This topic is for the purpose of discussing #1893: Inzult's NES Zelda II: The Adventure of Link "glitched" in 05:43.47

Glitchy goodness that's 30 seconds faster = very yes! The approach to and from the hammer seemed a little bizarre, being slower than max speed and everything. But it appears to have worked well, so... Good job!

Notice that NESVideoAgent renamed the ROM listing. This is part of the continuing issue that the video file doesn't store full CRC codes to properly indicate which version of the ROM it is.

FCM indicates all the required information. The ROM selected by NVA was an exact match for the MD5 indicated in your movie file header. Only VBM and GMV don't.

Also, I noticed that this run has the normal (U) [!] rom but the last run had a (U) [b4] rom. Dosen't that mean it is a bad dump?

I just use the rom I have. As far as I know there has never been any problem between it and the good dump. It probably just has a dirty header or something.

I don't have a clue what Link did, but apparently it managed to save the world or a princess or something. yes-vote to an obvious improvement.

So how exactly does the strange warping work?

So how exactly does the strange warping work?

If you're looking for a technical explanation, I have no idea. From observation, Link is drawn in the wrong spot, and takes exits that don't normally exist. If you pause the game when Link first spawns, the menu is actually half off the screen on the left. The game handles the false exits by putting Link in unexpected places.

So how exactly does the strange warping work?

If you're looking for a technical explanation, I have no idea. From observation, Link is drawn in the wrong spot, and takes exits that don't normally exist. If you pause the game when Link first spawns, the menu is actually half off the screen on the left. The game handles the false exits by putting Link in unexpected places.

Probably how the rooms/screens are stored in the rom, I imagine it just detects the exit wrong from going so fast, and places you in the room/screen left/right of it instead of the world map. Oh yeah, yes vote by the way.

I made the Zelda 2 editor, so I'm sort of familiar with how the exits normally work. But this is more about how the exits work in wierded-out extreme cases. The first abnormal exit taken is in Rauru, which leads to Room #0 in the town set of levels (also known as "Freakytown"). Same place you end up if you use Fairy off the top of the screen. For some reason, all the towns maps have exits on top and bottom mapping to room #0 in the town set of levels, but I have no clue why the door led there. Maps #3 and #4 are also freakytown... Then you enter that house once, and completely screw up the game. All bets are off regarding how the game works.

Ha ha! Well, that was certainly...fast. Although part of me feels there could be some super-funky way of getting to the end nearly instantly. In fact, I'd wager that doing something funky with the Japanese double-sided FDS version could yield interesting results. Regardless, this is awesome and nets an obvious yes vote.

Still one of the most mind blowing runs on the site. I love the dark link fight by the way. Also-- If the warping is so poorly understood, how was the current route decided on? was it strictly trial and error, or was enough of a pattern known to decide on a reasonably coherent path?

I made the Zelda 2 editor, so I'm sort of familiar with how the exits normally work.

Oh snap! You're that Dwedit? I had fun with that program. Hella fun. Yes, hella. Otherwise, yeah, I don't know why the door leads to freakytown or why entering the building there a certain number of times breaks the game this way. Entering more times puts you into a town with fall-through ground and no right (I think, memory's a bit fuzzy) exit. Of course, I'm really curious what exactly is happening to the game. I do not have the skills to find out, though.

If the warping is so poorly understood, how was the current route decided on? was it strictly trial and error, or was enough of a pattern known to decide on a reasonably coherent path?

Short answer: trial and error. Long answer: Now that I have tried and errored, I find the way it works somewhat predictable. If you exit to the right, it is (usually) the normal right exit. If you exit to the left without scrolling the screen, it is (usually) the normal left exit. When you scroll the screen a bit, and exit to the left, that's where the fun happens. This includes when the screen is already scrolled for you (spawning anywhere other than the left of a room). Using this, I worked out the route through palace 2 in advance, but there was a chance something I didn't expect would have happened. A lot of the work on the route was done by other TASers also. Sometimes it's sort of easy to tell where you will go. For rooms with an elevator, you usually go to the room that the elevator links to. For rooms in sequence, you will usually go one screen to the right. For rooms with only one exit, that's when the destination is less predictable. Occasionally you will wind up at the enterance of the palace. Occasionally you will wind up in the middle of nowhere. Using fairy on the top of a screen links you to similar places as the glitched exits. Of course, sometimes you spawn inside bricks or lava, so in general, every exit needs to be tested. That said, I have obviously not tested every exit in the game because I am generally not that bored. Trying to go into the basement with the magic jar in Hidden Kasuto will warp you to old Kasuto. Weird suff like that is probably impossible to predict without seeing (or knowing exactly what's happening).

I want to know if the room that you go to when you move left is one of the four rooms appearing in the layout list. If you run Zelda 2 Editor (use dosbox), you will see an area in the bottom-right labelled "Layout". First of these is the room numbers. The number in the middle is the room you are at, the numbers in each 4 directions are the destination room numbers. There's also information to the right about what side of the room you enter the next area at. (0-3, multiply by 256 basically) Room 3F is an exit, but is modified by the direction specified (which determines where you exit a 2-sided cave)

Would someone kindly guide me to the Zelda 2 thread? if you enter the house in freakytown with 0x75C = 02, then you get the freaky stuff happening after that. Any other value and it doesn't get freaky.

Okie doke. In Palace 2, You start in room 0E, with the down exit pointing to 13. That's where you go when you use the warp. In the glove room, (room 14), the top and bottom exits point to room 00. That seems to be the first room of the first palace. If you stop the run when that warp is used and enter the palace via the elevator, you do indeed enter palace #1 (with palace 2's colors) In the room with the jar under the rock (map #59), all exits point to 3F. This is what warps Link to East Hyrule. For some reason, this acts like it's the exit of another map. Just wandering around inside palace 1, it looks like it takes whatever is mapped to the bottom exit. e: the thread is over this a way. I'll edit the submission text once you get whatever it is figured out.

What happens is that the townsperson counts the steps she is taking to find out whether she is to the left or the right of the door. When you talk to her at the right moment, she will lose count and think she's to the right, and will enter the door in the wrong direction. She is supposed to stop you on the way in, but doesn't since she isn't where you are. You'll go to room 0. Then you go to a house within a house within a house. The game tries to remember where you came from, but overwrites the data for normal screen transitions. The second time you enter the door, entering from the left side becomes glitched and the fifth time, entering from the right side becomes glitched. The left initial screen becomes the new left column of the initial screen The left column of the initial screen becomes the new left screen The left screen becomes the new right screen The right screen becomes the new left column The left column becomes the new right column The right column becomes the new screen number The screen number becomes the new left side offset The left side offset becomes the new scrolling offset The scrolling offset becomes the new scrolling direction Screen direction 4th time becomes left bound for 3 screen rooms Screen direction 5th time becomes left bound for 4 screen rooms Screen direction 7th time becomes right bound for 3 screen rooms Screen direction 8th time becomes right bound for 4 screen rooms Because the left bound has changed, the screen doesn't scroll fully to the left when you try to go back. You're supposed to be to the left of the room, but you aren't so you'll be regarded as being 1 screen to the right and will go to the room below. (By the way I think you mean 075B, not 075C. 075B is the number of houses you're currently inside. 075C is the direction you enter from, where 0=Left, 1=Left middle, 2=Right middle, 3=Right, 4=Elevator, 5=House 1, 6=House 2)

In the submission text, you mention all these frame delays and frame savings, which gives one the impression that overall not a lot of frames were saved from a previously tightly optimized movie, yet this was 30 seconds faster. Could you elaborate a little more on how this movie is half a minute faster? Liked the old one, love the new one, yes from me.

This movie has been published. The posts before this message apply to the submission, and posts after this message apply to the published movie. ---- [1082] NES Zelda II: The Adventure of Link "warp glitch" by Inzult in 05:43.47

In the submission text, you mention all these frame delays and frame savings, which gives one the impression that overall not a lot of frames were saved from a previously tightly optimized movie, yet this was 30 seconds faster. Could you elaborate a little more on how this movie is half a minute faster?

Sure. In the old run, after the Jump spell is gotten, Link walks through the jump cave to level 2, playing it without the warping glitch active, then uses up&a (game restart trick) and travels to Saria via the warping glitch. Measured from getting Jump to entering Saria, this takes 9470 frames. In the new run, up&a is used after getting the Jump spell, then Link travels to Palace 2 via the warping glitch and walks back across the swamp to Saria. Measured from getting Jump to entering Saria, this takes 7830 frames. 1640 frames saved thanks the the better route choice. I overestimated the savings thanks to walking around to the north of the warp point to east Hyrule, it's only 109 frames. I should probably update the submission text.

The description says he doesn't kill a monster, but doesn't he kill a monster in the very first cave? Hit once on the way in, once on the way out... gets 10 exp.

theres a warp around the north cave? O_O I always thought you had to use the raft to get across ->. i havent watched the movie yet so it is probably a glitched warp.

Good job. I'm happy to see the progress of this run. Thanks

You don't need the raft if you can get from Death Mountain to Maze Island. Maze Island and Death Mountain are basically the same area. If you go east of Death Mountain (such as with a walk-thru-walls code), you come to Maze Island. So if there's a glitchwarp from Death Mountain to Maze Island, that will get you to East Hyrule without a raft. There are actually two copies of the Maze Island/Death Moutain, one for East Hyrule's enemy set, and one for West Hyrule's enemy set. (Nintendo was wasteful here and didn't share the data, instead they completely duplicated it). So when you warp from Death Mountain to Maze Island, you get West Hyrule's enemies there.