Posts for primorial_soup

1 2
7 8 9
13 14
Posted: 12/13/2006 5:14 PM
Post subject: Re: WIP downloading
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Chamale wrote:
You need to update the link for the WIP. The link takes you to some "This domain is expired" page.
Actually, I need to renew my domain. It's been down for about a week now.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 12/13/2006 5:05 PM
Post subject: Re: Movie published
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Movie Description wrote:
In this movie, we watch a Ninja who transcends to outer space and steals the options from Gradius then proceeds to wipe out a bunch of other Ninjas and evil spirits. This movie also contains many Ninjas, need I say more?
Not to be picky, but would it be possible to change the movie description to something that actually describes the movie? Perhaps mentioning the shadow ninjas, the variety of weaponry, and perhaps that the setting is ancient Japan (which, although similar to outer space, is not quite the same).
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 12/11/2006 5:10 AM
Post subject: Re: DV memory address
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Chamale wrote:
FYI, I'm currently 95 frames behind Primo's run because of a move expected to save 150 frames overall. So the net gain is around 55 frames.
This confuses me. Perhaps you could post a WIP? It would make it a lot easier to give meaningful feedback.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 12/9/2006 6:32 PM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Excellent work, as always. I noticed something during the clone battle. After he's defeated, it seems like it takes a relatively long time for the white puffs to exit the screen. Perhaps this is only an illusion, but the delay seemed significantly larger than for other bosses. My question is, would it be faster to manipulate him towards the center of the screen (perhaps mid-air) for the final blow, so that the puffs don't need to travel so far? If so, could this be used for other bosses as well? Or is the delay a fixed duration?
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 12/7/2006 5:24 PM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Mildly entertaining.
Warp wrote:
This must be the most inefficient use of image format I have ever seen in my life: A 2-color 600x200 image in bmp format, taking a whopping 352 kB of space.
Ditto.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 11/22/2006 5:25 AM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
That was a thing of beauty. Everything about this run was more entertaining than the last. Yes vote, of course.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 10/30/2006 1:05 PM
Post subject: Re: @Primo
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Chamale wrote:
OK, I just wasn't thinking straight. According to pokemon stadium 2, the only pokemon that learn cut and fly are: Farfetch'd Skarmory Charizard Mew
Charizard can only learn Fly from Pokémon Yellow onward. In Red and Blue he unfortunately cannot :\
Chamale wrote:
By the way, Primo, if critical hits are so bad, why do you beat Brock with critical hits in your run?
Simple answer is that I was using Charmander. I don't have an emulator in front of me at the moment, so I can't check exactly, but I think it made the difference between 5 points maximum damage non-critical, and 6 points maximum damage with criticals. At any rate, it meant one less attack needed over the course of the battle. If you have a Squirtle with a high enough starting special, you can two-hit KO with out critical hits, which is definitely the optimal strategy.
Chamale wrote:
Edit: OK, the glitch doesn't work. The idea was that you don't go to a pokemon center at any point, and use Escape Rope while in Mount Moon. The game, not knowing where to send you, sends you to Hall of Fame. There's another alleged version, involving an Abra's teleport, and I'll have to try in the interest of TAS science. I'm wondering how Squirtle will have the PP to make it that far (Actually, not really. I plan to replace Water Gun with Mega Punch, then replace something with Water Gun.)
Sure doesn't. If you haven't used a Poké Center, your last save point is at your house in Pallet town (I think this is because your mother counts as a Poké Center, in some strange manner). Another variation would be to simply die somewhere along the way. If this sent you to the Elite Four, that is a bug that most certainly would have been noticed by the developers. Mega Punch is a strong attack obtainable very early, but it has its drawbacks. I've mentioned elsewhere that certain attacks have an animation which is significantly slower than others... Mega Punch is one of them (55 frames slower than Bubblebeam, for example). Just something else to think about.
Chamale wrote:
Edit 2: OK, the other variation didn't work either. I'm glad I transferred all my pokemon to pokemon stadium... all I deleted was a ditto.
If you have any questions about anything that might possibly compromise your game data, feel free to ask beforehand. Many people have already done a lot of the experimentation for you :)
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 10/29/2006 3:31 PM
Post subject: Re: RE: Fractal Fusion
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Chamale wrote:
Oooh, here's a thought: Taking the name "ACE". It wastes no time at all on the run, allows glitching of a level 130 Snorlax (He'll just go to 100 anyway), and flows much better than "EEE defeated Youngster".
This should actually be two frames faster than 'EEE': 
'EEE': > > > >A A AS
'ACE':  A> >A> >AS
Why two and not three? For some reason when selecting the letter A as the very first character, you have wait one frame longer than you do to move the cursor. Optimal names would be 'BCE', 'BDE', or 'CDE', which all save one frame over 'ACE'. There's still a few things about using a Spearow which confuse me. When passing through Veridian, you won't have any Poké Balls. This means that you will have to stop at the Poké Mart first, and then go over to Rt. 22 to catch him, and then get back onto course. This seems like it would add a lot more than 25 seconds, as you proposed. Also, you proposed keeping the Spearow at a low level to learn Cut and Fly... but Spearow can't learn Cut. Trading for a Farfetch'd would also be a much slower choice. And also, since this is one of your main tactics, Pidgey has the fastest call of any other pokémon. About Squirtle. In addition to fighting the unavoidable Bug Catcher in Veridian Forest, Squirtle will also need to fight a random encounter, because he's about 8-9 Exp. Points short of leveling. See Tilus' original Pokémon Blue Movie or even my 151 WIP for details. Also, I don't think you'll save as much time in the Gym as you think; Critcal hits are very broken at low levels. Without critical hits, Squirtle can 2-hit KO both the Geodude and the Onix. With critical hits, Squirtle can 2-hit KO both the Geodude and the Onix... and waste about 80 frames from message box text. Just some things to think about.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 10/22/2006 3:06 PM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
I had actually considered redoing this run myself, but I haven't had a lot of free time as of late. I don't think that there's any problem with naming the hero 'EEE'. I had been considering using the name 'I', because it flows well in the dialogue; 'I defeated so-and-so', etc. For choosing a rival name, I would probably use pokémon blue, and choose the name 'ASH'. It's around 65 frames slower than using a one character name, but still 77 frames faster than what I did, which I think is a reasonable stylistic choice. Don't forget that you can exit the name screens with the start button (for pokémon as well). Further analysis. Also note the bit about switching attacks. I also made a route mistake just before Bill's. The Hiker with four pokémon is avoidable, and you can fight the Jr. Trainer above him instead which only has two. This will also give you two extra Bubblebeams to work with, which means you can avoid using Hydro Pump for the rival's Pigeotto on SS Anne (Hydro Pump has an 'no-animation animation' which is 55 frames slower than most other attacks, so it's best to avoid using it). How 'difficult' a pokémon is to catch is independent of their level. I know from experience that a Snorlax can be caught with just a plain poké ball. When will you be catching Snorlax exactly? There are a few sections that you could probably avoid doing with Gyarados, and then do later with Snorlax. Celedon Gym comes to mind, and of course the Silphco. Building. A L132 Snorlax should have an openning moveset of Body Slam, Harden, Double-Edge, and Hyper Beam. Harden is worthless, and Double-Edge can't be used if you intend to do a no damage run. For a third attack, Rock Slide might be a good choice, mainly for the ghosts. It's a normal attack (as opposed to special), which means it will utilize Snorlax's high attack, and it won't produce a 'super effective' message as Earthquake or Psychic would. The fourth attack could be Surf or Ice Beam. I think the Veridian Gym will require one of the two. Surf seems like a better choice, because it's just as powerful, and it has more PP. I like the idea, and I think it could be a good run. Good Luck :)
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 10/7/2006 4:20 AM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
It seems likely that this test is intentionally easy... you know, to get people spend money on a piece of paper, sent from an internet site, which tells them how smart they are. 142, since people seem to be interested.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 9/24/2006 5:12 PM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Mukki wrote:
I'm thinking this run can be improved, Primo's that is. I reckon there could be a big time saver (a minute or so maybe...) in manipulating the enemy's moves so that it always uses the move with the shortest name, this could usually save several frames per battle over Primo's run. This sounds extremely tedious to TAS however....
For the majority of the run, I chose attacks which had a miss ratio better than 1:256. The were a few battles where this wasn't possible, for example Lt. Surge's Raichu, which refused to do anything except Thunderbolt or Thundershock. I spent somewhere in the neighborhood of ~2000 rerecords on this single attack alone. Short synopsis, this could potentially save a few hundred frames over the course of the run. Whether the run would be able to be completed in a single lifetime is another question.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 9/13/2006 11:57 AM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
AzHP wrote:
Too bad the author hasn't updated in over a month =/
*Hint, nudge...* I was on vactaion for about two weeks, so that's where some of that time went. I'm still trying to work out a few details on the route. As soon as I have it, it'll be posted :)
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 9/12/2006 12:01 AM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Nemo2342 wrote:
Now we just need someone to do a run for the Gameboy version :D
The games are fairly identical, except that the gameboy version seems to have annoying things like not being able to buy the alarm flute (you have to find it in a cave) and Zebu popping up everywhere to explain the item you just found (like ball of wind and rabbit boots). I think the gameboy version is best left undone.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 9/11/2006 8:57 PM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Harvest Moon 64. It's an entertaining game to play, but watching someone farming for a few hours would be downright boring.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 9/10/2006 11:04 PM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
The name of the rom I used was "Ninja Spirit (J).gb", although I know it's mislabelled, because it is very clearly the english version. A correctly labelled version I was unable to find.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 9/10/2006 10:48 PM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
JXQ wrote:
It seems to me that certain things make the music lag in this game, but not the actual gameplay. Is this right?
For the most part. There's a few places which cause the music to lag, but don't lag the gameplay. Stage 3 does contain three frames of actual lag though. I decided to leave these parts in because I thought they looked impressive.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 9/10/2006 9:39 PM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
I'm glad to see that they didn't change the game physics too much in the SNES version. I had a few questions about 8-4: 1. There's a place where you wait for a pirhana plant when you have fire power. Why not just shoot it? 2. The swimming didn't look optimal... at the end it looks like you wait for no reason. Were you manipulating something? Other than that, the run was very entertaining and for the most part impressive. Yes vote.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 9/10/2006 7:52 PM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
I voted yes for the last, so I'll be voting yes for this submission as well.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 9/10/2006 7:18 PM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Run completed and submitted.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 9/10/2006 11:44 AM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Bag of Magic Food wrote:
What's wrong with Pac-Man? I want to see Pac-Man!
It... never ends.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 9/9/2006 1:27 PM
Post subject: Re: hey, AdmiralJonB
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Highness wrote:
Brainfuck is also a funny one! http://en.wikipedia.org/wiki/Brainfuck
This is, in fact, nothing but a Turing machine.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 9/9/2006 11:41 AM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
It depends on who you are and what you're doing. If you're a company wanting to develop a stable application for all platforms as quickly as possible, Java is probably your best choice. If you're a researcher doing numerical analysis, C/C++ would obviously be faster. Is Java easier to learn to learn than C++? Certainly. I can't possibly see how this could be construed as a bad thing. The code is also easier to read, especially for those who didn't write it themselves. This makes software maintenance more cost effective, and less of a headache for the one actually doing it.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 9/9/2006 10:40 AM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
qFox wrote:
You're crashing into cars and hitting the dirt a lot. Sure you don't break a lot, but this is the same. I'd say, you have the power of slowdown and premonition at your command... use it. If at some point you hit the dirt in a corner, rerecord and slow down until you don't hit the dirt. If you bump into a car, rerecord and don't. This of course not for the kisses you seem to make on purpose (but only if these dont slow you down either, if that's the case, more kisses are a pre ;) and the crashing at the end. I'm voting no because the beginning is rather sloppy and since this game doesn't have a lot of things to take care of I think not slowing down unless absolutely required should be one of them.
This submission should have had a much better description. Basically everything I wrote in this post: http://tasvideos.org/forum/viewtopic.php?p=89293#89293 Basically, braking is a lot slower than going off the road for a bit. Once you're off the road, you can use the accelerator and turbo charger again, and so it hardly slows you down at all. The other alternative is to let off the accelerator while going around a curve. This doesn't always work though. In the first track, for example, there a curve on which you need to be going 180 MPH or so not to skid, whereas going off the road a bit only drops you down to 205 or so. Naturally going off road is a faster choice. Hitting cars on from the side doesn't slow you down at all. It doesn't take a whole lot of observation skill to see that he is still travelling top speed after a side collision. It's also very useful for going around corners. Hitting a car from behind does slow you down; however, this run doesn't do this at all. I think it is an unfair analysis to say that this run is sloppy without knowing the game at all.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 9/8/2006 10:39 PM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
This run is much, much better than your last. Especially tracks 6 and 8 (which I actually mentioned specifically in your last run ;). You also pull off a lot of impressive maneuvers, particularly in track 6, which was nice to see. If these times could be beaten, it's only by a matter of a couple of frames. Voting yes :)
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Posted: 9/8/2006 8:21 PM
Quote
primorial_soup
Experienced Forum User, Published Author, Player (87)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
I hope you do well. Let us know the results afterwards :)
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
1 2
7 8 9
13 14