Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
The NES page already has many worse games than Air on it. The vast majority of these movies would greatly improve just about any page that's lucky enough to be graced by their presence.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
Slowking wrote:
I googled a little. Guess it's relatively new. That explains the lack of support through hardware and that it's still fringe in general.
Is there a short 1080p video encoded with 10 Bit somewhere? Would like to test if this computer can handle it.
Why limit yourself to just 1080p? While you're at it, go for a real test.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
I'd gladly share tales of strife and the blinding fury of a thousand suns, but the executive summary is that no revision of mednafen-rr can be built on Linux.
Kiss your PC-FX TASes goodbye, gentlemen.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
The goals don't make sense to me. Collecting all instruments and then proceeding to enter the Egg without using them seems pointless.
The point of a run like this is to show off glitches that a straight speedrun wouldn't feature; I don't think the goals used are optimal to that end. As such, I'm voting No.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
Version 1.4 features a glitch that allows Keen to skip all keystones in the final level without using IBT. This could be useful in a 100% run that forgoes some or all forms of bullet collision abuse.
Version 1.0 features numerous clipping errors that were fixed in later releases; it is definitely the version of choice for glitched runs.
For a low-glitch run, defining the goals is tricky. IBT isn't the only way to abuse bullet collision; the rest aren't quite as game-breaking, but they do make enormous time savings possible.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
Tseralith wrote:
Which difficulty are you talking about? Is there one that's even harder than Lord Of The Game? I can easily access this difficulty from a new game with no saves in beforehand. Could you please link the page where it's mentioned?
Radiant wrote:
I don't think that's true.
turska wrote:
see OpenTyrian source for details.
Language: C
// increases game difficulty based on player's total score / total of players' scores
void adjust_difficulty( void )
{
const float score_multiplier[10] =
{
0, // Wimp (doesn't exist)
0.4f, // Easy
0.8f, // Normal
1.3f, // Hard
1.6f, // Impossible
2, // Insanity
2, // Suicide
3, // Maniacal
3, // Zinglon
3, // Nortaneous
};
assert(initialDifficulty > 0 && initialDifficulty < 10);
const ulong score = twoPlayerMode ? (player[0].cash + player[1].cash) : JE_totalScore(&player[0]),
adjusted_score = roundf(score * score_multiplier[initialDifficulty]);
uint new_difficulty = 0;
if (twoPlayerMode)
{
if (adjusted_score < 10000)
new_difficulty = 1; // Easy
else if (adjusted_score < 20000)
new_difficulty = 2; // Normal
else if (adjusted_score < 50000)
new_difficulty = 3; // Hard
else if (adjusted_score < 80000)
new_difficulty = 4; // Impossible
else if (adjusted_score < 125000)
new_difficulty = 5; // Insanity
else if (adjusted_score < 200000)
new_difficulty = 6; // Suicide
else if (adjusted_score < 400000)
new_difficulty = 7; // Maniacal
else if (adjusted_score < 600000)
new_difficulty = 8; // Zinglon
else
new_difficulty = 9; // Nortaneous
}
else
{
if (adjusted_score < 40000)
new_difficulty = 1; // Easy
else if (adjusted_score < 70000)
new_difficulty = 2; // Normal
else if (adjusted_score < 150000)
new_difficulty = 3; // Hard
else if (adjusted_score < 300000)
new_difficulty = 4; // Impossible
else if (adjusted_score < 600000)
new_difficulty = 5; // Insanity
else if (adjusted_score < 1000000)
new_difficulty = 6; // Suicide
else if (adjusted_score < 2000000)
new_difficulty = 7; // Maniacal
else if (adjusted_score < 3000000)
new_difficulty = 8; // Zinglon
else
new_difficulty = 9; // Nortaneous
}
difficultyLevel = MAX((unsigned)difficultyLevel, new_difficulty);
}
Language: C
case 0: /* New game */
if (select_gameplay() && select_episode() && select_difficulty())
{
gameLoaded = true;
initialDifficulty = difficultyLevel;
if (onePlayerAction)
{
player[0].cash = 0;
player[0].items.ship = 8;
}
else if (twoPlayerMode)
{
for (uint i = 0; i < COUNTOF(player); ++i)
player[i].cash = 0;
player[0].items.ship = 11;
difficultyLevel++;
inputDevice[0] = 1;
inputDevice[1] = 2;
}
else if (richMode)
{
player[0].cash = 1000000;
}
else
{
ulong initial_cash[] = { 10000, 15000, 20000, 30000 };
assert(episodeNum >= 1 && episodeNum <= 4);
player[0].cash = initial_cash[episodeNum-1];
}
}
fadeIn = true;
break;
Language: C
bool select_difficulty( void )
{
JE_loadPic(VGAScreen, 2, false);
JE_dString(VGAScreen, JE_fontCenter(difficulty_name[0], FONT_SHAPES), 20, difficulty_name[0], FONT_SHAPES);
difficultyLevel = 2;
int difficulty_max = 3;
bool fade_in = true;
for (; ; )
{
for (int i = 1; i <= difficulty_max; i++)
{
JE_outTextAdjust(VGAScreen, JE_fontCenter(difficulty_name[i], SMALL_FONT_SHAPES), i * 24 + 30, difficulty_name[i], 15, -4 + (i == difficultyLevel ? 2 : 0), SMALL_FONT_SHAPES, true);
}
JE_showVGA();
if (fade_in)
{
fade_palette(colors, 10, 0, 255);
fade_in = false;
}
JE_word temp = 0;
JE_textMenuWait(&temp, false);
if (SDL_GetModState() & KMOD_SHIFT)
{
if ((difficulty_max < 4 && keysactive[SDLK_g]) ||
(difficulty_max == 4 && keysactive[SDLK_RIGHTBRACKET]))
{
difficulty_max++;
}
} else if (difficulty_max == 5 && keysactive[SDLK_l] && keysactive[SDLK_o] && keysactive[SDLK_r] && keysactive[SDLK_d]) {
difficulty_max++;
}
if (newkey)
{
switch (lastkey_sym)
{
case SDLK_UP:
difficultyLevel--;
if (difficultyLevel < 1)
{
difficultyLevel = difficulty_max;
}
JE_playSampleNum(S_CURSOR);
break;
case SDLK_DOWN:
difficultyLevel++;
if (difficultyLevel > difficulty_max)
{
difficultyLevel = 1;
}
JE_playSampleNum(S_CURSOR);
break;
case SDLK_RETURN:
JE_playSampleNum(S_SELECT);
/* fading handled elsewhere
fade_black(10); */
if (difficultyLevel == 6)
{
difficultyLevel = 8;
} else if (difficultyLevel == 5) {
difficultyLevel = 6;
}
return true;
case SDLK_ESCAPE:
JE_playSampleNum(S_SPRING);
/* fading handled elsewhere
fade_black(10); */
return false;
default:
break;
}
}
}
}
The moral of the story is that laziness always prevails.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
I've always wanted to do a playaround two-player arcade run with mouse input for Player 2 and keyboard input for Player 1. Too bad JPC-RR doesn't run Tyrian and that Hourglass doesn't support mouse input.
I don't like max% destroyed as a goal; the game has a ton of weapons to show off and in Arcade modes, you can already get Zica Laser and Constant Laser on a secret level accessible from the first level.
The official v2.1 data files have several boring glitches; most notably, the Charge Beam bug that can be used to gain orbiting balls of destruction much like the ones you get on that one Asteroid level, except that they destroy almost everything instantly. The data files provided on the OpenTyrian site don't feature this glitch.
Also, the very highest difficulties can't be accessed without using a predetermined/modified save; see OpenTyrian source for details.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
She runs in JPC-RR. She's real slow (about 1/8th of realtime), but she runs alright.
The Internet doesn't hold too much information on this beauty, so things like differences between the various versions (1.0, 1.1 and 1.2; 2.x are the Windows port) are largely unknown.
The game scrolls somewhat slowly, so it might not be the most entertaining game choice on the planet. Howewer, playing on the hardest difficulty and destroying everything could still be interesting.
I'm currently using v1.0 of the full version; the boot movie can be located here.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
If you don't finish the secret level by breaking the fuse, you won't get the best ending. I'd consider reaching the best ending to be an essential part of 100% completition.
Playing the level "normally" and then finishing by breaking the fuse is really hard to justify; I don't find the secret level entertaining enough for that to be an option. Check out the level layout yourself here.
EDIT: I'll also address franpa's woes about versions. The published glitched Keen 5 run uses the latest version because the feature/bug that allows warping to the secret level doesn't exist in previous versions. The improvement will also do so.
In the case of Keen 4, there aren't really any speedrun-relevant differences to speak of; v1.0 was used because it was the first one I happened to download.
For Keen 6: v1.0 has a lot more clipping errors than other versions, so it'll probably be the version of choice.
EDIT: THE FINAL EDIT: Found a way to manipulate the Shikadi Master. Sub-2:10 (with longer loading times than the published run and ending input after entering nick for high score screen) is the goal, and it seems possible. Too bad we may have to redo fucking everything.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
The Shikadi Master is very unpredictable. In this POS (enters nick for the high score screen, ends input at 2:12.096), his malevolence costs about three seconds. As such, the entirety of QED will probably need to be redone.
EDIT: It seems that Volte-Faces, Shockshunds and the Shikadi Master all share a common property: not being manipulatable by delaying level entry and firing the Neural Stunner. As such, the three lost seconds are probably here to stay. The last few seconds in the v0 are very unoptimal since it was rushed out to get an estimate of the probable final time. Also, I'll be gone for a while, so don't stay tuned.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
The glitched any% has been started and memory addresses have been found. Also, being rocked to the bone is more enjoyable than expected. Good times.
100% would seem to be less frustrating to TAS than in Keen 4 since there are no pesky oracles that can cost seconds and be hell to manipulate. There's also more variety between the any% and the 100%. Using the death flag glitch wouldn't seem be too entertaining in a 100% run, so I'm against using it.
Also, I'm using different emulator settings than Grunt, so the boot and loading times will be longer. Measuring the actual improvements is simple enough, so this won't be a problem.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
deuxhero wrote:
Any reason the donut isn't grabed at 3:26 given you have to wait for the platform anyways?
Score was completely disregarded throughout the run. I didn't even notice the donuts when redoing PoGA for the umpteenth time.
Keen Dreams, Keen 3
Improving the Keen 5 glitched any% is what I'm currently planning. I might take a look at the first trilogy sometime in the future and if I'm going to do any of them, it'll probably be Marooned on Mars.
Keen Dreams never pleased me too much; the lack of a pogo stick makes both playing and TASing the game a chore.
The best one to me is the Isle of Fire. It's hard enough not to die there, and the route you used seems so impossible unassisted.
The Berkeloid positioning achieved is indeed quite unlikely. When testing if the route was even possible by hanging from the ledge and letting the emulator run freely, the Berkeloid behaved as desired after around 80 seconds had passed.
I was extremely lucky when manipulating the Isle of Fire - optimal Berkeloid behaviour and excellent Oracle positioning was achieved with minimal delay.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
The pyramids have now been redone. In the slower run, I had an optimal oracle positioning roll in the Pyramid of Darkness — no such luck this time around.
PoGA is slightly faster and PoD is more than slightly slower, both due to oracle positioning which I manipulated to the best of my ability. As such, I'm now a little less than a second ahead of the old run.
Edit: Crystalus is now done. Due to the Continuous Vertical Pogo (which is also used in the new Miragia) being figured out, gobs of time are saved and everything looks extremely fabulous. Unfortunately, the trick only seems to work on static platforms. Oracle positioning is less than ideal, but delaying level entry for superior oracle positioning resulted in horrible enemy positioning that prevented several cornerboosts.
Final Edit: The Cave of the Descendents and the The Perilous Pit are now done. Both look extremely fabulous with good and near-optimal oracle positioning, respectively. The The Perilous Pit features a new shortcut that enables reaching the Oracle door within a single elevator cycle.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
Even if the DAR is correct, the MKV header will contain bonky values. While they may not be harmful, they look fugly and are just plain wrong.
I'll post the proper workaround here for those who might have missed it:
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
Lifewater Oasis has now been redone. The Isle of Fire jump turned out to be possible — I don't know what I was smoking.
Pyramids should be redone Soon™ with PoGA optimization being all but useless thanks to the shortcut. Maybe I'll try to find something entertaining to do in the upper part of the level.
Edit: The Pyramid of Darkness has now been redone.
Another edit: it appears that oracle positioning is random. As such, we're now redoing everything.
Yet another edit: Miragia improvements implemented, Oasis redone with very good oracle positioning rolls. EDIT: The improvements save around a second.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
The current draft now completes Miragia, Oasis, PoD and PoGA.
It is almost 15 seconds faster than the old WIP because of a new shortcut in PoGA discovered by Cosmo, superior optimization, reduced rampant stupidity, and greater :effort:.
Grab the encode here. Poke me on #tasvideos for the most recent draft.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
Cool Matty wrote:
Fansubbers, where 10-bit really started to be pushed, are even switching back due to these problems.
Cool story, bro. Care to cite any sources for this? Last I checked, pretty much every group that isn't a complete joke or using shareraws uses 10-bit in their primary (and usually only) release.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
Phallosvogel wrote:
I would love to see the encode but I was not able to play that movie back. I only get sound and a green image
hegyak wrote:
I can't even get the Disk Image to load in JPCRR. The MKV crashes VLC.
Read the news item about it on the front page. Then get the second WIP, now with 120% more Crystalus. If you desire the movie file, come slap me on #tasvideos.
Also, I'll be attempting to hex out the trainwreck at around 4:30 when I recover from post-slowdown stress disorder.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 6/17/2008
Posts: 146
EricS wrote:
But if you do that in CotD, then you haven't rescued the Council Member in that level and the game is now unwinnable.
That is indeed the case. I hadn't done adequate testing at the time of writing the post.
I don't have strong enough insults to convey how I feel about the The Perilous Pit.
Rejoice, for the moment of rebirth is at hand.