Difficult or not, it shouldn't take almost 4 months to publish a 20 minute long run. I mean how is it that Muramasa, an hour long Wii run, got published in 3 weeks?
Can you give an example of memory.registerwrite being used? I tried it and got an error "bad argument #3 to 'registerwrite' (function expected, got no value)". In the documentation, it says:
memory.registerwrite (address, [size,] [cpuname,] func)
Calls the function whenever the given address is written to. Function can be nil.
I tried passing a simple test function that simply returns the value 8, but got the same error message. What exactly are you supposed to do with that 2nd parameter?
Why would you publish a run with 2 yes votes and 2 no votes? In addition, the only feedback for the run are highly critical posts by Tangent. So not only is there clearly insufficient feedback on which to base a decision, the existing feedback is negative enough to seriously doubt whether this run is publication worthy.
local buttons = {}
while true do
buttons = joypad.get(1)
xspd = 0x7E007B
yspd = 0x7E007D
rxspd = memory.readbyte(xspd)
ryspd = memory.readbyte(yspd)
flight_charge = 0x7E13E4
gui.text(190,40,"horz spd: "..rxspd)
gui.text(190,50,"vert spd: "..ryspd)
joypad.set(1,{Y=1})
memory.writebyte(flight_charge,112)
if rxspd >= 0 and rxspd <30>= 0 and ryspd <= 6 then
if buttons.right then
memory.writebyte(xspd,50)
end
if buttons.left then
memory.writebyte(xspd,-50)
end
end
if buttons.down then
memory.writebyte(xspd,50)
memory.writebyte(yspd,56)
end
if buttons.up then
memory.writebyte(xspd,50)
memory.writebyte(yspd,-56)
end
snes9x.frameadvance()
end
There's still some annoyances, like dying whenever you fly too far into a wall, which makes water levels nearly impossible to beat with these controls.
*Note that line 19 of the code isn't coming out correctly :( rxspd is less than or equal to 30 and ryspd is greater than equal to 0.
There's a lot of simple hacking (eg/ infinite hp, lv 99, max gold) you can do just by using RAM search and changing memory values. But what if you want to do something more elaborate, like messing with your character's acceleration? That's where lua comes in.
To illustrate, I had some fun messing around with Super Mario World. This simple lua script makes Mario go at full speed the frame you press left or right. Mario also floats upwards at full speed by holding B, and will fall just as fast by pressing down. It's a lot of fun to try out and there's some hilarious unintended side-effects, such as dying when touching a green ! block.
Download SMWCheat.lua
Language: lua
local buttons = {}
while true do
buttons = joypad.get(1)
xspd = 0x7E007B
yspd = 0x7E007D
rxspd = memory.readbyte(xspd)
ryspd = memory.readbyte(yspd)
gui.text(190,40,"horz spd: "..rxspd)
gui.text(190,50,"vert spd: "..ryspd)
if buttons.right then
memory.writebyte(xspd,50)
end
if buttons.left then
memory.writebyte(xspd,-50)
end
if buttons.B then
memory.writebyte(yspd,-56)
end
if buttons.down then
memory.writebyte(yspd,56)
end
snes9x.frameadvance()
end
Since I'm not very proficient in lua, I'm hoping other members will get on board with this idea, since there's tiny gameplay hacks that can make games dramatically more enjoyable to play. Some ideas off the top of my head:
Super Mario World:
-Able to fly with the cape without charging (ie running until you build enough speed)
-Increase the rate at which fireballs are released
-Press a button to make a P-switch fall out of the sky
-Make a duplicate of every enemy on-screen
Super Metroid:
-Instant shinespark charge
-Shinesparks don't cost hp: In fact, maybe they even increase hp!
-Double Samus' natural jumping height
-Beam shoots backwards (Samus must face away from the enemy to hit)
-Power bombs activate without the long explode animation
-A simple morph ball bomb sends Samus flying
-Samus' ship appears in the boss room after the boss is defeated
-Shoot missiles and super missiles simultaneously
Most of those ideas involve making the game easier, but of course you can manipulate the game into becoming harder. For instance, making the main character die in 1 hit.
I think it would be really fun, both from a coding standpoint (better understanding ram search and lua), and obviously the end result of a game on steroids. For now I'd put emphasis on lua scripts that could apply to a broad range of games, such as making the main character move faster.
Also, Saturn, when are you going to submit the Super Metroid Redesign run? I'm looking forward to it.
There is no decision yet. That run has a couple of small optimization errors in the early Crateria parts resulted from rushing and not testing all variations at that time, so I probably shouldn't bother until an improved version, if it gets done. Having a probability that your hard work may not get accepted as it is without manipulations isn't a good motivation factor to make submissions either.
You haven't yet decided if you want to submit a run you completed 2 years ago?!?
Nobody's going to vote no over small optimization errors early in the run, that happens all the time and there's nothing runners can do about it unless the game is very hex friendly. If people were truly bothered by that, we'd never see a SM64 or N64 Zelda run published, EVER.
I don't really get your logic about not submitting. Even if it gets rejected (it won't), you still dramatically increase the audience for your work, which must count for something given that run must've taken at least months to complete. And if the run is rejected, it will be over factors out of your control, like the inherent hatred most hacks get here. Even then I wouldn't worry too much, as the site's been more lenient with hack publications - look at SM Zero Mission any% and 100%. The shinesparking exploits in your Redesign TAS alone make it more entertaining than any SM hack the site's ever seen.
Now even if you STILL don't want to submit, you should at least post the smv somewhere so that people aren't forced to watch an 11-part 30fps encode to appreciate your TAS.
My reaction to seeing this in the workbench:
NO WAY! NO WAY! NO WAY!
Now time to actually watch the run :)
Edit: Watch the TAS streamed here. It might be twice as long (~4 hours) because of all of Trihex's replays, but it also dramatically increases your appreciation for the run with all the slowdown and commentary going on.
Also check out SMW2: Kamek's Revenge TASes. Carl said he might start working on these again at the end of Trihex's stream :)
So, it goes without saying, this is probably one of the only TASes that actually deserves a 10/10 entertainment and technical rating. A star doesn't do this run justice.
How did you get those weird camera angles with Ashley? Is that a J version thing? Also, the video had a pretty noticeable sound desynch by the end. Otherwise, keep up the good work. The dreaded water room was amazing :)
Allowing both RBA, while forcing the runner to get the dungeon stuff makes sense to us, but it will certainly not even make a bit of sense for the normal viewer. We'd suddenly get the Map in Shadow normally, but nowhere else etc.
Can you explain this in further detail? If you have to get the dungeon items legit, why would only the shadow map be acquired?
Again, it's better if 100% is defined as actually getting the items legit as opposed to glitching out your inventory for the same effect. This ensures a large portion of the game will still be played regardless of how many glitches get discovered over time, and that's what people want to see.
I think it would be cool to have 3 categories for this game: Glitched any% (ie the 8:13 run by Tompa), any% and 100%. Would basically follow the same logic as the glitched any% and any% Super Metroid categories.
Basically I think it's worth seeing a run that completes all stages without also doing all the bonus barrels, and I don't think it's too arbitrary to separate categories based on whether a game breaking level skipping glitch is used or not.
No, the problem with using savestates is that you can make unnoticeable changes to the game by messing with the game's memory. Not that I think it's likely, but without a verification movie, the savestate could have been hacked to make T.T 1% faster, or change hot top volcano's hitboxes, or any number of things. Conversely, with the verification movie, as long as the movie syncs you can be 100% sure there wasn't any form of ROM tampering involved.
I think the site rules make this pretty clear.
We do not allow save-anchored movies.
We want a standard starting point for movies (power-on). Saves introduce an infinite amount of possible variation that may cause the game to behave differently compared to starting from power-on. They can also can be hacked, allowing nearly transparent cheating.
However, there are certain games with unlockable modes, second quests, or other things of interest that can only be accessed if a save file (or an otherwise "dirty" SRAM) is present. If you really wish to submit a movie made on such a mode, you will need a verification movie made and provided alongside it. Any input file that starts from power-on (for example, a previously submitted movie for that game) and creates the exact circumstances for your submission to sync will generally do.
I wouldn't change my vote over this (voted yes earlier) since I've seen glitched any% streamed on realtime and highly doubt the save Swordless used was tampered with (intentionally or not). But this should never happen again and is setting a really bad precedent if the site stops caring about verification movies.
The reason I didn't provide the other movie file is because I didn't feel like making one when I can easily just grab a savefile that's already available that has the necessary criteria for the run already met. It's just a waste of my time and I doubt it'd be rejected for that, anyway, especially since Nahoc knows this game well and has already stated in IRC that he knows the savefile is legit.
Cabin fight start- 16:52
Cabin fight snipe fest- 17:46
Mendez fight- 21:45
Looking good so far :) Cabin fight was a highlight obviously. Two things though. One, does grenade tossing in the cabin really save time compared to just sniping from the beginning? The other thing, the lift was terrible, like bad enough to make the run unpublishable. What I was expecting was you to snipe all the enemies from the start (except maybe the enemies who jump at you), then waste away your remaining bullets aiming straight ahead, or something like that. I don’t get why you spent so much time aiming at certain enemies before firing, but it looks really sloppy imo.