SEI-RYU and the monster trio more difficult, eh? Still skipping AGILITY, but here's what grabbing a few STRONGs look like. Compare that to actually fighting 5 battles (True, some of those 5 would be better off being from boss fights). Okay, how many strats do we have for SEI-RYU by now? - P-BLAST only; Needs high initial Mana; 3 rounds - POWER support; 2~3 rounds, depending on how much Mana we pick up. - QUAKE support; 2~3 round, but we can build two mutants effectively. - BEETLE SAW; 1 round, if we're "lucky" with the RNG. Unknown path to get... Never thought I'd end up reworking the route again to see if there's a way we can pull it off with only 46 Mana on our game-starting Mutant.
Posted: 10/16/2010 7:25 PM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
Bobo the King wrote:
Yikes! My lack of attention to this run really showed with my last post!
It may have. But you've noticed, so we should be good.
Bobo the King wrote:
I somehow missed your 40 minute run. [...]
I got that impression from your previous post. Missing that would be painful, so I wanted to make sure you were aware of it. About the initial party, I was thinking there might be trouble killing off a few in World I if they're consistently too fast to let an enemy get a turn. The difference between M and F is that great, and since it is only important right in the first world, should we make a few tests? My GAZE, P-BLAST, and TELEPOR plan indicates that it isn't anywhere close to impossible to do, while still wasting no steps out of the main path. Well, actually, the GAZE, P-BLAST, and TELEPOR route takes some incredible amount of luck at the TELEPOR manipulation, since you need one address to be one value and another to be something else, and in this particular case, is a very rare happening. But possible. P-BLAST versus ASIGARU might simply be because it was faster than stabbing said foe dead. I can't remember. My notes about making 6 steps first thing before restarting was to make the manipulation for GAZE easier, I believe. Even if you manipulate the right value for GAZE, it is still possible to end up with an encounter we simply can't beat with SABER and RAPIER/HAMMER goodness, but being able to choose the forest or plains at will for the encounter allows us to avoid such nasty 40 or 60 HP foe no matter how the encounter stuff rolls. My test run did pick up two Humans F and some monster. As "luck" would have it, a Human F sacrificed herself for the greater good at SEI-RYU, and the monster was a good bodyguard at the three monster battle and was sacrificed at ASHURA. I think there wasn't any real purpose for the other Human F, so I made specific attempts to have her die somewhere. Not sure what changes should be done here, really. Offense from them were practically nonexistent. But I'm starting to think a QUAKE-toting spare MUTANT would be handy.
Bobo the King wrote:
Don't forget we can luck manipulate additional charges of the spells. What slots are you expecting them in? As long as there's nothing in slot 1, we can manipulate more uses of any spell we need. I think...
Apparently, from my really short luck manipulation test... 1) TELEPOR 2) P-BLAST 3) GAZE (death) 4) [empty] Not that it matters too much, as if we had nothing that has uses in slot 1, the game will skip over that slot when shuffling our uses around. I'm not exactly inclined towards BEETLE, myself. It would probably take at least 5 battles to get a monster at a meat level of 6 from the World II battles and earlier. I wouldn't worry too much about what encounters we can fight, but more what we can possibly face in the trip. I'm just sort of wondering if, in the case we do bother with a BEETLE, what we would do instead here. Besides, it seems impossible to win with a solo Mutant F -- someone must take a hit or kill SEI-RYU earlier. For the three monster party, it's possible we can pick a monster to supply the target-all ability for us for that three-monster battle. Otherwise, we need a living shield once again to protect our horribly undefended self.
Bobo the King wrote:
All right, let me try this again... If our Mutant M has 56 Str., we can expect the three battles to last 1.786 turns each for a total of 5.36 turns. [...]
Wrong idea. When we reset, we have the same old battle RNG. An action needs a 76% chance to connect for it to hit at all first time out (and will do so guaranteed). A Str. of 75 meets this for the SAW item, as does having a Mana stat at enemy-4 or greater. Anything less, and we're guaranteed to have to wait until the RNG decides to behave.
Bobo the King wrote:
[...] Tangentially, have you considered conscripting a female party member in World 4 and giving her a WHIP? I really do love stunning enemies with the WHIP...
I haven't considered that possibility. Stopping ASHURA's actions of doom has appeal, but I'm not sure if 52 Agl. is going to go first.
Bobo the King wrote:
What if we just pick up GAZE (stone) as a mutant ability?
If we do, we get unstoppable power for a huge portion of the game. SEI-RYU falls to us without any chance to put up a fight. The detour to the bSTONE is unneeded. We wouldn't need GAZE (death), as nothing that falls to insta-kill would resist insta-art.
Bobo the King wrote:
What are the drawbacks of that strategy?
The fact I checked every single RNG value and not a single one allowed us to learn this ability. For once, a faulty RNG has done horrible things to our plans. For the secondary target-all killer, I'm thinking monster. Here's the list of things that hit all with damage: - bFLARE (Uh, the expensive book? No. Not happening) - QUAKE - TORNADO - ACID - WHIRL - FLARE - P-BLAST Nothing we can really buy to get a target-all. That leaves a Mutant or a monster. Either that, or single-target kill the ones that survive with our spare characters (they're not left with much HP), or wait until round 2 for the second P-BLAST of death.
Bobo the King wrote:
[...] would it be possible to TELEPOR back to floor 1 after the ASHURA battle? [...]
I just checked. Unfortunately, we need to go through the cutscene. ASHURA's death apparently isn't the trigger.
Bobo the King wrote:
[...] please elaborate on using the Sound Test menu to manipulate luck. [...]
There's some tough piece of manipulation I've tried to do. Best I could get was 10 straight seconds of waiting just to get it (Slot 1 TELEPOR with encounters GGF). The Sound Test can be accessed during these 10 seconds. Apparently, by playing different music, at different times, the game shuffles the RNG differently. It doesn't matter how good our lua scripting is, it's still limited by the input we feed to the game (or outright cheating, but that's out of bounds for obvious reasons). This input can be directed into the Sound Test to play music that somehow messes with the RNG next time we reset. But this applies only to those pieces of manipulation that we can't get within a few seconds dedicated to a reset or two.
Bobo the King wrote:
[...] I guess it doesn't hurt to re-tread old ground, but I'm feeling kind of stupid for asking questions that have already been answered.
You forget things. I may well have forgotten a few myself. What's important is that we get back to where we were at before our break and continue from there. Reading older stuff is a good start, and if you have any of your own notes you can refer to, use them.
Bobo the King wrote:
[...] POWER and QUAKE. [...]
POWER is of no real help. Anything that could use a good SAWing dies in one or two rounds anyway. QUAKE might be a decent alternative to P-BLAST, at least, for that one battle. Or maybe SEI-RYU, too. It can be obtained in one battle, even with a two-mutant party. address C30B == 0x52 gets us QUAKE 3 "steps" down the line, while 0x53 is only 2 "steps". There's even a Mana+5 7 or 6 "steps", so we can pull that boost out from another battle without an intervening reset, assuming we use a 3 or 4 mutant party. There's potential here.
Posted: 10/15/2010 5:00 PM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
A previous post of mine seems to say that my plans involve grabbing GAZE (death), P-BLAST, and finally TELEPOR all by the time we dust off the P-FROG bandit. Another post seems to outline what we need to use for every single battle. Still a question mark around SEI-RYU, though. The three monster battle in World III can't be cleared in a single 99 Mana P-BLAST, which might encourage having another multi-target user handy. And don't forget my 40 minute test run from this post. It skips building Str. and Agl. on the main Mutant, in favor of dusting things dead with spells and picking up a stronger Mutant M who begins with 54 Str., where the SAW is handed to him instead. Finally, there's the resources I've filled up. I've indicated that there are magic numbers that will let us hit with SAW or bSTONE first thing. We won't need more than 75 Str. or 96 Mana by endgame, as the first action we throw out there will always hit at that level. (I seem to have forgotten STEWARD in the forced encounter notes. So trivial even I forgot.) In any case, here's how things look in my own world: As we start * Assemble party of four Mutants. The primary character needs to be MUTANT F, thanks to things like higher starting Agl. and the ever yummy SABER. The other three may be either gender, but I'm thinking 2 M and one F here. * Names. We can use 0-character names for the odd fact the game allows nameless characters and the fact it's fastest, or we can give single letter names so to distinguish who is doing what. World I * Get GAZE (death), P-BLAST, and TELEPOR, in that order, by the time we defeat P-FROG bandit. * * My notes indicate an optimal first reset time is 6 steps out into the overworld. * Abuse TELEPOR to leave Bandit Cave and go to King Armor. * Abuse TELEPOR to shorten the trip to King Sword. * * He dies to GAZE. Can't take us staring at him? * Abuse TELEPOR (poor spell...) to get to King Shield in a hurry. * * ... But we're still too late. Darn plot. P-BLAST the resulting battle. * * STEWARD dies to... Well, just about anything. * * Somewhere along the way of collecting these pieces of junk, there may be reason to... dispose... of the extras. I can't seem to remember why, though. * * It would be best to spend the spare spell charges we got on various battles to jack up our Mana among our TELEPORing. * Statue stuff. Joy. * * Abuse TELEPOR (That's just plain mean!) to skip GEN-BU. * Go on our merry way to the tower. * * Definitely talk to that NPC. We don't want game-stopping glitches, do we? * * The inn sounds favorable at this time. We'll need to reload our GAZEs and P-BLASTs. * * Possibly get replacement characters at this time. World II * Let's see... 5 GAZEs and 2 P-BLASTs means we can get another 35 Mana on our way to SEI-RYU. We may need that Mana -- SEI-RYU doesn't look beatable through instant win means at this time. Except maybe a BEETLE's SAW, but I'm not sure how that would fit in my plans. * Sanities like the BATTLE sword, sailing island, AIRSEED, the usual path to SEI-RYU. * Kill SEI-RYU dead. * * Probably with 3 P-BLASTs. * * If we did bring a BEETLE, we might want to fight an ATOMCRAB to change up the RNG for the well-anticipated SAW. * Old man's riddle. Good thing we got the BATTLE sword ahead of time. TELEPOR to Tower. Fun. World III * Our approximate stats by this point should be 4 Str., 8 Agl., and piles of Mana. * * The turn order when we use spells doesn't depend on our Agility, but Mana instead. Or so it seems. * * Our Strength is useless. We can't make use of it. Because we're going to pick up someone who starts with decent Strength later. * Advance the story. Possibly an inn stop if we need extra spells. * * MOSQUITO dies to GAZE. * * P-BLAST the trio of monsters. Though, one won't be enough... * * SABERCAT dies to GAZE. * TELEPOR out of the prison. Note: If you're going to put someone in jail, make sure they don't have an inherent escape mechanism! * Detour for Book of STONE in the hidden town. * Our freshly gained "Instant Art Kit" comes in use immediately against BYAK-KO. * Sphere, TELEPOR, fun times. World IV * The SAW, man. Don't forget the SAW! * Advance the story. * * Going the "usual route" through the sewers seem to be the way. * * ATOM ANT dies to GAZE * Go straight to the key items. Do not pass GO, do not collect 200 G. Or any optional item. * * The path over some of the buildings don't seem to increment the encounters. Use them! * * SU-ZAKU's encounter rate off the bike is 64. Ouch. * * On the way to depositing our ROM, summon a MUTANT M to our cause. He has the 54 Str. that we don't have, and didn't feel like fighting 10 battles for. * More underground stuff! * * EVIL EYE dies to GAZE. * * MACHINE dies to SAW. We might want to avoid resetting since EVIL EYE so the RNG is good for our otherwise incompetent 54~59 Str. * * We got a BAND from So-Cho. This is just about unavoidable anyway. For some reason, I decided to equip it on my run. * Okay, time for the walk to SU-ZAKU. * * I really think walking there is fastest. Naturally, we should time it. * * SU-ZAKU dies to bSTONE. I love our Instant Art Kit! * * That BAND protects versus Blind (and everything else). Let SU-ZAKU fail to blind the wearer if it's a problem. * More tower stuff. Probably no TELEPOR this time, I guess. Final climb * The NPC we talked to earlier doesn't exist anymore. Hence, isn't messing with our door. If we had talked to the NPC now, we must leave the area and reenter to get Tower2 working properly. * ASHURA dies to SAW. This battle may be troublesome... We might need a spare meatshield, considering my lack of any Agility and whipping won't do any good. Either that, or buff the Mutant M's Str. to 75, but that requires extra battles. * GEN-BU2 dies to SAW. With no resets, this should be easy. * SEI-RYU2 becomes art. bSTONE, go! We'll need 96 Mana on the Mutant F by now. * BYAK-KO2 becomes art. Hey, this book is a good read! * SU-ZAKU2 becomes art. Our power is simply petrifying! * CREATOR dies to SAW. We should have 75 Str. on our Mutant M by now. That's my plans. We differ on stat growth. You say we should get Str., Agl., and Mana on the same Mutant. I say we get Mana on the starter and pick up a boyfriend for this mutant and let him handle the Strength needs for us. We differ in some battle plans. Odd, you don't have bSTONE considered. I do believe we'll need it. See my test run on the writings in action! We differ in the use of EYEDROPs. ... I didn't need it in my test run, anyway. Silly detail, I guess. Do we want a second character to blast things with? Well, besides the Mutant M I pick up in the test run specifically for the SAW. The only times this matters much is against SEI-RYU and the trio of monsters. Against every other boss out there, I either deal enough damage to instantly kill or skip the damage numbers entirely. It's hard to say whether we need it or not. I'm leaning no, but again... Hard to say. There's probably a few details I'm missing. Feel free to consider all our plans and tell me of any holes.
Posted: 10/14/2010 6:59 PM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
After a revelation and/or epiphany, I would like to sign up. I have something to do, but this might be a good distraction. I generally don't like to put myself into using other emulators I'm unfamiliar with, but this gives me a perfect excuse to actually use it instead of staying away from it. (Those I used are FCEUX, Snes9x, VBA, and DeSmuME. Perhaps I will add one more to this list as contest stuff gets decided.)
Posted: 10/14/2010 6:40 PM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
Bobo the King wrote:
[...] I hope it'll still be a good pep talk.
And indeed, it has. I have been successful in making three published TASes so far. Although, one of them is pathetically short, it did still go through more than one revision (but it was still quick). Another also went through a revision, and this did take some more dedication. The last of the three I'm mentioning did have some breaks in its creation, but I had to go through two different long term routes of nearly equal lengths with strong TAS effort to see which one's better in my test run. Every run had some sort of test run going through the whole game first. This talk made me recall my current successes thus far, and the troubles they had. Well, [1199] NES Dungeon Magic: Sword of the Elements by FatRatKnight in 01:02.83 wasn't much trouble at the time, now that I think about it. But one of the runs, at the least, did have trouble. Indeed, as you say, when the two of us were working together, I really felt that everything I'm doing was getting well-received and I could actually ping some information into someone's hands and get a useful result coming back. When you stopped, it became a one-way channel. With little interaction, my motivation took a solid hit. Those Guidelines seem to have nailed it. The first advice that comes out, the first sentence after the overview, says going alone isn't usually a good idea. Right now, I'm somewhat burnt out for this run. I need you to lead the way for the moment. Even if all you do is ask "So... What's the plan?" That's all I need, as far as I believe. This gives me a direction to reach for, something that an outside force wants from me. A destination in which I feel I can get to and tell you, "here it is." Incidentally, this also made me think the Dream Team Contests have a lot more value in them than what I had at first glance. I agree. We need to push. Try to get this new run out to the end. One last time.
Posted: 10/14/2010 5:57 PM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
DarkKobold wrote:
There currently is not a tool like that which I am aware of. I'm sure someone could whip up a pretty simple Lua script using FatRatKnight's buttons script for you. I'll see what I can do. I've been meaning to try that script.
It'll fail for this application -- Snes9x doesn't have run-while-paused lua code. Well, 1.43 and 1.51 don't. 1.52 just barely does, but it's lacking in many lua tools at the moment, to my knowledge. But basically, if you get it to work, I will ask: How?
Posted: 10/13/2010 9:34 PM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
Here's my current progress. Haven't made further attempts to manipulate luck, but I think accessing the Sound Test is the way to go in this instance. At the moment, it's... Sort of abandoned by me. The "names" were decided over IRC, but I'm still not sure if they're the best way to go. Technically fastest, and it's interesting that the game allows 0 character names, but... I stopped in the attempt at manipulating luck by Sound Test. I can't remember my exact plans now, but I have plenty of notes, and I think I have a bunch of them posted here anyway. I recall worrying about how to successfully kill off the three other characters I have tagging along, all Mutant F. I may actually need to use slower Mutant M so that they can get a beating somewhere down the line. I can't actually remember why, though. Maybe for a monster later? It's been too long since I last looked at this. The JPN run uses JPN-exclusive glitches, so no worries about this run being "suboptimal". Even so, game-skipping glitches can be ignored for the sake of showing more of the game, with a few examples already published. Regardless, I tend to do one project, then suddenly drop it forever and decide that another is more interesting at the time. I keep thinking about my Life Force/Salamander run, for example, and even open up the emulator on occasion for it, but can't even add a single frame into it. If you can pick up on what I have so far for FFL, I believe I will do what I can to support that, but I can't seem to get things done myself. I can get started, and pretty well at that, but finishing is not a strong point of mine.
Posted: 10/6/2010 3:46 AM
Post subject: adelikat, here's a King Hippo script
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
Apparently, I was asked to go ahead and mess with more lua stuff. So here's more. It seems to work well, I think, but maybe the lower BND should be more negative. Maybe the cheat function needs tweaking. Hrm... Still, much of the same old code is used. Yay. Download PunchOutTraining_Hippo.lua
Language: lua
-- For Mike Tyson's Punch Out!! when fighting King Hippo
-- Intended to help time hits in real time.
--FatRatKnight

local AutoCheat= true

local CLOCK= 0x0306 -- Game clock
local HEART= 0x0324 -- Mac's low digit for his heart counter.
local ETMR=  0x0039 -- Enemy timer address
local EHP=   0x0398 -- Enemy HP address
local HRT=   0x0325 -- Mac's Heart's been dropped indicator
local RN=    0x0018 -- Some sort of random number stuff?
local ACT=   0x0090 -- Something to do with Hippo's thoughts.

local TMR=120      -- For the silly line.
local UPP= 25      -- Frames prior to the golden zone to display.
local BND= -8      -- KEEP NEGATIVE!! Frames after the golden zone.
local OFFSET= 13   -- Timing frames after Hippo decides his punch.

local DISPx= 180
local DISPy= 180
local DISPx2= DISPx+11 -- Right side of box. Adjust that plus to your need
local Dsp= {line= true, bar= true, stealth= 0, stick= 45}

local deviation= 0
local accuracy= 0
local perfect= 0
local earlies= 0
local lates= 0
local accearly= 0
local acclate= 0

local LastHit
local HitTiming

local DEBUG_Status= "nil"

local keys= input.get()
local lastkeys= keys
--*****************************************************************************
function UpdateKeys()  lastkeys= keys; keys= input.get()  end
function Press(btn)  return (keys[btn] and not lastkeys[btn])  end
function NULLFn() end
--*****************************************************************************

--*****************************************************************************
local function ForceRNToBehave(addr)
--*****************************************************************************
    if memory.readbyte(addr) ~= 9 then
        memory.writebyte(addr,9)
    end
end

if AutoCheat then
    memory.registerwrite(RN,ForceRNToBehave)
end

local Pad, LastPad= {}, {}
--*****************************************************************************
local function AB()
--*****************************************************************************
    LastPad= Pad
    Pad= joypad.get(1)
    return ( (Pad.A and LastPad.A) or (Pad.B and LastPad.B) )
end


local timer= 0
--*****************************************************************************
local function IdleStuffs()
--*****************************************************************************
    timer= timer-1
    if timer <=0 then
        LastHit= nil
        emu.registerafter(NULLFn)
    end
    memory.writebyte(RN,    9)
end

local MainLoop  -- There exists a MainLoop. Needed for a few function.

--*****************************************************************************
local function ForgetIt()
--*****************************************************************************
    DEBUG_Status= "... Yeesh..."
    HitTiming= nil
    emu.registerbefore(MainLoop)
end

--*****************************************************************************
local function ScanHit()
--*****************************************************************************
    DEBUG_Status= "Scanned!"
    timer= Dsp.stick
    emu.registerafter(IdleStuffs)
    LastHit= HitTiming

    deviation= deviation + HitTiming
    if HitTiming < 0 then
        accuracy= accuracy - HitTiming
        lates= lates+1
        acclate= acclate - HitTiming
    elseif HitTiming > 0 then
        accuracy= accuracy + HitTiming
        earlies= earlies+1
        accearly= accearly + HitTiming
    else
        perfect= perfect+1
    end
    HitTiming= nil
    emu.registerbefore(MainLoop)
end

--*****************************************************************************
local function PunchLoop()
--*****************************************************************************
    DEBUG_Status= "He's punching!"
    HitTiming= HitTiming-1
    if HitTiming < BND then ForgetIt()
    elseif AB() then
        ScanHit()
    end
end

--*****************************************************************************
local function NewPunch()
--*****************************************************************************
    DEBUG_Status= "Oh, snap!"
    emu.registerbefore(PunchLoop)
    HitTiming= OFFSET
end

--*****************************************************************************
local function WatchLoop()
--*****************************************************************************
    DEBUG_Status= "He's waiting..."
    local HippoStatus= memory.readbyte(ACT)

    if (HippoStatus == 23) or (HippoStatus == 24) then
        NewPunch()

    elseif HippoStatus ~= 1 then
        HitTiming= nil
        emu.registerbefore(MainLoop)

    else
        HitTiming= OFFSET+memory.readbyte(ETMR)
    end
end

--*****************************************************************************
local function NewWatch()
--*****************************************************************************
    DEBUG_Status= "Oh, shiny!"
    HitTiming= OFFSET+memory.readbyte(ETMR)
    emu.registerbefore(WatchLoop)
end

--*****************************************************************************
function MainLoop()
--*****************************************************************************
    DEBUG_Status= "Sleeping"
    local HippoStatus= memory.readbyte(ACT)
    if HippoStatus == 1 then
        NewWatch()
    end
end

emu.registerbefore(MainLoop)




--*****************************************************************************
function ResetStats()
--*****************************************************************************
    deviation= 0
    accuracy= 0
    perfect= 0
    earlies= 0
    lates= 0
    accearly= 0
    acclate= 0
end

local SparklyJoy= 0
local SparkyColors={"red","white","green","blue"}
--*****************************************************************************
local function DisplayStuffs()
--*****************************************************************************

--I'm handling options just for you!!
    UpdateKeys()
    if Press("insert")   then Dsp.line= not Dsp.line end
    if Press("delete")   then Dsp.bar=  not Dsp.bar  end
    if Press("home")     then Dsp.stealth= Dsp.stealth+1 end
    if Press("end")      then Dsp.stealth= math.max(Dsp.stealth-1,0) end
    if Press("pageup")   then Dsp.stick= Dsp.stick+1 end
    if Press("pagedown") then Dsp.stick= math.max(Dsp.stick-1,0) end
    if Press("numpad5")  then ResetStats() end
    if keys.leftclick and lastkeys.leftclick then
        DISPx=  DISPx  + keys.xmouse - lastkeys.xmouse
        DISPx2= DISPx2 + keys.xmouse - lastkeys.xmouse
        DISPy=  DISPy  + keys.ymouse - lastkeys.ymouse
    end

    if Press("numpad0")  then
        AutoCheat= not AutoCheat
        if AutoCheat then
            memory.registerwrite(RN,ForceRNToBehave)
        else
            memory.registerwrite(RN,nil)
        end
    end

-- Okay, enough options handling. On we go!
  if Dsp.bar then
    for i= BND, UPP do
        if i <= 0 or i > Dsp.stealth or LastHit then
            local Y = DISPy - i*4
            if     i < 0 then Y= Y+2
            elseif i > 0 then Y= Y-2 end

            local color= "black"
            if     i == LastHit   then color= "red"
            elseif i == HitTiming then color= "green" end

            gui.box(DISPx, Y, DISPx2, Y+2, 0, color)
        end
    end

    if LastHit == 0 then  --Sparkly joy!
        SparklyJoy= (SparklyJoy%3)+1
        gui.drawbox(DISPx-2, DISPy-2, DISPx2+2, DISPy+4, 0, SparkyColors[SparklyJoy])
        gui.drawbox(DISPx, DISPy, DISPx2, DISPy+2, 0, SparkyColors[SparklyJoy+1])
    else
        SparklyJoy= 0
        gui.drawbox(DISPx-2, DISPy-2, DISPx2+2, DISPy+4, 0, "green")
    end
  end

  if Dsp.line then
    local ThisX= 240
    local ThisY=  60
    gui.box(ThisX-1,ThisY,ThisX+1,ThisY+2+TMR-BND, "black", "white")
    gui.pixel(ThisX,ThisY+1+TMR,"white")
    if HitTiming then gui.pixel(ThisX,ThisY+1+TMR-HitTiming,"green") end
  end

    if LastHit then
        if LastHit < 0 then
            gui.text(128,80,"Late  " .. -LastHit)
        elseif LastHit > 0 then
            gui.text(128,80,"Early  " .. LastHit)
        else
            gui.drawbox(127,79,142,88,"green")
            gui.text(128,80,"OK")
        end
    end

    gui.text(  1,110,"Timing error: " .. accuracy)
    gui.text(  1,120,"Deviation: " .. deviation)

    gui.text(  1,140,"Perfect hits: " .. perfect)
    gui.text(  1,160,"Early: " .. earlies .. " : " .. accearly)
    gui.text(  1,170,"Late: " .. lates .. " : " .. acclate)

    local Tot=(perfect + earlies + lates)
    if Tot ~= 0 then
        local temp= perfect / Tot
        temp= math.floor(temp*1000)/10
        gui.text(  1,185,"Accuracy: " .. temp .. "%")

        temp= deviation / Tot
        temp= math.floor(temp*100)/100
        gui.text(1,195,"Dev.Rate: " .. temp)
    else
        gui.text(  1,185,"Accuracy: --.-%")
        gui.text(1,195,"Dev.Rate: -.--")
    end

    gui.text(145, 14,memory.readbyte(EHP))
    gui.text(  1,  1,DEBUG_Status)
    gui.text(230,  1,Dsp.stick)
    gui.text(  1, 50,"Punches: " .. Tot)

    if AutoCheat then   gui.text(88,1,"Eternal Cheat Acitve!")  end
end

gui.register(DisplayStuffs)


--*****************************************************************************
function CheatEternalBattle()
--*****************************************************************************
    memory.writebyte(CLOCK, 0)
    memory.writebyte(EHP,  96)
--    memory.writebyte(HEART, 7)
end

while true do
    if AutoCheat then CheatEternalBattle() end
    emu.frameadvance()
end
Edit: Now forces the memory to stay put sub-frame.
Posted: 10/6/2010 2:09 AM
Post subject: Someone use the subject headers, too!
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
There's the subject header in posts. True, they're small and sort of out of the way -- Who noticed the "Someone use the subject headers, too!" I put as my subject? Regardless, might not be a bad idea to put a useful note in there. Take a look at Critical five's posts, the header is getting use there! System, game name, length of WIP, something. If someone actually looks there, why not show something of value? Besides, TASVideoAgent tells the IRC channel what the subject is, right? Just a thought about making use of that tiny bit of text.
Posted: 10/5/2010 2:43 AM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
I have a bad habit of abandoning projects. I have started on 3 other runs that I never finished, and the two GB ones I picked looks like they may end up numbers 4 and 5 on my list, regardless of the result here. Even if the submission never existed, there was still a high risk of that happening with me. As it may be, I'm still waiting on knowing what glitches work only on the JPN version.
Posted: 10/5/2010 1:48 AM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
With a little luck, 4 out of 6 battles involved a single Clawbug. Any encoders? Currently deciding my shopping list. I originally considered grabbing 8 Red Charms, but I think I might want to pick up a few Hide Belts. Two on the Survivalist should be enough to have him survive one strike from a Fender, where Lv3 Poison wins the day. The Mole + Fender group is worth 80 Exp, which divided on a party of 5 gets 16 per person. Not bad, considering the alternatives (Venomfly is immune to poison, which cripples some early encounter-related options on the second floor). In order to help me plan things, I've studied the encounter data a bit and now holding on to these in my txt files, among other maps: 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ A 40% Treerat  Treerat              12xp
~,FFFFFFFFF~DDD~CCCC~III~ III~~~~~~   30% Mole                          15xp
~~~~~~~~~~E~,DD~C~~C~I~I~~~~I~~~~~~___30%_Treerat__Woodfly__Treerat_____24xp
~E,E~~~~~~E~,DDCC~~C~I~IIIIII~~~~~~ B 30% Treerat  Treerat              12xp
~EEE~DDD~~E~,DD~~~~C~I~~~~~ ~~~~~~~   40% Mole     Woodfly              27xp
~EEE,DDDEEE~DDD~CC~C~I~G~~HHH~~~~~~___30%_Treerat__Woodfly__Treerat_____24xp
~EEE~DDD~~~~~~~~~C~C~I~GG HH ~~~~~~ C 25% Treerat  Woodfly  Treerat     24xp
~~~~~~A~~~~~~~~~~C~C~I~G~~HHH~~~~~~   25% Mole                          15xp
~BBBBB.BBBBCCCCCCCCC~I~G~~~~~~~~~~~___50%_Woodfly__Woodfly__Woodfly_____36xp
~B~~~~A~~~~~~~~~~~~C,I~G~~HHH~~~~~~ D 30% Mole     Woodfly              27xp
~B~AAAAAAABBBBBCCC~C~I~GG HH ~~~~~~   30% Mole     Mole     Mole        45xp
~B~AAAAAAA~~~~~~~C~C~I~G~~HHH~~~~~~___40%_Clawbug_______________________34xp
~B~AAAAAAA~,EEEE~C~C~I~G~~~~~~~~~~~ E 40% Mole                          15xp
~B~AAAAAAA~,EEEE~B~C~I~G~~HHH~~~~~~   30% Clawbug                       34xp
~~~~~~.~~~~~~~~,~BCC~I~GG HH ~~~~~~___30%_Mole_____Mole_____Mole________45xp
~EEE~~.~~DDD~~~C~B~C~I~G~~HHH~~~~~~ F 40% Treerat  Woodfly  Treerat     24xp
~EDE~~.~~DDDCC~C~B~C~I~G~~~~~~~~~~~   20% Treerat  Clawbug  Treerat     46xp
~EEE~...,DDD~CCCCB~C~~~G~~~~G~~~~~~___40%_Mole_____Mole_____Mole________45xp
~~D~~...~~~~~~~C~C~C,GGGGGGGG~~~~~~
~~DD~~.~~D~~~~~C~C~C~~~~~~~~~~~~~~~
~~~D~~,~~DDDDCCC~C~C~~~~~~~~~~~~~~~
~D~D~~~~~~~~~D~~~C~C~~~~~~~~~~~~~~~ G 30% Venomgel Venomgel            462xp
~DDDDDDDDDDDDDD~EC~~~~~~~~~~~~~~~~~   30% Sleepgel Sleepgel            432xp
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~___40%_Sleepgel_Venomgel____________447xp
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ H 30% Sleepgel Venomgel            447xp
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~   30% Waspior  Sleepgel Waspior    744xp
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~___40%_Waspior__Waspior_____________528xp
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I 30% Spider   Sloth    Spider    1042xp
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~   30% Sleepgel Sloth    Sleepgel   852xp
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~___40%_Waspior__Sloth____Waspior____948xp


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ A 30% Venomfly Venomfly             78xp
~,AAAAAA~.BBBB~B~BCC~CCCCCCCCDDD~,~   30% Mole     Venomfly Mole        69xp
~~~~~A~~~~~~~BBB~B~C~C~D~~~~~~~D~D~   40% Treerat  Hare                 42xp
~AAA~A~AAA~BBB~B~B~C~C~D~,DDDD~D~D~
~B~A~A~A~A~~~B~BBB~CCC~D~~~~~D~D~D~ B 40% Hare     Hare                 72xp
~B~A~A~A~ABB~B~~~B~D~C~DDDDDDD~D~D~   20% Venomfly Woodfly  Venomfly    90xp
~B~A~A~A~E~B~B~C~B~D~C~~~~~~~~~D~D~   40% Fender   Mole                 80xp
~B~AAAAA~E~BBB~BBB~D~CCCCCCDDD~D~D~
~B~~~~~A~E~~~~~~~~~~~~~~~~~~~DDDDD~ C 40% Mole     Venomfly Mole        69xp
~BBBBBBB~EBBBDBBDEBBBBEBBB~DDD~D~~~   20% Venomfly Venomfly Venomfly   117xp
~~~~~~~~~~~~~~~~~~~~~~B~~~~~~~~~~,~   40% Treerat  Hare                 42xp
~,~DDD~DDD~DEEED~C~CCECCCE~DDDDDDD~
~D~D~E~D~D~E~~~ECE~C~~~~~C,D~~~~~~~ D 50% Fender   Mole                 80xp
~EEE~EEE~EEEEEEE~CCCECCC~C~DDDDDD,~   25% Hare     Hare                 72xp
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~   25% Venomfly Woodfly  Venomfly    90xp
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ E 30% Venomfly Woodfly  Venomfly    90xp
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~   20% Venomfly Fender   Venomfly   143xp
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~   50% Venomfly Fender   Venomfly   143xp
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
I seem to be able to manipulate encounters just fine. I'm considering picking up a bow for the Survivalist and hanging onto a knife for manipulation purposes. A pierce attack and a slash attack rolls R2 a different number of times, and I'm not really worried about a single point of difference in ATK. As I see it, I have two obvious options: Beat down Fender population, or do immediate quests. Either way, I need to use the inn to reload my poisons. As for the map... Well, me and lag got into an argument. Now I don't want to deal with lag anymore. Seeing as it's a test run, I don't see it critical that I mess around with the map, but I sort of have it sitting in a corner collecting dust and doing nothing useful for anyone. Hoping for a few thoughts. Wondering how poisoning Fenrir through that 95% resistance sounds...
Posted: 10/2/2010 6:17 PM
Post subject: Super Mario World -- Powerup buttons script
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
Something small. Use with Snes9x while playing Super Mario World. The purpose is to let you click one of four buttons using the mouse to change powerups at will. I hope it's a nice example on how to make use of these buttons. Download SMWPowerupButtons.lua
Language: lua
require("YummyButton")

local PowerUpAddr= 0x7E0019

local function SetPowerUp(B)
    memory.writebyte(PowerUpAddr, B.tag)
end

local T= {
[0]= NewBtn{x=  1, y= 3, fn= SetPowerUp, text= "s", tag= 0, color= 0x00FF00C0},
     NewBtn{x= 17, y= 3, fn= SetPowerUp, text= "B", tag= 1, color= 0xFF0000C0},
     NewBtn{x= 33, y= 3, fn= SetPowerUp, text= "C", tag= 2, color= 0xFF0000C0},
     NewBtn{x= 49, y= 3, fn= SetPowerUp, text= "F", tag= 3, color= 0xFF0000C0}
}

local LastScan= 0
local function CheckPower()
    local Scan= memory.readbyte(PowerUpAddr)
    if Scan ~= LastScan then
        T[LastScan]:recolor(0xFF0000C0)
        T[Scan]:recolor(0x00FF00C0)
        LastScan= Scan
    end
end
emu.registerafter(CheckPower)
Posted: 9/30/2010 2:11 AM
Quote
FatRatKnight
Editor, Experienced Forum User, Published Author, Skilled player (1174)
Joined: 9/27/2008
Posts: 1085
Try a playaround. There's this SMB3 example here, which doesn't place time as high priority. "Fastest possible" is already done or being done on practically every stage already, so an any% warpless probably wouldn't show anything that the other categories already have shown, other than a curiosity detail or two. Speed is covered a bit too well already, in my opinion. As I understand TASVideos, they want to create entertaining runs. There have been rejections, even when it is a "technically perfect" run of a game that hasn't been done yet. The run just isn't appealing to the viewers. And there are runs that don't have any particular goal other than fooling around and reaching some sort of ending so it doesn't feel abruptly stopped. If you don't use time as the main factor of your run, I'm sure you'll find plenty to do that other runs haven't done. Let Mario die in the most unexpected ways possible. Devise unforeseen challenges that becomes apparent to the audience as you're doing them. Find ways to make the audience think there's just no way you'll survive that, then survive it of course! Although, I'm not sure how simple a playaround is. It would take creativity to keep it entertaining throughout without placing time at high priority. This is simply a suggestion from me.
1 2
31 32 33
43 44