Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
Vykan12 wrote:
I don't think you guys are following. Rebecca Black got signed to Vevo, a major record label. She is probably going to make millions... for making a music video so bad it went furiously viral. In essence, she's going to make more than a surgeon...for trolling the internet.
Doesn't that make your blood boil with rage? The fact that pure irony is making people rich and famous instead of, you know, talent?
Yeah, I used to think about this a lot, now I simply accept this stuff and don't care about it. I live in a country where most athletes struggle to find sponsorship while popular soccer players who are almost always not nearly as prepared or hard working make more money in a year than what I'm gonna make in my whole life.
Every carnival, some stupid band has to make a song that has at most 20 different words in it that will be so successful that will be played to exhaustion until my ears can no longer tolerate it, while some people that are very knowledgeable of music and can make good stuff will simply be ignored.
However, despite these setbacks I still think that if people try hard enough, they will be rewarded. My advice is: "if you don't know how to kick a ball or how to sing pop music, stay in school" :P
4matsy wrote:
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
Evaluate
Suggestion: It's not a coincidence that this is posted after Bobo the King's problem. Try starting from n+1 = sqrt((n+1)^2) and see what you'll get.
EDIT: It's actually simple:
We have
And, similarly, and .
Now, just keep nesting:
Continuing this process indefinitely:
For n=2, we have:
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
Thanks a lot, PJBoy! However, I think the run won't sync in v23.5. Apparently the desync is caused by the newer versions not understanding a hard reset, perhaps the conversion of the movie file can fix this.
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
Por favor, quando for pedir ajuda, explique claramente o que você deseja, isso facilita a vida de quem for ajudar. As duas postagens que você fez nesse fórum, feitas com frases isoladas, fazem pouco sentido.
O emulador adotado pelo site para MSX é o openMSX, minha experiência com ele é nula, mas posso indicar quem pode ajudar se você tiver problemas: http://openmsx.sourceforge.net/GingeYamano wrote:
Será que posso fazer um tas criando uma página web?
Sim, porém recomendo que você crie um tópico sobre o jogo no fórum da plataforma dele, é lá que as pessoas que podem te ajudar a fazer o TAS vão olhar. Alguns TASers gostam de criar uma página de projetos deles no googlecode também, mas eles são mais experientes e têm domínio melhor de inglês.
GingeYamano wrote:
posso fazer um run em Resident Evil 5?
Em tempo real, sim, por que não? Fazer um TAS é imposível no momento, esse jogo só tem versões para PS3 e Xbox 360, e não há nenhum emulador estável o suficiente para implementar as ferramentas necessárias para fazer um TAS, e não parece que estamos próximos de ter um.
EDIT: Agora, vendo melhor, parece que tem uma versão para Windows também, recentemente um "emulador" para fazer TASes de jogos de Windows (hourglass) foi aprovado, mas a funcionalidade dele é limitada, e só deve rodar os jogos mais simples, o que RE5 com certeza não é.
GingeYamano wrote:
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
Username_Required wrote:
Still, it bothers me that I can't seem to find-the data for the keys of a level (probably bools)
-the data for the keys of the current level
1.I just go for the number of changes (which should occur ONCE in a level
and be reset when I exit the level before completing it/when completing it)
But after 4 iterations (the change counter should be 8 by then) my search
gets reset because I have none left... what to do?
-the x and y coordinates *doh*
2.I only found the Scroll coordinates (which seem to be bound to the Map,
like seen in the map viewer)
which wrap around (so no absolute digits)
Is there another way to determine increasing values than using the greater than?
For the first two, what did you do exactly? I couldn't figure from what you wrote. One thing I'd try first is search "Equal to" for some frames while inside the level, exit it and use "Not Equal to".
For the coordinates I don't know about WL3, but many games store only the camera position and your character's position within the screen. Try to move your character while keeping the camera static and you may find this other address.
In the sub-30 Spiritia run, is there any reason that he went through the Prologue stage? It's been some time since I played this game, but I remember it could be skipped. Good luck to whoever takes this project, that was one of the best TASes I've seen for quite a while.
Your YT encode is not aspect corrected and the audio desynched for me, did you simply dump it from the emulator and uploaded it? oO
Anyway, watching on the emulator...
EDIT: It had its boring moments, but it's an overall solid run, this one is easy to decide. Yes vote.
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
You can always have everything saved at the script's tables. The idea of using an external file is that the stored data won't go away after the user closes the emulator.
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
To avoid going through the entire movie file every time you want to check presses, you could try to have a presses counter and increment it as the frames are advanced. The only drawback of this method is that whenever a rerecord happens, your values will no longer be correct. To circumvent this, you could store this information when a savestate is saved and retrieve it when one is loaded
Since I don't know a way to put data inside a savestate, a good option would be to create an external file that links the savestate to the amount of presses that occured up to it. Use functions savestate.registersave() and savestate.registerload(). Whenever, a savestate is saved, hash it somehow (frame number, buttons pressed, checksum of a memory range) and store it in this file with the amount of presses. When a savestate is loaded, do the same hash function, look for it in the external file and replace your current data with it, if it's not found, make it go through the movie file again to find it and store it in the external file.
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
Mister Epic wrote:
Sorry, but an usual combo video is more impressive than your TAS, and it's not even a TAS.
This is true for nearly all of the fighting games TASes published here, even one that's a recommended movie (I only don't say all of them, because I think the MK4 TAS deserves its tool-assisted label). Through proper segmentation, experienced unassisted players can achieve similar results.
Although admittedly there's technical complication in making these types of movies, such as AI manipulation, I don't see how this alone can be enough technical merit to warrant publication, as movies that consist of simple manipulation of dice throws/coin flips and perfect clicking of menus are most likely rejected.
IMO the difference between "this movie is entertaining" and "this movie should be published" must be addressed more carefully when voting. It's pointless to TAS a fighting game unless you can do something that really needs tool-assistance to be done.
I've already noticed this behavior when I submitted a joke run for April Fools that was entertaining but I wouldn't publish in a thousand years. I don't know whether to follow my opinion alone or to take into account the common opinion, so I'm abstaining from voting due to my strong bias against these playarounds.
Sorry for the negative comment, x2poet. I'm actually a silent admirer of your work and appreciate your effort to expand the Arcade TASes section.
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
The title says almost everything. Lately, I've been analyzing some games, trying to find out their internal algorithms. This is, unless I'm completely mistaken about the terminology, reverse engineering of software. Someone told me months ago that there's no problem with that as long as you own a legal copy of the program (which I do, I have the cartridge for all the games I'm analyzing) and you use that information for personal use only, that is, not release it to the public or sell it.
The person who told me that, however, is not a reliable source and I'm no good at laws either, so I decided to create this topic to ask if it's really OK to share information obtained through reverse engineering of copyrighted software or to distribute a file of key presses made with this information. And if not, what legal consequences that would implicate.
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
* Because I didn't know that o_O
* It doesn't work the same way as the Old man version. You get Missingno. in Cinnabar coast, but it's not your name that goes there and those pokémon will not appear.
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
@System Error
No, we haven't considered using a glitch pokémon as a primary fighter for the E4. For story battles, we "borrowed" most of our strategy from primo's 1:17 Blue run. IIRC the usage of a glitch pokémon as a fighter was suggested many times in the R/B thread. The problem with glitched mons is that they cause extra lag when used because of their buggy graphics and require at least one T-fly to get (taking into account that they don't get registered in pokédex, it's extra time, probably more than E4 extra messages). Some are even higher than 200, which must be got with the Old Man glitch, requiring to change the name and, as a consequence, our route.
Mukki must be almost finished with the subtitle script. It may clarify some things once released.
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
System Error wrote:
Also, on the subject of the movie, obvious yes, but one curiousity. Was it ever considered using a glitch Pokemon in order to cause more damage to the stronger foes (Elite 4)?
Do you mean the high level ones at Cinnabar coast? The game resets them to L100 as soon as they gain experience.
Glitcher wrote:
ElectroSpecter wrote:
stuff
That's because he held A + B + Down.
It's hard for me to tell whether you're being sarcastic or not. In case you're not or anyone believes this, it's a myth, no chance to catch a pokémon increases when you do any button combination during the pokéball animation.
Kuwaga wrote:
Catching Mewtwo with an ordinary pokeball and full energy was also possible on GameBoy after about 256 resets
This sounds coherent. Mukki determined that the chance of catching a full health Mewtwo with a Pokéball is 3/256. I spent a freaking long time manipulating that catch to get a reasonable waiting time, mostly because there are some frames that the RNG "lags" and you have no chance to catch the pokemon if you press A that time. Mukki found this out too. Still today, I don't know how they could program the game like that.
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
L64 Chansey in Unknown Dungeon is the only thing found in the wild that has a special so high to allow glitching Bulbasaur. Fighting 'M is to dupe Rare Candies again, and since we had to fight him anyway, we decided to evolve it to Kangaskhan, since it saves one encounter in Safari Zone and the pokédex registration.
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
hezekiah wrote:
Could there have been time saved in the Elite Four by teaching a different move over Thunderbolt? Especially on Bruno, those "It's Super Effective" messages feel like they should be avoidable. A quick glance at Bulbapedia seems to indicate that a possible candidate would be Psywave, which is the reward for beating Sabrina. All I'm not sure of is if the time saved by using a different move (IE, removal of Super Effective and perhaps some Critical Hit messages) would be enough to make it worth taking the time to teach it the move.
Wouldn't a level 100 pokemon using Psywave deal at most 150 damage? Most pokemon at the Elite Four surpass this iirc, so there would be crits anyway, a move I considered was Fire Blast, but as you seem to have noticed, Thunderbolt is the only move we can delete and is superior to Fire Blast in terms of PP. One could change moves in the middle of the league, but it still seems unlikely that calling the menu and going through those annoying messages is faster than some crit/SE messages, if the switch to Fire Blast could actually get rid of them.
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
Traduzindo...
Master of pigs wrote:
o boss de Diamout Dust, Gene Gadge e Panic Puppet não tem como melhorar, está muito no limite mas parece que a versão PC tem a vinheta mais rápida que da versão genesis o que me leva a parecer perder 1 segundo.
Para a Diamond Dust Zone Act 3, 0:49 já está bom, você não precisa se preocupar em melhorar esse tempo. Porém, como você pode ver em um dos meus vídeos, você pode derrotar aquele chefão em 0:49 sem tomar dano, e isso passa uma impressão muito melhor. Na sua última TAS você tomou dano o tempo todo. http://www.youtube.com/watch?v=9Mz0DERTfyA
Para a Gene Gadget Zone Act 3, não acho que o SpringGod usou a versão para PC para conseguir o tempo dele, então deve ser possível conseguir 1:13 seguindo a estratégia do meu vídeo, executando com precisão de TAS: http://www.youtube.com/watch?v=DkNUCA4S9iA
Para a Panic Puppet Zone Act 3, 1:39 muito provavelmente é o mais rápido possível para a versão em Genesis, então não se preocupe se você não conseguir superar essa marca.
Estou ansioso pela sua próxima tentativa. Boa sorte!
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
1. If we don't teach him Slash, he'll try to learn it again immediately after evolving.
2. Yes, but when we use the Escape Rope on the teleporter the warp animation is faster.
3. We can only have one active box at a time, if we let him in the box, we'd have to switch back to that box, losing some time.
4. The repeated Trainer-Fly is to just leave the area glitched to fight the pokemon later, we don't need to do the encounter immediately after doing the glitch. Chaining it this way is faster because there's less walking.
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
Um bom lugar pra se começar a desenvolver táticas contra chefões é aqui: http://tasvideos.org/GameResources/BossFightingGuide.html
Pode perguntar se não entender alguma coisa. Agora, se quiser alguma coisa específica de Sonic, vai ter que esperar alguém que conhece bem o jogo passar por aqui. Procurei no site e parece que não tem nenhum tópico sobre esse jogo no fórum de Genesis, aconselho você a criar um lá, pois é o lugar que a maioria das pessoas que pode te ajudar vai ollhar.
Experienced Forum User, Published Author, Player
(42)
Joined: 12/27/2008
Posts: 873
Location: Germany
To solve it I used a small lemma I deduced. It can be shown by trial and error that a + b + | a - b | = 2 max(a,b)
This allows one to trade the variables a and b for only one, I'll call it u=max(a,b) to simplify things. So, we have:
2u + 2 | c - d | + 2 | 2u - c - d - | c - d | | = | 2u - 2c | + 2d + | 2u + 2c + | 2u - 2c | - 4d |
Simplify by two:
u + | c - d | + | 2u - c - d - | c - d | | = | u - c | + d + | u + c + | u - c | - 2d |
Sum u+c+d on both sides:
2u + c + d + | c - d | + | 2u - c - d - | c - d | | = u + c + 2d + | u - c | + | u + c + | u - c | - 2d |
Can't see the beauty yet? Making it clearer:
2u + (c + d + | c - d |) + | 2u - (c + d + | c - d |) | = (u + c + | u - c |) + 2d + | (u + c + | u - c |) - 2d |
Applying the lemma:
2u + 2max(c,d) + | 2u - 2max(c,d) | = 2max(u,c) + 2d + | 2max(u,c) - 2d |
And using it yet again:
4max(u,c,d) = 4max(u,c,d)
Equality is obvious. I'm not very sure how Bobo got to this, but my best guess is starting from max(a,b,c,d)=max(a,b,c,d) and using repeated nesting of the property 2max(a,b) = a+b+|a-b|
Suggestion: It's not a coincidence that this is posted after Bobo the King's problem. Try starting from n+1 = sqrt((n+1)^2) and see what you'll get.
EDIT: It's actually simple:
We have 
And, similarly, 
and 
.
Now, just keep nesting:
Continuing this process indefinitely:
For n=2, we have:
