Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
As I remember, on real console the game was damn impossible to survive. But it's still enjoyable and even kinda fast-paced if not for those ghost rooms and slow spikeshafts. I guess these stops were inevitable, although you could try to perform some antics during waiting periods.
Okay, you sure polished this run pretty well, so it's a Yes vote. Too bad you coulnd't afford using more damageboosts, they looked like real opportunities to speed up the game.
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
Just thinking about working on some new game (instead of improving one of published TASes) makes me feel VERY uncomfortable. I catch myself on not even trying to distinguish pros and cons, because even strong motivation quickly fades away from pressure of huge risk to waste time for non-publishable material.
I don't remember having such feeling when I only began TASing. So it must be something I gathered from these 2 years of TASing experience and communication. Something that was born from judging policy, voting/rating trends and common discussions/feedbacks.
Now I know for sure: making new runs is far more difficult than improving other. Of course, polishing is tedious, but at least it's ...safe, yeah, this word suits it. Starting new run requires a lot of encouragement, which currently cannot be found anywhere on tasvideos.
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
BenniCR wrote:
Also, what about the flying thing? a bit of explaining would make this publication perfect.
Do you mean that flying enemy from level 3-1 which accompamies players for most of the level? Well, I managed to prevent it from causing any lag, so why not. :)
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
"Need nothing else than just your pride to get there" ;)
Last year, when watching "God playing" video on YouTube, I was wondering why they included such sloppy moment as climbing brick stairs in SMB3. So I made an assumption it's the best way to cross the level.
Now what, another ground-shaking TAS! Congrats, guys. If I'm not mistaken, nearly all runs presented in the promo video now are obsolete.
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
Before sticking to TAS I had several different hobbies, and now all of them kinda stagnate. And it's not that I'm anywhere near productive on the field of TASing! Ehhh, what the heck...
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
Horray for new world record!
According to doppelganger's SMB disassembly the trick works because Mario Y coordinate > 0xA2 at the moment of touching the flagpole.
lda Enemy_Y_Position,x ;check flagpole flag's vertical coordinate
cmp #$aa ;if flagpole flag down to a certain point,
bcs GiveFPScr ;branch to end the level
lda Player_Y_Position ;check player's vertical coordinate
cmp #$a2 ;if player down to a certain point,
bcs GiveFPScr ;branch to end the level
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
Nice improvement. As I see, published run was very challenging to beat, although it's quite old.
Despite that I find this TAS a little bit less entertaining than Genisto's run, because some levels were played with "chip-carries-dale" style. Of course I understand it was necessary to reduce lag by halving amount of collision detection.
Anyway, that glitching on Zone F was truly amazing, and I'll vote Yes just because of that.
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
Excellent improvement in both speed and entertainment. Dancing on stage 1-2 was so perfectly timed that I hoped you'd repeat it later. Final level was also entertaining as never before, the most hilarious trick was near frame 29150. Voting YeS!
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
Baxter wrote:
Sorry for going completely offtopic:
Are you currently working on some TAS AnS?
And if so, which one?
Currently I'm trying to finish old Werewolf improvement, but it goes slowly and uninteresting. Also I have some plans to polish my first published movie (Little Mermaid), but not this year. So, nothing really special, I guess.
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
I don't think that longer version would obsolete published run if it's arguably entertaining (if you meant that 2players version is gonna be more interesting to watch). Anyway, while I'm speaking here, you probably recorded most of that TAS, so why not. :)
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
Great that you decided to improve this run, Dasrik! Will you insert some old input dances? ;)
I've found these adresses that Baxter mentioned, but there are some details necessary to explain.
You may know that the game has 8 slots for enemy objects. Thus there are 8 memory addresses for life (x621-x628), 8 X and 8 Y coordinates and so on. Also, x529-x530 adresses contain 8 enemy identificators (0 = free slot, 5 = foot soldier1 here, ...)
So, naturally there are 8 invulnerability timers stored at x689-x690. They work in somewhat unusual way. Actually there are two types of enemies - ones that have no invulnerability time (you can hurt them right when they appear) and ones that have standard_invulnerability_period. And this period depends on a slot where invulnerable object appears.
For example, if an enemy appears on first slot (slot0, which is unlikely), it will have whole 224 frames of invulnerability (x689=32*(7-0)=224). If it appears on slot 3 then it'll have 96 frames of invulnerability (x68C=96). Fortunately, enemies usually try to appear from last slot (7), but if it's already used, they appear in next-to-last slot, then try other slot and so on. When there's only three enemies, they take slots 7,6,5 and their invulnerability timers are 0, 32, 64.
So it seems you can't manipulate these timers - the only thing you can do is to make enemy from slot7 (with timer=0) appear closer to you and then wait 32 frames for 2nd enemy / 64 for 3rd.
And of course you'll want new enemies to appear on slot7, so try to clear this slot before new enemy decides to appear (slot7 is cleared when x530=0, see above). Well, I'm speaking only about enemies with invulnerability timers, as for those without invulnerability, you can let them take any slot - they still will have timer=0.
Ok, hope this helps. :)
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
While still voting Yes, after watching the run three times I have to admit it definitely can be "improved" in terms of entertainment (or better to say "polished by authors"). Everything was mentioned in the topic, like suggestions to reduce final battle glitching, make more hilarious stacking and maybe add something new. And sure this movie shouldn't obsolete time-attack run. I don't see anything bad in adding second movie for such famous game.
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
Ow, how could I forget such awesome game I enjoyed a lot as a kid! Well, good thing that you made the TAS look so smooth and varying. Also I agree No-damage aspect is wise decision here. So, it's total yes vote!
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
Whoa, 18seconds improvement is unexpectedly high result for NG.
However, I think the run is still improvable. For example, in 4-1 you lose ~7 frames on first moving platform (2nd platform could appear earlier if you wouldn't stop for ninpo change). Do you really need that fire weapon later in the level?
Anyway, yes vote for entertaining and fast movie.
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
Haha, cool stuff! The movie really shows the game's best. Fastest time, maximum score/carrots, awesome antics here and there - I wonder what aspect of the game can be improved anymore.
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
Kitsune wrote:
Unfortunately, this can't really be called a no damage TAS, since you use Death as a shortcut, you obviously take damage.
Thus defeating the cause. I am voting yes however.
"Death as a shortcut" isn't equal to "Takes damage". These are completely different categories, especially in the case.
In Bucky O'Hare the moment works this way: when playable character collides with enemy, first the programm verifies if x3B is set, and if it is, programm kills you immediately - even if you're in invulnerable state or whatever. Then another useful subroutine nullifies LIFE gauge and makes other stuff like changing sprites and so on.
But if that byte isn't set, the programm applies to another branch of code - to check if the character's invulnerability timer is > 0, then it checks how strong is that enemy and after that it decrements LIFE gauge by respective value. If this process results in LIFE energy becoming less or equal to 0, then above mentioned another subroutine will be called to make some workaround.
So technically it's still no-damage run. And damage-taking will change a lot of tactics in the game.
Emulator Coder, Experienced Forum User, Published Author, Experienced player
(724)
Joined: 2/23/2006
Posts: 682
Moozooh explained my assumptions perfectly. :)
Personally, I'd like to see two Bucky O'Hare runs: "anything-goes" and "entertainment/longplay". My current run is something inbetween...
1. "Any%" run - may be done in approx. 21:20+
The run should find theoretical limit of the game.
Aims for fastest time
Takes damage to save time
Uses death as shortcut
Uses password (to skip intro movie)
Abuses programm errors
Manipulates luck
2. "Full" run - can be done in 23:43
The run plays all levels and tries to show viewers the game's best.
Aims for fastest time BUT
Contains speed/entertainment tradeoffs
Takes no damage
No warps
[optionally] Uses password / Plays at hardest level
Abuses programm errors
Manipulates luck
Probably next year I may be thinking about making full run.
As for "any%" run, I've found some arguable timesaver. There's interesting password "KNJ7J" which doesn't contain ANYTHING in its bits. After entering it, you'll appear on STAGE SELECT screen and all four planets are untouched, you have only Bucky with initial level of power/life. Just like if you chose GAME START on title screen.
So, IMO, in terms of fair play this password can be considered as playing the game normally from beginning. Am I right with the opinion?
The password only skips intro movie (about 13 seconds) which could be useful for "Any%" run, but not for my current TAS.
Xipo wrote:
It looks very good.
I like your show in green planet,ice planet and stage Center of magma tanker.
Great run!
I think if someone vote no,he should tell us the reason instead of be silent.
PS:Good luck!
BTW, xipo! I'll be waiting for your run of the game. ;) Hope my descriptions would help you too.
