Editor, Experienced Forum User, Judge, Published Author, Expert player
(2070)
Joined: 8/21/2016
Posts: 1012
Location: US
Radiant wrote:
This page outlines an easy AI exploit in the fighting robots minigame.
The robots were what I was most worried about as far as RNG.
After reading the above article, I think I'll try both that method as well as actually fighting him to see which is fastest. Truthfully, with save states, outright fighting him may be fairly simple.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2070)
Joined: 8/21/2016
Posts: 1012
Location: US
boct1584 wrote:
I played it a bit when I was young. Never beat the harder difficulties.
Me neither, which is part of why I'm considering it.
Animation only happens roughly every 7 frames, so it's going to be a bit tedious to TAS. I have worked up through collecting the first set of artifact pieces and putting the puzzle together. it's just a slow process due to the animation rate.
I may bounce around between the three games just to keep my sanity.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2070)
Joined: 8/21/2016
Posts: 1012
Location: US
c-square wrote:
One quick point, you may want to replace the suggested screenshot with a one from this run that doesn't have the glitched text at the top.
Done
Also, on a similar note to potentially receiving a nomination for Speedy TAS of the year, would Radiant be qualified for Rookie of the year as this is his first submitted TAS?
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2070)
Joined: 8/21/2016
Posts: 1012
Location: US
fsvgm777 wrote:
Page 2 clearly displays a joystick (but that's already established). However, it only allows for 8 directional inputs. Therefore, the input in this case is semi-analog (as is the case with many other Arcade games), like in the Atari 2600 controller.
It's fully digital. The 8 directions are achieved through 4 switches and combinations of the signals coming from them just like in an NES controller. The only difference is that in the NES controller, the 'wires' are the electrical traces attached to the internal circuit board, while the arcade cabinet uses actual copper wires. Gluing a thumbstick to the top of an NES controller wouldn't magically make it analog anymore than arcade joysticks make their internal switches analog.
I've had multiple jobs through college repairing arcade machines. The VAST majority of older games (and many) newer ones use on/off switches (either leaf or button style) underneath the joystick and are thus digital input, not analog. There were a few analog controlled games, but they were rare in my experience.
This means an arcade joystick is essentially the same as an NES control pad. Where the only limitation to pressing U+D/L+R on the NES is the rigid D-Pad on the stock NES controller, the only limitation to pressing U+D/L+R on the arcade cabinet is the rigid joystick structure.
If NES TASes are accepted because these multidirectional inputs can be physically performed on non-stock controllers, the argument could be made that modifying a cabinet's controls to accept these inputs would be that same as using an aftermarket/non-official controller.
To further support the side of accepting these inputs on arcade machine TASes... While I was working in those repair jobs, I took the opportunity to press the switches for opposite directions simultaneously during internal maintenance on various cabinets. I can confirm that this would indeed glitch some games, in some cases requiring a complete reboot of the unit to return the game to normal function.
EDIT: My apologies to Ferret Warlord, I was in the midst of typing my response when you posted yours saying essentially the same thing in a much more concise manner.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2070)
Joined: 8/21/2016
Posts: 1012
Location: US
I was able to edit the save file to eliminate the sound during the self destruct keypad sequence. I've updated the video links above.
I'll submit the final video a bit later tonight.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2070)
Joined: 8/21/2016
Posts: 1012
Location: US
c-square wrote:
That's completely up to you. I'm glad to have helped, but you're the one who ran the run in the end. I'm happy to have my name added to the list, but I won't be offended if it isn't. To submit with multiple authors, just put commas or ampersands in the nickname field to separate the authors.
Just because you didn't directly work on the final TAS movie file, doesn't mean your work wasn't included. Without the input from you and Radiant on the various randomization calculations and Radiant's understanding/clarifications of the AGI system, I would never have accomplished this result.
I was asking to both be polite and to give you two the opportunity to actively request not to be included in authorship, if you so desired. But my opinion is that you've both earned a place.
I'll wait to hear from Radiant before publishing.
On a different and completely unnecessary note...
Hypothetically, if Radiant were to do a TAS of one of his own games, would it be deemed unfair? Because theoretically, he could just tweak a few files here and there to make a TAS extremely easy and then claim it a new version of the game with which he could then perform the TAS.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2070)
Joined: 8/21/2016
Posts: 1012
Location: US
Radiant wrote:
SQ contains an internal clock that few people know about (because it's not in the menu for some reason). It may be interesting to show this in the speedrun? Or perhaps you find it distracting. To try it out, on any screen type "clock"; or if you want to be fancy about it, "First, please clock me in, sir." Since this doesn't affect gameplay it may be possible to hex it into an early room.
I tried to manipulate the save-state. I was able to get the command squeezed in after getting the keycard on the Arcada. However, even though I was able to get the game to sync through opening the bay doors, it desynced after that room.
I don't think it' s worth the effort to figure out all the adjustments to add it in.
ANYWAY: I've finished the runs. Here's the temp encodes for full speed and 1/5th speed. BEWARE! The sound during the screen with the code panel on the Start Generator is REALLY ANNOYING; mute or skip this LONG section.
Normal:
Link to video
1/5 Speed:
Link to video
Now then...who all wants/gets authorship credit? I was thinking myself, c-square, and Radiant.
To follow up that question...how do I submit with multiple authors?
EDIT: Updated video links with annoying alarm sound disabled.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2070)
Joined: 8/21/2016
Posts: 1012
Location: US
Start: 30 Bz
1) Bet 1 -> EDE 29 Bz
2) Bet 1 -> SCE 28 Bz
3) Y- Enter
4) Bet 3 -> EEE 55 Bz
5) Y- Enter
6) Bet 3 -> EEE 82 Bz
7) Y-Enter x 3
8) Bet 3 -> EEE 109 Bz
9) Y-Enter x 5
10) Bet 3 -> EEE 136 Bz
11) Bet 3 -> CCD 148 Bz
12) Y-Enter
13) Bet 3 -> DDD 205 Bz
14) Bet 1 -> SDS 204 Bz
15) Bet 2 -> CDS 204 Bz
16) Y-Enter
17) Bet 3 -> DDD 261 Bz
18) Machine Broken
END FRAME =
EDIT: Actually, we can set it up so that you get to 250 but not actually break the machine, meaning the text pop-up won't appear. I don't know if that will save anything, but may be worth checking out:
Start: 30 Bz
1) Bet 1 -> EDE 29 Bz
2) Bet 1 -> SCE 28 Bz
3) Y- Enter
4) Bet 3 -> EEE 55 Bz
5) Y- Enter
6) Bet 3 -> EEE 82 Bz
7) Y-Enter x 3
8) Bet 3 -> EEE 109 Bz
9) Y-Enter x 5
10) Bet 3 -> EEE 136 Bz
11) Bet 1 -> CCD 140 Bz
12) Y-Enter
13) Bet 3 -> DDD 197 Bz
14) Bet 1 -> SDS 196 Bz
15) Bet 2 -> CDS 196 Bz
16) Y-Enter
17) Bet 3 -> DDD 253 Bz
18) F10 Exit
Both of these will break the machine. Anything greater than 249 will.
The only way to not break the machine and get enough cash for the purchases is to get between 240-249 on the machine then pick up the buckazoids from the droids-b-us building or the bartender. With our current seed order, neither extra pickup is faster than just breaking the bank.
Both of the above should take the same amount of time. I compared the first to my old sequence and the top one is 3 frames faster. I may be able to shave off another frame or two with better timing on the inputs, but I think this one is the best we've found.
Start: 30 Bz
1) Bet 1 -> EDE 29 Bz
2) Bet 1 -> SCE 28 Bz
3) Y- Enter
4) Bet 3 -> EEE 55 Bz
5) Y- Enter
6) Bet 3 -> EEE 82 Bz
7) Y-Enter x 3
8) Bet 3 -> EEE 109 Bz
9) Y-Enter x 5
10) Bet 3 -> EEE 136 Bz
11) Bet 3 -> CCD 148 Bz
12) Y-Enter
13) Bet 3 -> DDD 205 Bz
14) Bet 1 -> SDS 204 Bz
15) Bet 2 -> CDS 204 Bz
16) Y-Enter
17) Bet 3 -> DDD 261 Bz
18) Machine Broken
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2070)
Joined: 8/21/2016
Posts: 1012
Location: US
Radiant wrote:
SQ contains an internal clock that few people know about (because it's not in the menu for some reason). It may be interesting to show this in the speedrun? Or perhaps you find it distracting. To try it out, on any screen type "clock"; or if you want to be fancy about it, "First, please clock me in, sir." Since this doesn't affect gameplay it may be possible to hex it into an early room.
It would take some fancy save-state editing, but I think it would be possible.
Hoping not to test every key, is there a list of key IDs for JPC-rr? For example, 'Enter' is key 28, I believe.
The only concern I have with displaying the internal clock is that it may run faster than actual time, making it look like the playthrough has been artificially sped up.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2070)
Joined: 8/21/2016
Posts: 1012
Location: US
c-square wrote:
DrD2k9 wrote:
Ok, I just set JPC-rr to run instead of doing frame advance.
I followed your sequence, and here's what I got.
My apologies. I was doing this after midnight last night and made some mistakes.
Please give this a try:
1) Bet 1 -> Eye | Diamond | Eye
2) Bet 1 -> Skull | Cherry | Eye
3) Y-Enter
4) Bet 3 -> Eye | Eye | Eye
5) Y-Enter
6) Bet 3 -> Eye | Eye | Eye
7) Y-Enter
8) Bet 3 -> Cherry | Cherry | Cherry
9) Bet 2 -> Cherry | Eye| Cherry
10) Y-Enter
11) Bet 3 -> Eye | Eye | Eye
12) Bet 3 -> Cherry | Cherry | Diamond
13) Y-Enter
14) Bet 3 -> Diamond | Diamond | Diamond
15) Bet 1 -> Skull | Diamond | Skull
16) Bet 2 -> Cherry | Diamond | Skull
17) Y-Enter
18) Bet 3 -> Diamond | Diamond | Diamond
No problem. Fatigue hinders everyone's productivity.
And not to be difficult, but that only yields 246 buckazoids. It takes another 3 spins to get the needed 250.
The last $4 could come from the coupon with the barkeeper or from picking up the few lying behind the droids r us building (probably the least efficient).
Ii'll test and see which is faster between the extra 3 spins or redeeming the coupon at the bar.
Some estimates before testing...the coupon will require at least 2 text boxes (8-10 frames) in addition to the frames to get to the bar, then a few more waiting for the barkeeper to get to Roger. This would be probably a minimum total of 20-25ish frames, but wouldn't need the text box for breaking the slot machine.
The 3 extra spins would require 15-21 frames for spins and another 4-5 for the text box. So..it could be close. I'll test them as soon as I am able....which won't be before tomorrow.
