...What in the name of Yuji Naka is Dolphin doing that causes the pitch of the music to bend sometimes when you're dashing that much? I love me a good Sonic Team soundtrack, and the oh-so-subtle sound of the tunes warping behind Billy's yell makes the hairs on the back of my brain itch.
Dolphin isn't doing that. What Dolphin is actually doing is not rendering enough audio for how much video there is. What RGamma does to compensate is stretch the audio samples to fill out gaps.
There's a huge cognitive dissonance in my mind about the concept of "rooting" an open-source OS.
It's a big like saying "you can download a pirated version of Linux for free here!" Duh...
Android is open source, but getting a phone with android on it, not so much, because the phone manufacturers control the specific distribution you're getting and what you can do with it.
0:27: Lol, clipping through the flipper
Nice work! It's amusing to me how much of the stage forces you into 'ball mode', and whenever you're out of it you immediately go back to normal speed techniques.
My suggestions for Android gaming (and I'm biased towards skill games, so <.<) are:
Rhythm games:
jubeat plus (the iPad version is better, if you ever get a chance. BTW, I think this is only in the japanese google play store, so you might need to fiddle with stuff)
Cytus and Deemo (Rayark isn't the best at rhythm games, but - they're playable, they come with a ton of songs for free, so I am obliged to recommend nonetheless)
Beats (Stepmania for Android)
Shmups:
http://shmups.system11.org/viewtopic.php?f=1&t=43212
All the ones he lists are good. There are (or should be) Lite versions for all of them, so you can see if they feel and play right before shelling out money. I've been enjoying Espgaluda II lately.
I also hear there is MAME for Android: https://play.google.com/store/apps/details?id=com.seleuco.mame4all
I have no idea how well it works, but if it's good, that will open up a huge amount of games available.
I think that if this is re-visited, it should be renamed <character> ring attack.
The advantage of calling it a ring attack is that getting the perfect bonuses is just a side-effect. It doesn't lead to the weird ambiguity on levels where perfect bonus is impossible of 'well, do you try as hard as you can, or just ignore rings'. Ring attack = you always get as many rings as possible.
True, but maybe two versions could be available, a monitor less, and a full run with the max monitors too.
Nah. It should be these:
-Maximum amount of 'field' rings
-Maximum amount of ring monitors broken
-Maximum amount of rings from non-respawning field objects/enemies that produce them
But not these:
-Rings from checkpoints bonuses, special stages, big rings, etc
Same as in S3&K.
"Xom thinks this is hilarious!"
http://crawl.develz.org/wordpress/downloads
The popular roguelike DCSS can technically be TASed now, because of this:
https://crawl.develz.org/tavern/viewtopic.php?p=172145#p172145
You can start the game with -seed <num> and that fixes the starting RNG to a certain number. From thereon, if you always input exactly the same actions, you always get exactly the same result back. And you can use Crawl's built-in LUA scripting or otherwise to re-input actions for you, for example.
Like in Nethack, there is one RNG used for everything, and there is a 0 turn action you can perform to change the RNG - Start as a Chaos Knight. Praying to Xom prints a random message, which advances the RNG. Xom is particularly useful for such a TAS because he has a ridiculously wide variety of actions he can perform, including teleporting the player around, granting him gifts, polymorphing monsters in line of sight and so on. With either that or a ring of teleportation (or both) and RNG manipulation, you could go anywhere you wanted really fast.
This would probably be really horrible to do without writing some helper scripts (unless it works in Hourglass which I seriously doubt), I'm just raising it now since it is technically possible now.
(I'm also interested in seeing a TAS of Brogue, which has player-selectable seeds that determine the dungeon unfixingly in advance as well as the starting RNG seed for the game, and a deterministic replay system.)
I think that if this is re-visited, it should be renamed <character> ring attack.
The advantage of calling it a ring attack is that getting the perfect bonuses is just a side-effect. It doesn't lead to the weird ambiguity on levels where perfect bonus is impossible of 'well, do you try as hard as you can, or just ignore rings'. Ring attack = you always get as many rings as possible.
A lot has been happening on the Viking front. I'll try to outline it here so if there's anyone who wants to get into this game, they can be up on current knowledge more effectively. Miketheviewer has been very active breaking the game repeatedly, and I've done my share applying it to human speedruns, finding setups, etc. Yagamoth has also helped some, though he has other projects on his back right now.
First, there's a new core movement technique called Erik-launching. When you start a headbash with Erik using Y, then immediately switch to another Viking (1-player) or switch the camera to the other controller (2-player, done with L+R), Erik will continue moving through the remainder of his headbash cycle, meaning he moves a decent distance without needing to be controlled. You can get extra free distance with him like this.
Generally, collisions are not detected for objects offscreen. There are exceptions, even within the same object type (some electrical fields will kill offscreen, some won't), but even walls aren't invincible. To abuse this, you need some form of automated movement - a fall, a spring, Erik's launch, inflation, an elevator, etc. This creates the following exploits:
- Electrical fields can often be passed through.
- Fans will not blow you into spikes, unless you were already in the wind.
- Due to his speed and the launching described above, Erik can pass through block doors, such as the ones seen in CMB0 and TRPD.
- Enemies will almost never hurt someone while they're offscreen.
- Items can be duplicated, provided you switch away from the relevant Viking at the right time. Room for error varies, but is never more than a half-block wide. The current human route only duplicates keys.
Pause-buffering can help with many of the tricks in this game. Both Start and Select work for this, but I recommend Select because Start can end the level if you press Left in your buffer, and Select doesn't block part of the screen. The D-pad, Y, and shoulder buttons can all be buffered, and you can buffer as many inputs as you like. In 2-player, you can use both Select buttons to rapidly re-buffer to advance a few frames. If you buffer left/right plus a shoulder button, the Viking who paused the game will move exactly one pixel.
Yagamoth found that Erik's fall speed is decreased while holding the B button. This is of course familiar to avid players in antigravity fields, but also holds true outside them, and can slightly lengthen a jump.
When you attempt to open a keyhole with a key, your collision disappears for a moment, preventing fall damage, and allowing you to fall through key-activated floors, assuming you can set up a fall that would collide with the floor - with a jump, spring, or higher ledge. The key doesn't have to match the keyhole - in fact it's preferable that it doesn't.
Miketheviewer calculated the load time difference between SNES and DOS versions, and it totals 2 minutes, 36 seconds (or 2:38, Mike can correct me, I can never remember which). For deciding the world record, I recommend making the conversion - though my current record is ahead of Chansey, the next fastest runner, with or without the conversion. The only other difference I have seen so far is that the screen is four blocks wider on DOS, forcing a few minor route changes due to onscreening of some things that would be offscreen on SNES. That difference doesn't seem to cause a significant advantage or disadvantage based on version - in fact, sometimes it helps, and sometimes it hurts, and it's negligible either way.
Current record is 1:02:58 by me, 2-player. I currently need to incorporate some minor route changes, learn to do the new Erik-launch practically everywhere, and move closer to deathless. After that, the "switch-staggering" I mentioned in an earlier post will be the only remaining hurdle... unless Mike keeps finding new things. Here's my run video.
Link to video
Above comment is for the SNES version, but should apply to the Genesis version too.
Found this nicovideo. (I searched for the game's japanese name + TAS.)
http://www.nicovideo.jp/watch/sm22441823
Doesn't seem like there's anything else (except the 10 minute mode TAS that was already linked in the first reply.)
Yearly reminder that this still needs a TAS and that it (all of Remar's games) TASes fine in Hourglass.
I will take this on as my next project
Cool :) There's no shortage of speedruns and routes and information to look at, you should have everything you need.
Will you do easiest or hardest difficulty?
I wonder if this is possible, but...
Can you guys add a feature to BizHawk that lets you change the game's Saturation/Brightness/Contrast ?
Not many emulators have this afaik, the only one I've seen so far was PCSX2. But it's pretty helpful especially when you can raise the brightness on dark games.
Would a minimum A press run also aim for time? So lowest a presses but in the fastest time? If so than a minimum A press run would be really hard to make.
So what does this mean for SMW TASing? Is chuck eat now banned in non-ACE categories? Or is it OK as long as you don't get an 'executes data as code' kind of sprite in your powerup box?
Even if technically this and the previous TAS are the same category, it would be a big shame for the site if one of the best, glitchiest, most entertaining TASes became obsoleted (which is as good as 'gone from the site' in a few senses).
It should be allowed to co-exist at least until a 'purer' TAS replaces it. In the mean time the category labels obviously won't be very accurate/meaningful, but I think that's an OK tradeoff to make. (I wouldn't ask for it if the previous TAS wasn't as good as it was!)
I see no reason why a similar concept would not work as applied to the "minimum A presses" category of SM64, once the rate of finding improvements had slowed down a bit. The publication description can always be used to indicate where something can be done better, but in the interest of making the movie more publishable, was not done that way.
Main problem is that almost all of the most interesting minimum A button press strategies have as a byproduct of their interestingness incredibly long real time (all the cloning-requiring solutions for example). It would also be arbitrary - how 'long' or 'boring' does it have to get before you start deciding not to save A presses?
To all people hoping a minimum A button presses TAS will ever come out: Get annihilated.
Link to video
Pannekoek2012 finds a glitch that takes 11 hours and saves 1 A button press.
Next TAS I wanna see if people like the idea of it:
"Hey hey hey! it's time to make some crazy money! Are you ready? Here! We! Go!"
Crazy Taxi! on hardest mode with shortest amount of time. try to see if it's possible to get the Crazy license and maybe even 1st Rank (to get Crazy license, you need to rake in 20,000$ in taxi fares). Shortest amount of time you can start with is a mere 35 seconds.
The thing is that tasvideos is not meant to be a repository of all TASes, or even all good/amazing/perfect TASes.
For starters, it won't publish a TAS unless the creator of the TAS him/herself submits it. That should be your first clue.
