Posts for Bisqwit


Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
adelikat wrote:
Couldn't the method used in these .mpg's be used in other areas as well?
Possibly. Have any idea where? (Both of these techniques have already been used in the existing movie in different places.)
Post subject: New beam usage plan
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
Previously, 4 beams were used in Wily2 and 2 in Wily3. Now, 2 beams are used in Wily2 and 1 in Wily3. In other words, 3 beams less are used. I need 7 beams in the beginning of Wily4. Therefore, I need 10 beams after the refills in Wily2. In the beginning of Wily2, I have 1. If I take two refills, I have 11. If I omit one small refill in Wily1, I'll have the exact amount after the two refills (I need to remanipulate the luck though). I could also use one extra beam in Wily4... I should compare whether the saving of extra beam in Wily4 is larger than the cost of 1 small refill in Wily1.
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
finalfighter wrote:
Morimoto find the new way to Wily2 & Wily3. It is fast.
Cool, I can get back to work at this movie :P
Post subject: September 2005
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
In September 2005 - Globally on my webserver, there were 1761250 hits, 13.6 GB of traffic, with average of 2446 hits per hour. 865k hits in the name of MSIE 6, 750k in the name of Mozilla, and 60k hits in the name of Opera. Konqueror has 3k, and Googlebot (heh) has 31k. The site was down at days 11-14. In September, Webalizer can't specify an audience favourite movie, but the most active referrer was, rather surprisingly, http://www.collegehumor.com/.
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
NrgSpoon wrote:
Or maybe it's the thing involving hitting Bowser and the axe at the same time, so mushrooms make you visibly shrink instead.
That would be related to the fiery Mario glitch.
http://tasvideos.org/SuperMarioBrosTricks.html wrote:
Note 2: The small-fiery-Mario trick is not explained on this page, because it's not really useful in making speedruns. It's however explained in detail at Wikipedia.
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
This name does not tell me much of anything. Please provide some details.
Post subject: Source code
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
http://bisqwit.iki.fi/src/fceu-bisqbot-patch-v2.txt (patch of FCEU source code) BisqBot code updated. It's now separated into a separate file, bisqbot.cc, which contains the settings that can be changed per-game and per-task. It still requires the recompilation of the source code every time the subjective is changed, though.
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
Thanks for this tip. I don't know how many people actually know this trick, but I have already been aware of it for more than 10 years -- I read it from a Nintendo magazine when the game was still new. Naturally, this is of no use in speedruns :)
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
Ah, yes. This trick was known already by then, but what you claimed (at least, how I remember it) was that the trick can be done in abritrary positions in the ladder. So far, we can only do it at the bottom of it (and at the too-high-position, that is, near the upper surface of it). Your claim is still unproven. After analyzing the code this carefully, I don't think it will be proven, either.
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
AngerFist wrote:
Remember when I wrote that I experienced this strange trick/glitch? I guess you proved me right with that .gif?
I have no idea what you're talking about, and that particular GIF has been around since April. :)
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
BlueBlue wrote:
But, as I said before, I think the pertinent part for a speedrun is that by jumping on the first frame, you can bypass the 7 frames it takes to start running.
Naturally. This was used in all Rockman 1-6 movies so far except the first Rockman 1 movie (Morimoto's). This is rather basic knowledge.
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
Zakatos wrote:
what could be the explanation of why you earn 1000 bonus orbs at the end of each boss fight?
Umm, no I don't. It's saying that each bonus orb is 1000 points, but I didn't get any. 1000 × 00. Thank you for the feedback! > Also, one question: how do you turn into teleporting-form without bringing the weapon select screen? By pressing the Select button.
Post subject: ladder functions partially disassembled
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
I analyzed the Rockman code a bit, and I found out why the "grabbing ladder top too high" trick works in the way it does. When the game thinks that Rockman has cleared the ladder (by climbing up) (that is, he's climbing up but the tile his upper portion lies on is not a ladder), it resets rockman's position with this formula: Y = (Y & 0xF0) + 0x0C The idea of that formula is to ensure that when he clears the ladder, he's properly positioned right on top of the ladder, and will not start falling. But if his climbing position was a little too high, that formula will actually place him 16 pixels above the ladder, as demonstrated in this image: https://files.tasvideos.org/gameresources/nes/rockman/updowntrick2.gif The same code is also probably responsible of this effect: https://files.tasvideos.org/gameresources/nes/rockman/vertgrab.gif I also found out, how climbing down works. When Rockman is trying to climb down, the game first checks whether Rockman's body is on a ladder. If that's not the case, it will drop Rockman by 12 pixels. Then it will actually climb him. (Wow, that sounds quite obvious and bland when I think of it.) So why does pressing "up+down" over a ladder work like it does? At first, I had no idea. According to the code, the game should ignore "down" if "up" is pressed. But we already know that's not true. Then I found that there's more to it. The game completely ignores up/down buttons if there is no ladder anywhere near Rockman. By ignoring, I mean it won't even check if they're pressed. Nice. When there is a ladder in sight, it checks for two things: * If Up (but not Down) is pressed, but the ladder in sight is below Rockman, the key is ignored. * If Down (but not Up) is pressed, but the ladder in sight covers whole Rockman (both the upper and lower part of his torso) but not below him, the key is ignored. These tests are rather dubious. Here's where up+down steps in: When you hold up+down, neither of those tests match, so it will go straight into Ladder_Handler, assuming that Rockman is now climbing. After that, it will hold priority to the Up key (Down is ignored if Up is pressed), and if Up was pressed and there's no ladder behind Rockman's upper torso, he'll perform an exit from the ladder, finally explaining the up+down trick. According to this analysis, the "down" key is also ignored when Rockman is walking past a ladder, but not if the ladder is shorter than 2 blocks. Edit: Analysis proven correct. See http://bisqwit.iki.fi/kala/tinyladder.zip (Cutman level beginning). Press or hold down when you pass the ladder. This quirk is responsible for this behavior (note closely what happens when Rockman passes below the ladder): I can't think of ways to abuse the "drop Rockman by 12 pixels" feature. To do it, you would need to find a way to put Rockman so that the game thinks he's now located near a ladder (but not over a ladder), but so that when he drops by 12 pixels, he's still not over the ladder.
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
It's a strange finding. I would have expected it to be more straightforward. I found this code, which is somehow related to Megaman's movement, but I'm not yet sure whether it's applies to A)climbing B)walking C)jumping. $94E7> A5 94: LDA MegamanWalkTimer $94E9> F0 0E: BEQ + ; $94F9 $94EB> A5 23: LDA FrameCounter $94ED> 29 03: AND #$03 $94EF> D0 08: BNE + ; $94F9 $94F1> A9 01: LDA #$01 $94F3> 85 95: STA MegamanStallTimer $94F5> EE A006: INC $06A0 $94F8> 60: RTS + $94F9> A9 00: LDA #$00 $94FB> 85 95: STA MegamanStallTimer $94FD> 85 94: STA MegamanWalkTimer Ps: This all applies to Rockman 1. This is a Mega Man 2 thread. What should we do about it? Edit: Oh, I'll know. I'll move'em to the right thread.
Post subject: Re: warn: image heavy
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
Omega wrote:
Maybe you can make a custom map that's just one long corridor, Bisqwit?
Here you go! http://bisqwit.iki.fi/kala/megacorridor.zip
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
Omega wrote:
Actually, I remember Bisqwit saying that the speed in Rockman was 1.33 pixels per frame, meaning he runs at 1, 1, 2 pixels per frame.
I may have mistaked in many ways. Better you verify it. :)
Post subject: Robot-assisted speedruns
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
As many people here know, I used a computer program to help me in some parts of my Rockman Movie. I programmed "BisqBot" to attempt to find the fastest ways to play some sequences of the game. It works best when the sequences are short (30 to 90 frames long) and the user has at least some understanding of what kind of input is required. This kind of approach has plenty of uses for speedrun purposes. Such robot can be used to invoke "once-in-a-lifetime" occurances in many games, for example the mountain-warp trick in Castlevania 2. Therefore I'm trying to formalize my code to make it general-purpose. Here's my current code. It's not an existing programming language, and there is no interpreter written for it yet. It's merely a plan. However, these rules imply precisely what kind of settings for BisqBot I used at some point in my run.
/* GLOBAL SETTING */
MAX_FRAMES 120

--------------------------
/* GAME MONITORING */
/* A, X, Y, PC correspond to the values of the CPU registers. */

// D57D = FrameUpd
// C5AB = RandomFunc
// BF25 = CreateDrop
// A231 = HurtFunc
// C833 = GotItem
// C219 = KillFunc
// BEEC = EnemyKill
WHEN PC = 0xBF25
    DEFINE RockmanBonus = Y  /* Y identifies the item that was generated */
WHEN PC = 0xA231
    DEFINE RockmanHurt = 1
WHEN PC = 0xC219
    DEFINE RockmanHurt = 2
WHEN PC = 0xC833
    DEFINE RockmanItem = A  /* A identifies the item that was collected */
WHEN PC = 0xBEEC
    DEFINE RockmanKill = A /* A identifies the enemy that was destroyed */

/* Bonus items in Rockman:
 * 0x3B = no item
 * 0x3c = score
 * 0x3D, 0x3E =small refills (wpn,nrj)
 * 0x3F, 0x40 = big refills (wpn,nrj)
 * 0x41 = 1up
 */

--------------------------

/* How to restart */
CLEAR RockmanBonus
CLEAR RockmanHurt
CLEAR RockmanItem
CLEAR RockmanKill

--------------------------

/* RULES, HOW TO CONTINUE AT EACH FRAME */

WHEN RockmanHurt COUNT_AT_LEAST 1
  OR RockmanItem COUNT_AT_LEAST 1
  OR RockmanBonus IS_NOT 0x3F,0x3B
    RESTART
WHEN RockmanBonus HAS 0x3F COUNT_AT_LEAST 2
  /* When two refills have been found, accept this result */
  ACCEPT
  /* Set this as a new cap (don't try to find longer sequences) */
  SET_LIMIT
  /* And go find more solutions. */
  RESTART

--------------------------

/* HOW TO CREATE INPUT */
/* These hex numbers are bitwise ORs of controller input
 * bits. 0x01 = A, 0x40 = left, 0x80 = right, 0x02 = B
 */
INPUT_TABLE 0x00 0x01
            0x40 0x41
            0x80 0x81
            0x01 0x41 0x81
            0x01 0x41 0x81
            0x01 0x41 0x81

// initialize input with zero.
INPUT 0x00
// 1:20 chance that B will be held this frame
IF_RANDOM 20 INPUT_OR 0x02
// 1:20 chance that LastInput will be changed
IF_RANDOM 20 RANDOM_ASSIGN LastInput INPUT_TABLE
// Include the LastInput in this frame
INPUT_OR LastInput
The purpose of this post is to call programming-oriented people to further refine this plan and to make it actually useful.
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
In my Rockman movie, it was useful for the fire shield not to last long ― or rather, it wouldn't have been useful for it to last longer. The reason is that I needed usually the fire spot to be in particular angle at a certain frame, and to accomplish that, it was very important to shoot the fire certain number of frames earlier.
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
CtrlAltDestroy wrote:
Is it just my eyes, or is there a missed shot on Bomb Man's stage (Frame 24152) ?
You're likely right. Where ever I had to manipulate luck for the weapon refills (well, BisqBot had to), excess shots were often fired. The reason to this is that the random seed is chaotically affected by the number of objects on screen at each frame, and shooting at the right frame might just create the right desirable result. (I.e. it's not only the frame when enemy dies, that matters.)
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
nitsuja wrote:
Many movies refer to other movies in their description without providing a link or even the exact name of the other movie. For instance, the full Zelda: A Link to the Past run refers to the "short version" of the run, and I can imagine some people reading that description wondering what the short version is and not knowing where to find it. Maybe it would helpful for all the references to movies to get links, except when it's talking about the movie that it directly obsoleted.
That a good idea, but the problem with movie pages is that movies.cgi is basically a search engine. It takes different types of search options, and the id= is just one of them. There are no pages dedicated to single movies, even though it's a common practice to link to them with the id= setting. I'm open to suggestions to help this.
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
Warp wrote:
Anyone else think this is so cool matrix-style running that it could deserve a star?
Many seem to think so. Thus, it will happen.
Post subject: "BisqBot" source code
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
Boco wrote:
The full code is at the discussion topic in the NES forum.
The patch is here: http://bisqwit.iki.fi/src/fceu-bisqbot-patch.txt you can ignore the TILETRACKER part, because that one's related to GIF making. The patch is to be applied to the Linux version of FCEU (after applying Blip's movie support patch). It should contain only the bare relevant parts of this patch, with the exception to my NES pad reading code, the compose mode and minor input remappings.
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
RT-55J wrote:
H. M. D. must be crying himself to sleep right now.
Nah, as a programmer myself, I can say that if I were in his place, I would be quite amused. These are (mostly) not failures at programming, just oversights or intentional compromises, that I am abusing in this movie. Even if there are actually failureous logic errors, it wouldn't fit in the programmer's mindscape to be insulted or hurt from seeing someone using creativity against his own creation. So I'm fairly confident, that if he ever sees this movie, it will be fascinating, and amusing to him. (I would be glad to hear it from him though.)
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
I definitely did.
Post subject: Re: Megaman 1 glitch
Editor, Experienced Forum User, Published Author, Active player (296)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
I haven't heard, but also not studied, of the duration of the fire shield being different in different circumstances (except lag or pause).