Posts for Nitrodon

1 2 3
13 14
Posted: 2/10/2011 5:44 PM
Post subject: Re: slots
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
antd wrote:
Slots is acquired by using Dice 8 times.
The guides I've seen indicate that Slots is acquired by making Cait Sith kill 40 enemies. Are these guides in error?
Posted: 2/4/2011 6:43 AM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
1) The total number of points is 39 = 3*13. This means that there were either 13 matches with 3 points/match or 3 matches with 13 points/match. The former is impossible, since a+b+c must be at least 6. Bob won the second match, and he got fewer points than he would if he got in second and third in the other two matches. Hence, he got in third in the other two matches, and b-c=3. Thus b is at least 4, and c is at least 1. Charlie must therefore have gotten in at least second place in the first and third matches, and at least third place in the second match. This gives a minimum of 4+1+4=9 points, which is exactly what he got. Hence c=1, b=4, and a=8. The results are as follows: Alice: 8+4+8=20 points. Bob: 1+8+1=10 points. Charlie: 4+1+4=9 points. 2) Let x be an element of the intersection of S1 and S2. By the third statement, x is either in P1 or in P2. If x is in P1, then it is in P1 and S2, and hence P2. If x is in P2, then it is in P2 and S1, and hence P1. Thus, either way, x must be in both P1 and P2, and this proves the result.
Posted: 1/16/2011 9:51 PM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
f(x,y) = (xy+1)^2 + x^2 As a sum of two squares, this is clearly nonnegative everywhere, and zero if and only if x=0 and xy=-1 (and thus can't be zero anywhere). For arbitrarily small x, we can find a y satisfying xy+1=0, and thus f gets arbitrarily close to zero. I actually got this from considering x as a quadratic with y fixed, and making c - b^2/4a arbitrarily small as y varies. I eventually came up with (y^2+1)x^2 + 2xy + 1, which turned out to simplify to the form above.
Posted: 11/14/2010 5:50 AM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
criticaluser wrote:
So instead of mashing (right) () (right) () (right) () () (A), It's faster by doing (right/down) () (right/down) () (right/up) () (up/A). (tell me if you don't get it)
I determined this a while back in AW2, but was unable to explain it well. (right/down) () (right/down) (up) (right) (up/A) should be even faster. For 5 spaces, (right) (down) (right) (up) (right) (up) (right) (down) (right) should be the fastest method.
Posted: 11/14/2010 5:47 AM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
1) I don't recall noticing anything wrong with that formula. It's off by 1 if the result of a certain calculation is an integer, but that's not really important enough to be worth mentioning. I can look at it again if you like. 2) I'm probably the wrong person to ask about this. PJ would certainly know more than I do on this subject.
Posted: 11/4/2010 8:07 PM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
In my attempt to generalize to non-convex polygons, I figured that it should be possible to triangulate any polygon, but I was getting nowhere in my attempts to find an algorithm to do the triangulation. At some point, I realized that I could allow the triangles to include some area outside the polygon, and another triangle would just cancel it out. A better possibility I just thought of (and I'm not sure why I didn't think of it earlier): Consider a ray extending from the given point in any direction. The point is inside the polygon if and only if the ray intersects an odd number of edges. If it goes through a vertex, this counts as one intersection, even though two edges are involved.
Posted: 11/4/2010 7:12 PM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
Are the polygons assumed to be convex? If not, then the intersection in part (b) can be a disjoint union of two or more polygons. Solution to part (a) for convex polygons: For each line segment forming the boundary of the polygon, determine which side of the line the given point in on. The point is inside the polygon if and only if it is on the "correct" side of all of these lines. If you allow non-convex and self-intersecting polygons: Consider the triangles A1-A2-A3, A1-A3-A4, A1-A4-A5, and so forth. (If the polygon is convex, this forms a triangulation of the polygon.) A point is inside the polygon if and only if it is inside an odd number of these triangles. I'm not quite sure what to do if it's on the boundary of one or more triangles. (Disclaimer: I am fallible, and thus I could be wrong about any or all statements in this post.)
Posted: 9/13/2010 3:56 AM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
I completed 7 more levels. In addition, DiamondPhoenix made some improvements which have been spliced in. http://dehacked.2y.net/microstorage.php/info/75746856/Puzznic%20TAS.fm2 The next level (4-4) is stupid and will probably be annoying to optimize. At least it's not 4-9.
Posted: 9/10/2010 2:32 AM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
I also discovered how to skip him. If the solid block is far enough to the left, you can jump off it while picking it up, then jump out of the room through the exit (which the game designer forgot to close). For reference, I tend to finish around 27 minutes unassisted (single-segment, without save/death warping).
Posted: 9/8/2010 3:00 AM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
This TAS uses a glitch that was fixed in the English translation. I think it's safe to say that it won't sync with said translation.
Posted: 9/6/2010 7:04 AM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
I got interested in this game again recently. My current WIP is about 12 seconds shorter than the old one despite completing 3 more levels. http://dehacked.2y.net/microstorage.php/info/428628734/Puzznic%20TAS.fm2
Posted: 9/2/2010 6:26 AM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
The running formula involves the speed of the character trying to run and the average speed among all living enemies. All relevant speeds are affected by the speed-up status. 
speed difference | run%
-----------------+------
   -inf to -42   |   2%
    -41 to -12   |   5%
    -11 to -7    |  10%
     -6 to +3    |  30%
     +4 to +18   |  40%
    +19 to +48   |  60%
    +49 to +inf  |  80%
If all enemies are petrified, you will always successfully run. There's also a formula for when enemies run away from you, but I doubt you're interested in that.
Posted: 7/5/2010 12:02 AM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
This was a long time ago, so my memory has a lot of holes, but here goes: Input is polled once every 3 frames or so, I think. Changing characters is done by pressing L or R for two consecutive input frames. Changing which player gets focus is done by pressing both L and R for two consecutive input frames. (At the beginning of that WIP, I didn't even know it was possible to do that.) After Erik runs into something to break it, he is dazed for a few seconds. This can be stopped by interrupting the animation. Ways to do this include getting him hit, or having Olaf drop his shield while Erik is standing on it. All of the individual level shortcuts I know are probably in that SMV.
Posted: 7/4/2010 11:30 PM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
There's already a run of the Genesis version of the first game: http://tasvideos.org/703M.html As for the second game, I started a WIP of it about 3 years ago, but stopped (probably because I didn't want to figure out the swimming physics): http://tasvideos.org/forum/viewtopic.php?t=5145
Posted: 6/23/2010 10:09 AM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
Bisqwit wrote:
Here's my attempt, below!
This does not behave as expected in the test cases GRRY and GRBRY.
Posted: 6/21/2010 4:12 AM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
Just to get this level started, here's Metatron in 84 parts (1:40). I haven't tried optimizing it yet. ?lvl=31&code=g12:2f2;y13:2f3;c9:11f3;r9:12f2;q10:11f3;p10:12f7;q10:13f3;c11:11f3;b11:12f0;q11:13f3;b7:5f2;y7:6f1;c7:7f1;c7:8f1;c7:9f1;c7:10f1;q8:3f5;c8:4f1;p8:5f5;q8:6f1;c8:7f1;g8:8f1;c8:10f0;c9:3f2;r9:5f0;y9:6f1;r9:7f2;q9:8f1;c10:3f2;g10:5f3;c10:6f3;c10:7f3;p10:8f7;i10:10f6;c11:3f2;r11:4f2;q11:5f1;c11:6f1;b11:7f0;q11:8f5;c11:10f0;g12:3f3;c12:4f3;p12:5f7;c12:6f0;g12:8f3;c12:9f3;q12:10f3;c13:3f0;b13:4f0;q13:5f5;c13:6f1;r13:7f2;q13:8f1;c13:9f1;c13:10f2;c14:3f0;g14:5f3;c14:6f3;c14:7f3;p14:8f7;c14:9f0;c14:10f2;c15:3f0;b15:5f2;g15:6f1;b15:7f0;q15:8f5;c15:10f2;q16:3f1;c16:4f1;p16:5f5;q16:6f1;c16:7f1;g16:8f1;c16:10f2;r17:5f0;g17:6f1;c17:7f1;c17:8f1;c17:9f1;c17:10f1;c9:10f0;q10:9f0;
Posted: 6/18/2010 10:24 PM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
As far as I could gather, there are two facts that led to the obsoletion of the non-glitched run: 1. We did not predict the existence of a glitch that skips the entire game (and thus were unable to explicitly mention that we weren't using it). 2. An unrelated Chrono Trigger run was horribly outdated.
Posted: 6/16/2010 7:53 PM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
Gunty wrote:
Rocket Planes in 3:09 using 19 parts: ?lvl=27&code=c12:6f3;p12:7f3;r13:8f1;c12:10f3;p13:7f6;c14:8f3;b11:7f2;b13:6f2;r14:6f1;p14:5f0;b14:4f3;c12:4f3;g12:5f3;c14:9f0;c13:9f0;q12:9f0;q13:5f1;r14:7f3;c12:8f3;
This can be done faster. Specifically, my solution is O(n^2) instead of O(n^3). Rocket Planes in 1:37 using 19 parts: ?lvl=27&code=g12:4f3;c11:9f3;c11:10f3;q11:11f3;r10:8f2;c11:7f3;p11:8f7;p12:6f7;c12:7f0;b13:6f0;c12:5f3;c12:8f2;c14:8f1;r13:9f1;p13:8f6;b13:7f3;q14:6f1;g14:7f1;r11:6f3;
Posted: 5/4/2010 4:46 AM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
The RNG is volatile enough on the field that it's pretty much impossible to avoid encounters via luck manipulation without savestates. It is possible to avoid many encounters via skillful movements, but those techniques have nothing to do with RNG manipulation.
Posted: 4/21/2010 5:11 PM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
If you pick up a normal honeycomb piece for the first time in the Klungo battle, you will not get that particular Bozzeye scene (even if you pick some up later).
Posted: 3/31/2010 6:08 PM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
The following balls stopped moving somehow: (shot numbers exclude suicides) Stage 22, shot 1: 3 ball Stage 46, shot 1: 1 ball Stage 46, shot 3: cue ball Stage 47, shot 1: cue ball Stage 50, shot 1: 1 ball and cue ball (and several honorable mentions) Stage 51, shot 1: cue ball Stage 58, shot 2: cue ball How is this consistent with the law of conservation of energy?
Posted: 3/16/2010 7:18 PM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
The tune-up gives +30 HP and MP, and +20 power, guard, speed, and magic (the "power up by 1" text is most likely a bad translation). The event flag that prevents you from getting another tune-up is set by the first two bytes of the text after the tune-up. It turns out that wall collisions aren't checked between the text before the tune-up and after it, but you can still move. If you walk as fast as possible to the stairs, the text will load when you reach the next room. However, if you manage to make the room fade out while the window is opening (but before it's completely open), you can skip the text entirely, including the part which sets that event flag. After that, you can just walk downstairs and get a tune-up again, since the game doesn't think you've done that yet.
Posted: 2/21/2010 12:02 AM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
http://dehacked.2y.net/microstorage.php/info/1870157010/Gariso%20Defeated.smv With this, we get the seventh and final rune. This is the stated goal of the game, so we're obviously done, right?
Posted: 2/3/2010 6:46 PM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
Try looking up "linear feedback shift register" instead. The tap sequence is [47, 21], and the most significant bit is generated first.
Posted: 12/2/2009 7:47 AM
Quote
Nitrodon
Experienced Forum User, Published Author, Active player (283)
Joined: 3/4/2006
Posts: 341
Might as well reveal some other stuff I found while hacking this game last year. I can't be bothered to format anything, so be warned. base damage = strength*(level+7)/8 + weapon power - defense*(enemy level+11)/12 minimum base damage = weapon power/32 rushing: 3/4 damage spinner: 5/4 damage slicer: 6/4 damage slider: 8/4 damage elemental properties (when applicable): 25%/50%/100%/150%/200% damage critical hit rate: (strength + 8 - target's strength)/128, minimum 4/128. Critical hits deal double damage. Attack damage varies based on certain bits in the timer (0042): bit 3 (0x008): do not adjust damage if set (ignore other bits) bit 2 (0x004): adjust by 1/4 if set, 1/8 if clear bit 8 (0x100): subtract damage if set, add if clear
1 2 3
13 14