Posts for asteron


1 2
10 11 12
18 19
Experienced Forum User
Joined: 2/16/2005
Posts: 462
Its weird how it always goes straight.... I wonder if there is ever a place where you would like to cut across a steep slope.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
Well it will have to be saved for the 16-star run... Hopefully there will be an oppurtunity to show it off somewhere.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
I agree with all the comments on the awesomeness of THI :) Nice btw back in WDW did you ever check to see if the chuckya glitch was slower than "quick race through downtown" (the one in the city you need either vanish cap or chuckya)? I guess it has to be but its a shame that it can't be showed off in TAS perfection.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
The 70 star run has its flaws though, a suboptimal route selection and pretty crappy bowser fights. The 120 star route is much easier to come up with.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
Does the gender of someone on the internet whom you will never meet make any sort of difference? Maybe I'm the only one who tends to think of everyone I see online as sexless avatars.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
Just finished watching the speedrun. It is a work of art for the bowser levels alone! I wouldnt say it is up to par with the 70 star run but it is pretty close and uses more modern strategies. You can tell from the camera manipulation how much practice was involved.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
In "watch for falling rocks" (the one with the jump kick to the ledge), isnt it faster to slide down the slope and jump over the middle pit as opposed to going around? I saw a video of that around here somewhere and it looked pretty fast. The first zipping looked like it could have gone farther but I trust from your statement you investigated that. I really liked the metal cap star :)
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
Congratulations on Corneria! Maybe 4 more years from now and you can start the expert run :)
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
Ahh thanks!
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
I honestly looked through the faq and the graphics title "toppub.png" isnt very clear so what does mean? Is it the best video on the front page? I suggest that this graphic (and the star one) link to a legend with descriptions.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
Yeah I guess it is kind of a grey area...
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
There isnt really any teleportation in this game, you can however get very fast with a set of stairs, an upwards moving platform, or a sufficiently steep slope. You can also squeeze through some thin walls/ceilings. These techiniques are already used in the 16-star run.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
I really should stop but its so hard!!! Yeah integrals are continuous and reimann sums are not but they are really pretty much the same thing. Integrals are defined as the limit of a summation and the reason an integral symbol looks like an S is because it stands for the word 'summation' and can be thought of as such. The summation symbol is a Sigma which is just a greek S. As you noted there is a topological difference between the two operators in the spaces over which they work but it is still pretty much the same operator. There are tons of other operators out there (like the addition operator +) that we dont really bother defining seperately over different spaces like real numbers and integers. But yeah I am probably more likely to think of reimann sums as discrete integrals than think of integrals as continuous reimann sums.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
Well anyway good came out of this. You can BLJ up steepish slopes that you can stand on without sliding.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
Im probably just speaking too much math/physics for my own good. Please disregard this if its boring. Well for every frame you have position p, velocity v, and acceleration a. If frame n is dt long then for frame n+1 we have p[n+1] = p[n] + v[n]*dt v[n+1] = v[n] + a[n]*dt ... extending this to k frames later p[n+k] = p[n] + v[n]*dt + v[n+1]*dt + v[n+2]*dt + .... + v[n+k-1]*dt v[n+k] = v[n] + a[n]*dt + a[n+1]*dt + a[n+2]*dt + .... + a[n+k-1]*dt factoring p[n+k] = p[n] + sum(v[n] ... v[n+k-1])*dt v[n+k] = v[n] + sum(a[n] ... a[n+k-1])*dt anyway the term sum(a[n] ... a[n+k-1])*dt is called an integral. sum(a[n] ... a[n+k-1])/k is the net accelleration (or average acceleration) As you can see the only way for the velocity to be constant is for that to be zero. If a[n] starts positive then it must later be negative for the velocity v[n+k] to equal v[n]. I am just calling the negative acceleration friction. I think I am spending too much time talking about this... sorry... please ignore me.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
???? Velocity = Integral Acceleration dt Are you thinking like net acceleration or something? As in it starts positive and ends up negative and so the net is zero? I would call the negative part the friction. Air friction :) Really i think we are just using different terminology.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
1. If the velocity doesnt decrease what keeps the non-zero acceleration from building up? 2. Air friction
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
I call the decreasing velocity you described friction.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
FODA wrote:
It doesn't get any more complicated than what i already said. Still, i'll talk about what you said, because there are some imprecisions. What friction are you talking about? ground friction? BLJ can be done by touching the ground for a single frame time, and this doesn't change whether you jump high or low. The amount of ground friction is the same for every jump. This is not the reason for the accumulated speed gain.
No friction is not the reason of the boost but is always present. It explains why you cant gain speed by BLJ on flat terrain. The speed boost though overcomes it when you are doing hyper BLJs.
FODA wrote:
Yes, that's what i said, when you jump to a higher plataform, you land on it.
Sorry for repeating you, it took me a while to write that post and didnt see yours.
FODA wrote:
It's not possible to jump on every stair... there's no limit to the speed you can acquire with BLJ.
I meant to use 'max' as a simple upper limit.
FODA wrote:
What you're basically saying is that if i miss landing on an upper plataform i can try and land on the next BLJ? I don't see any place where this could be used, because there aren't stairs in front of slopes or anything like that. It's either get it right or try again.
I was thinking along the lines of angling diagonally up stairs so you can store speed in a blocked direction. The point was that at the beginning you are not necessarily jumping at each frame.
FODA wrote:
here you contradict what you said earlier, yep, each jump may send you through several stairs.
Again I was thinking of an upper bound without necessarily considering aliasing effects
FODA wrote:
The way to do this is real time for me, is press "A" as fast as possible. There's not much science about it, just get the angle straight and then tap A. You may miss some of the first jumps but in no time mario will start gaining enough speed so that on every jump he lands on the upper stair.
Yeah I was trying to describe the effect in more detail. The point was that you need to get some boosts off of the stair skips until you are guanarteed to hit stairs.
FODA wrote:
I don't think i follow. You mean the height a plataform must be above the one mario is standing so that when he pushes against it he climbs it instead of being blocked? Obviously. But this is no BLJ effect, it's the normal phisics of the game. That's why mario can walk on stairs without needing to jump every step.
Yeah this was actually just me thinking out loud about stairs. I was actually considering if moving up slopes uses the same code... and maybe a similar downward skip exists for walking downstairs and going down slopes.
FODA wrote:
slides backwards...? you can't starts longjumping from a sliding position, or an standing sliding position even. You can only start LJ from forward movement. So if during a BLJ effect you stop doing the continuous BLJ and gets into sliding, there's no way to start BLJ again unless you start moving forward.
I have slid backwards to get a significant speed and upon jumping out of it landed on flat ground with considerable backwards walking speed. I was thinking about this effect. Actually I just remembered about all the crazy steep slopes in Bowser 3. I think it might be possible to BLJ up those if you do frame perfect jumps... hmm EDIT: Well I tried out bowser 3 and there ARE a few slopes you can BLJ up :) The super steep ones have very weird physics (you can tiptoe up them but you cant stand on them without falling), but they still might be possible with perfect frame jumps. There are though a few fairly steep ones that you can just stand on and they are pretty easy to BLJ up. I found them in the steep green metal area before the arrow lift. I dont see a compelling reason to BLJ there but if its possible there it could be possible on other slopes in the game :) A 45 degree slope you can stand on is pretty easy to manage on a console but Im sure in a TAS you could handle more shallow angles, especially if there is a way to get the initial backwards velocity.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
FILIM0NAS wrote:
@FODA What is exactly the key to do the BLJ? In the real console you must tap the A as fast as possible, but that doesn't always mean Mario can go fast enough. Do you have to meet the step right after you press A? Is the angle important?
A BLJ always gives you a speed boost (even on flat ground) but there is always friction which decreases your speed. Normally you cant do BLJ's frequently enough to overcome this friction. So BLJing stairs and upwards platforms works by allowing you to do BLJ more frequently than you would normally be able to. A forwards LJ gives you a boost too but the forward speed is bounded while backwards speed is not. I think the way BLJ works is that at the very first frame of a LJ mario is raised instantaneously from the ground so he can't jump again. If this amount is less than a traversed stair height or the movement of an upwards platform then the game puts mario back on the ground and mario can immediately jump again. When mario walks up stairs you can see how he is instantaneous transposed to higher ground levels. With stairs then that means that at most he can get the jump boosts at every stair (by abusing instantaneous transposition to higherground). Marios max speed on stairs is then limited by the number of stairs. The angle need only be sufficient to reach the next transposition (basically slamming into the front side of a stair is ideal). If he cant make it then I think FODA can add an extra jump to position himself better to hit the transposition. Each positioning jump though adds time and mario's speed is naturally decreasing over time (by friction) so this has to be taken into effect. So if you were to see what is happening in slow motion while BLJ up stairs you will see mario BLJ into a stair front face at which point he is raised up to ground and starts walking backwards (on the higher stairs). He will continue to walk backwards (even up several stairs) until he jumps again, at which point he gets another boost and slams into another stair again and starts walking backwards, etc. So on a console you will need to first start the BLJ cycle by BLJing into the front faces of stairs until you get enough speed to hit every stair, at which point as long as you just hit jump faster than the frequency of stairs, you will have optimal acceleration. There is also probably a limit in the code as to the maximum height that will allow a transposition (an upper limit on the stair effect). So mario should be able to walk over any small object. Actually when I think about it this way... if mario could obtain enough initial backwards momentum he should be able to BLJ up slopes. The trouble though is getting the initial momentum as with stairs you can just use a few multi boost-jumps (via transposition abuse) to slip into the hyper-BLJ mode. Maybe if mario slides backwards off a steep enough slope and hits an upwards slope he would have the required speed? Steeper upwards slopes would require less initial backwards speed. Of course sliding phyics might interfere... the upwards slope might be limited to ones you are allowed to walk on... FODA you might want to look out for these situations. I'll try to think of somewhere this would come into play...
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
I dont think there is any reason to believe the key doors can be passed through without a key. When mario walk ups to a door between different zones he enters a scripted animation which leads to the next area being loaded. The castle front door is an example of what happens if you get behind one of these doors. When mario is in the hidden black room he enters the opening door animation which causes the next area to be loaded just as normal (except he is on the wrong side). What this likely means is that if you somehow get into the fake area behind the key door and walk to the door you will just get the message that you need to have a key to pass (without loading the next zone). This can probably be tested easily by messing around with the memory. You could find the location of mario's position and simply add a value to his xyz to put him on the other side but Im not sure how much Mupen64 lets you try stuff like this. Now if you can find a way to get through all bowser doors (none involve loading), then you can do a 1-star, 2-key run (dire dire docks needs to be cleared to expose bowser 2).
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
Don't worry about it FODA... perfect use of the canon is plenty impressive and you already have a zip around the mountain in there.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
Just a head's up but Ive seen this video http://www.youtube.com/watch?v=TF_glNqW_hQ making the rounds around the internet. The video itself isnt too bad but some of the people posting it dont know it is tool-assisted. Just a head's up.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
Have you seen the record runs for Tall Tall Mountain? It's amazing. It's hard to imagine what kind of shortcut FODA could come up with to top it.
This signature is much better than its previous version.
Experienced Forum User
Joined: 2/16/2005
Posts: 462
Sketchie wrote:
Yes, doing a 120-star run is very long and tedious. You don't have to start over again just to try out new tricks. Just push on through and get this completed and get it published (hopefully!). :)
Well FODA has already completed a test run and knows what hes doing. The pace he is going at now is fine. Anyway the 100-coin route in SSL was very interesting. Really was not expected the teleporter in there. Anyway Im looking forward to LLL. I would like to see a double boost up the lavafall with a blind LJ over to the central poles :) I think that might be the fastest way to get there...
This signature is much better than its previous version.
1 2
10 11 12
18 19