Experienced Forum User, Judge, Published Author, Skilled player
(1279)
Joined: 9/12/2016
Posts: 1645
Location: Italy
I don't think this to be necessary. If someone wants to be grouped with a particular person, then should say it clearly before the deadline, otherwise I'll assume is giving me full control over the team decision.
Also, I think there will be plenty of time for people to say hi and stuff.
That's right, I am the only person behind the game choice, and I attest that the game itself is awesome; it's way better than any game were chosen in prior DTCs until now. I'm sure you will like it.
Experienced Forum User, Judge, Published Author, Skilled player
(1279)
Joined: 9/12/2016
Posts: 1645
Location: Italy
Yeah, I understand that. I should really take it easier.
Oh no way! I would rather make all my Crash TASing knowledge public instead of submitting a run that I'm aware to be suboptimal; maybe I could make exception for very very hard optimizations tough.
By the way, I'll write about all the tricks I've discovered in the submission text, when the time will come.
Experienced Forum User, Judge, Published Author, Skilled player
(1279)
Joined: 9/12/2016
Posts: 1645
Location: Italy
I'm recently experiencing a lot of boredom in optimizing Crash for GBA. TASing new levels was very exciting, but now that I am getting back at levels that I thought were already perfectly optimized, and discover that they're not, it's really frustrating, and thus TASing has become more like a job from my point of view. I really don't want this to happen, because it completely kills the main (but not only) purpose that makes me to TAS, which is to have fun.
What do you think about it? Is TASing a serious business for you or you only to it for fun? Also, what do you suggest me to do in order to make working on this run more enjoyable?
Edit: basically, my problem is that I want to both have fun and to optimize seriously my run.
Experienced Forum User, Judge, Published Author, Skilled player
(1279)
Joined: 9/12/2016
Posts: 1645
Location: Italy
I was waiting for this! And the run itself is magnificent! Absolutely Yes vote!
Edit: Maybe I'm overly entusiastic, but I simply happen to really like this game.
Experienced Forum User, Judge, Published Author, Skilled player
(1279)
Joined: 9/12/2016
Posts: 1645
Location: Italy
Absolutely no, for these reasons:
1) The contest must not clash with AGDQ
2) Everyone at AGDQ will need their time to travel back to home
3) Someone may need a second Sunday to get enough free time to form a team and sign-up
4) The contest will last for 3 months, so the signing-up time should be proportional.
Experienced Forum User, Judge, Published Author, Skilled player
(1279)
Joined: 9/12/2016
Posts: 1645
Location: Italy
mGBA emulator is available for both OS you mentioned, so anyone who doesn't have Windows can still help by playing the game without using TASing tools.
Edit: I mean, using a non-TASing emulator can make your able to study routes, tricks, memory addresses, etc. but you still need at least one team member that is able to use BizHawk 1.11.9.1 and create the movie file to submit.
Experienced Forum User, Judge, Published Author, Skilled player
(1279)
Joined: 9/12/2016
Posts: 1645
Location: Italy
I'm proud to announce the 6th Dream Team Contest!
This is a TASing contest that takes place between teams that compete to create a TAS that completes the given game as fast as possible. Anyone can join by sending a post to this topic. The winning team will be engraved in Dream Team Contest history.
Game info
The game name will be revealed when the signing up time expires but some details are available now:
The game platform is Game Boy Advance.
The game genre is platformer.
The estimated TAS length is about 25 minutes.
Rules
You sign up as a 4-people team or you can request to be put in a team; you may optionally also provide a team name and a logo; after contest start, teams can't be changed.
On 21 Jan 23:59 UTC signing up time expires, game name will be revealed and the contest will start.
The emulator to use will be BizHawk 1.11.9.1 (or newer) using the mGBA core.
The teams must provide their movie file to dreamteamcontest6@gmail.com before 22 Apr 23:59 UTC; improvements of the run can be still submitted within the deadline; the deadline can be anticipated if all teams agree.
In the case an Arbitrary Code Execution/Memory Corruption/SRAM Gltich/Game End Glitch/Wrong Warp Glitch is discovered, competition will be held for both exploiting such a glitch as well as not using it.
The team that provides the movie file that completes the game from power-on to credits in the fastest time (input end) wins.
The winning team will submit its movie to TASVideos and anyone who will contribute will be added to the authors list.
Here are the current sign-ups. Note that individual player sign-ups have time until contest start (21 Jan 23:59 UTC) to decide or change their team; if they don't, they will randomly shuffled in 4-people teams; 3 or 2 if the players amount is not divisible for 4.
Remember that team members, name and logo can be changed within the signing-up deadline.
Team sign-ups
* team members arbitrarily chosen by ThunderAxe31.
Individual player sign-upsThere are no invididual player sign-ups left.Sign-ups are closed. Teams which haven't still provided a name and a logo can still do so, but then it won't be changed or removed.
The game
Zook Man ZX4 is an unlicensed GBA game by Fast Fame. Most game content is taken from Mega Man 8 and Mega Man X, while some enemies come from previous Zook Hero games. Most musics are remixes of songs found in Soul Falchion, although there are new original compositions made for the game.
This game is fresh for TASing because of its past emulation issues. The currently last released version of BizHawk is literally the only TASing emulator able to run the game.
The game can be played without any known problem thanks to mGBA 0.5 core, allowing to beat the game and trigger the credits.
Checksums for the GBA ROM file
Controls
I decided to make the controls clear for people less familiar with Mega Man games, in order to avoid frustrating scenarios. I'm not going to give any more help.
A - Jump
A (on walls) - Climb wall
B - Shoot
B (keep) - Charge shoot
L or R - Change between available weapons
Left, Left (keep) or Right, Right (keep) - Dash (needs to be unlocked in a certain level)
Start - Open menu
Tips
Divide the tasks.
Everyone has own strengths and weaknesses. Make your team members work on different jobs in order to improve efficiency. This is crucial.
Keep in contact.
Use a private and secure chat group of your choice to talk with your teammates about your progress, discoveries, ideas, and share your files.
Play it a lot first.
Game mechanics are easier to understand by playing the game casually, with some code cheating if necessary. Start TASing when you feel that you know enough.
Take note of everything.
Any trick and glitch may come useful in a way or another, even if it does initially not seem. Also remember that your discoveries may come useful after the contest.
Try anything.
Even after you've finished TASing the game, use the remaining time to try and improve the run. A sudden discovery can turn the tide of the contest.
Don't give up.
I know this is going to be longest TASing contest until now, but remember that the journey, not the arrival, matters.
Experienced Forum User, Judge, Published Author, Skilled player
(1279)
Joined: 9/12/2016
Posts: 1645
Location: Italy
PikachuMan wrote:
2 Player Co-Op TASes on Game Boy is best to encode in 3D.
I'm not entirely sure about what do you mean but... Here is how I think it's best to encode it:
Link to video
Note that this is just a bad test. I had no clue the Wand of Uzi would deal half damage to bosses.
Experienced Forum User, Judge, Published Author, Skilled player
(1279)
Joined: 9/12/2016
Posts: 1645
Location: Italy
SmartFrame wrote:
since it would allow to make direct comparisons with RTA runs.
i wonder about that, RTA runs nowadays use that "super jump" i think it´s called, to skip major parts of the dungeons where you´re originally not supposed to go (e.g. skip small keys etc.. )
Then skipping ditches should be allowed.
You know, there is a SM64 run published that specifically don't uses the Backwards Long Jump glitch, so I think it should be ok to take an ad-hoc set of rules in order to create the most entertaining run possible.
Also, with the current game knowledge it should be also possible to improve the current record of the "warp glitch" brach.
Experienced Forum User, Judge, Published Author, Skilled player
(1279)
Joined: 9/12/2016
Posts: 1645
Location: Italy
Tompa wrote:
Yes, the "All dungeons" TAS is very outdated for this, and other, reasons. A new run would improve it by quite a lot. One of the problems, the same reason why I also quit TASing A Link to the Past, is a logical ruleset with defining what glitches that should be allowed or not... It's tricky.
I think it should go with not using glitches that allows walking throught walls, closed doors, or simply getting to places you're not normally supposed to be able to go to. I think it would result in both entertaining and technically interesting, since it would allow to make direct comparisons with RTA runs.
Also, no arbitrary code execution, game end glitch, memory corruption, SRAM glitch, or any other heavy game-breaking glitches possible.
Experienced Forum User, Judge, Published Author, Skilled player
(1279)
Joined: 9/12/2016
Posts: 1645
Location: Italy
E Mo Dao, also know as Devil Island or Devil Land, is an unlicensed GBC game developed by Vast Fame. (more)
The game itself is a copy of Getsu Fuuma Den for NES, with different graphics, musics, and without the pseudo-3D-view dungeons.
There is an overworld seen-from-above map that have to be travelled in order to get to the final boss. Along the way, there are many minor crossroads, but the game is actually very linear: in order to proceed to the next map, you're forced to beat several bosses in a set order.
The route is the same as for Getsu Fuuma Den, with the only expection that levels 30 and 49 are swapped: map
The gameplay in the levels may look similar to Castelvania series. You have to travel from left to right (or to right to left, depending on the case) thought the level in order to get past it on the map. The levels are packed with many kind of monsters, that when killed may drop a soul, that when picked restores 5 HP, or a "bag of dollars", literally. Every monster has its own movement pattern, health, dollars drop amount, exp gain amount, and some throw projectiles at you.
Like many unlicensed games, avoiding getting hit is often nearly impossible, due to the poor game development, but on the other hand you can easly get health drops from enemies.
You start the game with a whip as weapon, that can hit from very low distance. With money, is possible to buy other weapons and items to the shops. After beating the last but one boss, you get a powerful whip that fires a sort of air blade that can hit from any distance. Weapons and Items must be set on the pause screen into a level, in order to be used. Damage of whip weapons increases with experience.
Here is the list, with relative cost/s. Note that names are arbitrary:
Consumables Items
Gauntlet (S1/S6:150$, 300$): Doubles any weapon damage until you exit from the level.
Dragon Ally (S8:400$): kills every enemy on screen. deals 20 damage to bosses. can't be doubled by using gauntlet. Takes about 293 frames.
White Jug (S3:150$, S2:200$): Grants invulnerability for 12 seconds. No, Shops names are not swapped.
Armor (S4/S5:300$, S0:600$): Grants invulnerability for 12 seconds.
Red Potion (S1:100$, S2/S3:150$, S4/S6/S8/S9:200$, S7:400$, S0:500$): Restores some HP.
Hourglass (S5:100$, S9:350$): Stops enemy movements and spawning.
Crucifix (S3:200$, 300$): currently unknown
Weapons Items
Regular Flail: You start with this.
Fireball (S2:200$, 250$): Throws a projectile that pierces throught enemies and deals 1 damage for tick
Rock Flail (S1:200$): Same as regular weapon, but can also break Brick Blocks. Getting this is mandatory.
Kunai (S0:600$): Broken weapon that can deal up to 1 damage about every 2 frames if close enough to the enemy. Can kill Bugman bosses in about 2,5 seconds. Damage does not increase with experience.
Ultimate Flail: The only weapon that reasonably damages the final boss. Deals 16 HP damages, also fires a projectile. Damage can be doubled to 32 with Gauntlet.
Passive Items
Shield (S7:350$): Halves damage to player.
Winged Shoe (400$): Allows to hit enemies by jumping, orizontal velocity is slower when jumping.
Candle (S4:150$, S7:350$): currently unknown
Unlockables
Demon Pass: You get this when you beat a Bugman boss on "P" named levels. Needed for entering gate levels.
Ultimate Flail Piece: You get this when you beat a boss on "D" named levels. Need all 3 to unlock Ultimate Flail and access final boss battle.
Here are the Shops list:
S1: Red Potion (100$), Rock Flail (200$), Gauntlet (150$)
S2: Red Potion (150$), White Jug (200$), Fireball (200$)
S3: Red Potion (150$), Crucifix (200$), White Jug (150$)
S4: Candle (150$), Red Potion (200$), Armor (300$)
S5: Hourglass (100$), Fireball (250$), Armor (300$)
S6: Red Potion (200$), Crucifix (300$), Gauntlet (150$)
S7: Shield (350$), Red Potion (400$), Candle (350$)
S8: Red Potion (200$), Dragon Ally (400$), Gauntlet (300$)
S9: Winged Shoe (400$), Red Potion (200$), Hourglass (350$)
S0: Kunai (600$), Armor (600$), Red Potion (500$)
The game has a simple experience system that determines damage output of Regular Flail and Rock Flail. The experience amount is a Word value that start at 1024, and grows with every enemy kill. Here is the table:
Damage | Experience required
1 | 1024 - 3071
2 | 3072 - 6143
3 | 6144 - 8191
4 | 8192 - 10239
5 | 10240-12287
6 | 12288-14335
7 | 14336-16383
8 | 16384
Known glitches:
Quicksave Abuse
This is technically not a glitch, but more like a development overlook. Involves pressing Select to quicksave the game right before entering the levels, and then hard resetting. This effectively allows you to skip most levels. All credits go to Mothrayas for discovering this.
Hurted Mid-air
When the hero gets hit while in mid-air, he won't get recoil and will be able to attack much faster with Flail weapons. Can be used to hit faster regular enemies and Bugman bosses.
Landing Whipping
When the hero land in the right moment, the Whipping animation will be cut and get able to whip again faster.
Drop Manipulation
The enemy drops simply depends to the moment they are killed.
Level Entering
Not sure about its possible uses, but I have to note that right before the spinning animation before entering a stage on the overworld, the game thinks for a moment that you already are in a level. So it's possible to access the level pause menu if Start is pressed in the right moment, but I didn't found anything useful in this. If you set a consumable item in this situation, you will lose it without getting its effect during the level you're entering.
So that was the theory. Here is the route I'm planning for my run:
-go to level 01 and farm 400$ with the 3 centipedes, then exit on the right
-beat boss at area P1; take damage to save time
-go to shop S1 and buy everything
-travel through gate 13; avoid any damage
-enter level 16 by the left and farm 250$, full health, and get to ~3000 total experience
-go to stairs D1; set Rock Flail and Gautlet; get to 3072+ experience before facing the boss
-go back through 13
-go to shop S6 and buy Gauntlet
-go to shop S5 and buy Hourglass
-travel through gate 17, 18, 19, 20, 21; take damage to save time if necessary
-beat boss at area P2
-get back to 19 and travel through 22, 23, 24, 25, 26
-...to be continued
There may be some changes for the part after clearing D1, but it's probably already fine. I have to study a bit the next levels enemies before confirming.
Here is the current WIP encode:
Link to video
Here the .bk2 file: http://tasvideos.org/userfiles/download/35994737823374091
A quick lua script to help lag management: http://pastebin.com/2QfTpsFe]Movement
And memory addresses watch .wch file: http://pastebin.com/iLrASh1y
Experienced Forum User, Judge, Published Author, Skilled player
(1279)
Joined: 9/12/2016
Posts: 1645
Location: Italy
jlun2 wrote:
Given that you cancelled it, I suppose this means you failed to find a way to make the post-game contents work properly?
Yes. I need to study the game code seriously before being able to make it. Also, I've come to the conclusion that the resulting movie would be noticeably different from the one in this submission, so it's the case to cancel and start over from scratch.
Experienced Forum User, Judge, Published Author, Skilled player
(1279)
Joined: 9/12/2016
Posts: 1645
Location: Italy
Ramzi wrote:
And some day and AI will be able to make those personal artistic choices.
If you let a machine take that sort of choices in your place, then life stops having meaning.
If for example, someday there would be created an AI than can do perfect TASes, then human TASing would lose much meaning.
One of the good things abot TASing is that you can't never be absolutly certain that a run can't beaten. That's what makes you give it all to desperately search for a way to beat current records.
Experienced Forum User, Judge, Published Author, Skilled player
(1279)
Joined: 9/12/2016
Posts: 1645
Location: Italy
Ramzi wrote:
If a human can do it, and AI eventually can. Maybe one day we'll see such gorgeous remakes made easily.
Maybe it is possible to create a batter depixelizing algorithm, but that would still not be enought to remasterize a game. Remastering games need arbitrary choices that can't be automatized in any way. It needs a lot of personal choices and artistic sensibilities... that's what the "remastering" word itself means. That's why remastered games needs so much work in order to be made.
Experienced Forum User, Judge, Published Author, Skilled player
(1279)
Joined: 9/12/2016
Posts: 1645
Location: Italy
feos wrote:
Putting bitmaps on top of background takes little processing power. If the depixelized sprites are indeed taken from the ROM and saved separately, this is pretty doable (with some code tweaks probably):
Ok, it's not impossible, but it would still require a lot of time because every game's sprites would have to be processed differently. Using a single method for rendering any game would be ineffective, not to mention that the render mothod(s) would be very arbitrary.
Also, emulators used for TASes are built in order to get emulation accuracy, in order to be the most faithful to the actual platform, so this topic has probably little to do with TASing.
Ramzi wrote: