Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
rythin wrote:
I've always felt like Standard was missing a goal for something along the lines of "forgoes the use of one common glitch", where the movie can still feature many glitches of various severity, but omits one particular one that would otherwise be repeated throughout the movie.
#7874: GMP's GC Prince of Persia: Warrior Within "zipless" in 24:56.15 is a great example of this in recent memory, though I'm sure there's many others that also fit the description.
How do we decide which glitch can/should be avoided? In the case of "zipless" movies, they all fit very well into Alternative.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
InputEvelution wrote:
While not necessarily applicable to these two TASes, I feel that it's worth talking about the limitations of off-screen mouse input in the official Flash Player, as this seems to have been the main argument used for allowing it. While you can indeed click things before they appear within the regular bounds of the screen in Adobe's offline Flash Player, you can't really do so along diagonals. Expanding the window in two dimensions at once merely causes Flash to expand things up to fill in the gap. So while it's possible to click things that are off-screen in one dimension, it's not possible to do so for things off-screen in both at once. It's also not possible to see anything beyond the normal bounds, at least in the games I tested.
It's optional.
InputEvelution wrote:
That being said, I am still not against the idea of having these as two separate branches. While I can't say I find its usage here to fall under the category of a "major skip glitch"...I would prefer it if TASers had the option to submit movies that specifically choose not to exploit this kind of thing. It feels...cheap? Unintended? Unentertaining? It's hard for me to pinpoint anything in my gut reaction that makes for a coherent point here, and it seems the few of those who have participated in discussions about this agree on accepting it, but I'd very much prefer to be able to keep making TASes that choose not to use it, without such things being treated as an "optimisation error" or "not in an entertaining enough movie". Frankly, the most I've been able to see myself using this is choosing a female worker in the Papa's games where a male worker would otherwise be faster to select. Perhaps L+R/U+D is worth considering in this discussion as well, if it's considered a rather similar issue to this.
rythin wrote:
Option B - The technique is used to save major amounts of time by directly interacting with gameplay in some way, either allowing for glitches that lead to major skips or significantly skipping intended waiting sections as happens here.
[...]
As for option B, it absolutely should be a separate branch. My initial thought would be to put both branches in Standard, with the one using off-screen clicks as the baseline, since I believe that's how it currently works with L+R input in older console games? Though doing it another way may be more practical, I'm rather new to the site and don't yet fully understand the decision-making behind how the branches and classes are structured.
The tricky part is when this technique results in gameplay difference, but that difference is not enough to go to Alternative. I listed options we have with the current rules at the bottom of this post.
If we want this technique to be Standard eligible (meaning branches with and without it would co-exist regardless of entertainment), we need to come to a community agreement on which standard goal it would represent. If it's not about using and avoiding a major skip glitch, then we need a new standard goal for this. Which is in turn made tricky by similarities with L+R on old consoles: some games only get slightly faster over it, and some just break (there are probably other official intended limitations that we historically abuse, but I can't remember off the top of my head).
I always felt uneasy thinking about having a Flash-only one-off standard goal that wouldn't be applicable to any other platform, just like we're all seeing the similarity between it and L+R. Can "simultaneous cardinal inputs" be a separate standard goal on its own, regardless of its difference? I'm not sure it makes sense to me.
As I mentioned before, I'd be fine with having a standard goal along the lines of "fewest glitches", even if its definition is ultimately subjective. But at least once there's some kind of a definition, it becomes more or less clear what to expect, when watching and when TASing. But the current SA3 publication is not that.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
I didn't initially ask the community about this movie becoming its own branch or not, so here it goes.
Here's what off-screen shooting allows you to do. You have several missions where you have to wait for an object to appear on the screen so you can shoot it, skipping the forced wait. Or you can shoot it right away outside the screen.
My initial impression was that off-screen shooting only skips the waiting time, and the change is not visible to the viewer anyway. But there's also an argument that the wait is skipped by shooting, in a shooting game, so it's a direct part of normal gameplay.
So the main question is, do we make this a new branch or do we obsolete [4923] Flash Sniper Assassin 3 "inbound mouse" by Spikestuff in 02:37.50? If new branch, then what's the defining criterion? There's no visible difference so it can't go to Alternative, but maybe results of the off-screen shooting can be considered a major skip glitch, hence making both movies go to Standard alongside each other? If the argument is that glitchless can be a new standard branch, both movies abuse a glitch that skips some levels due to fast shooting (tho I'd personally be interested in a glitchless branch).
#8110: Spikestuff's Flash Arsenal in 01:31.36 is in a pretty similar situation, barring glitch usage in the current publication: Post #521966
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
I don't particularly remember what changed, but this run is definitely cool (pun unintended).
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
Saved 64 frames in the cave level by doing most things sooner:
User movie #638177837869179697
There are maybe other ways to tighten that level up, and probably other levels too.
We don't use rerecord count as argument in our judgments, but so far it does indeed feel like 3607 rerecords in 9.5 minutes is all that went into this submission (my rough test is already at 5k).
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
lexikiq wrote:
As far as I'm aware, the precedent has typically been to only accept one ACE category, maybe a second for playaround. I could see there being different branches for different releases of a game in cases where there's significant vanilla gameplay difference but for a branch whose goal is basically to trigger the credits as fast as possible using any means (and any version) necessary? I don't really see a reason for it, it should just obsolete the NTSC run, imho.
My thinking too.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Read the screen as an array of tests: XX-YYYY where XX is the byte precharged onto PPU open bus, and YYYY is how many loop iterations it look before PPU open bus changed its contents. Each loop is 15 CPU cycles
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
Can you just edit memory to simulate successful botting, as a proof of concept for whatever depends on it?
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
How do I make a fceux version of your script that shows which PPU register mirror addresses were read? I'm trying
Language: lua
memory.registerexecute(0xBD29, function()
local script_ptr = memory.readbyte(0x99)
if 0x2000 <= script_ptr and script_ptr <= 0x3FFF then
print(string.format(
"%d: $%04X $%04X $%02X",
emu.framecount(),
script_ptr,
0x2000 + (script_ptr % 8),
memory.readbyte(script_ptr)
))
end
end)
and it does nothing.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
Can you try reproducing this glitch on Mesen?
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
OtakuTAS wrote:
You can compare the tricks and see the improvement.
OtakuTAS wrote:
I obsoleted the original run (or... whatever you want to call it) based off of genuine improvement.
It helps a lot when improvements are mentioned in the submission text, because not everyone can watch the 2 movies side by side with enough precision to see all the differences, while also magically knowing which of those differences are not version differences but better optimization.
Which of the tricks used here to improve gameplay of the SMS version are also applicable to the SMS version?
OtakuTAS wrote:
Even the ROM with the identification given in the SMS publication runs at 50 FPS no matter where I source it from.
It doesn't run at 50fps until Bizhawk 1.13.2, even if you set it to PAL or Auto, both of which result in 50fps in later versions. And in 1.5.1 where the SMS movie was made, you didn't even have a way to set the display mode, so all games seemed to just run at 60fps.
Movie Rules wrote:
Gameplay must be accurate to hardware
Running 50Hz PAL games in 60Hz NTSC mode (and vice versa) is also not allowed, save for a couple exceptions:
Sega Master System games with an official release in Brazil can use NTSC mode, due to Brazil's PAL-M being functionally identical to NTSC.
I do wonder if the PAL-M scenario here created any loopholes of gain with Brazil SMS vs. NTSC Game Gear, i.e. if they adjusted the ROM at all to run better at 50 fps (higher walk speeds, etc.), it could be illegitimately faster on the PAL-M port.
The same exact game image was released in Europe, Australia, Brazil, and Korea. I don't know what's illegitimate in it being used in Brazilian 60fps consoles.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
<contest name> - <name of entry level> isn't something we've done so far, but for a new thing, it feels natural and intuitive for "games" that are just joke hacks, with potentially joke names. If there's a clash in names, we can still put author(s) in parentheses, as always.
Now regarding <name of entry level> (<contest name>), it emphasizes the actual name of the hack, and helps making the contest a completely separate notion, effectively preventing possible viewer misunderstanding, so I like it too.
Name of the hack is not a branch goal within an existing game, so it can't be a branch label.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
InputEvelution wrote:
I feel that Samsara's point regarding collaboration is particularly interesting. In cases where the person's submitted a movie for a game that isn't even on the site yet...why should we care so much about the state of the optimisation? Even if it's sub-optimal, they've put in more work than anyone else has so far for getting it on TASVideos. And even if one can demonstrate that the movie is unpolished, sub-optimal, etc., why not go and make some improvements to the movie yourself for a collaborative submission? No point complaining about a lack of effort from the submitter if you're not willing to put in the effort either.
This is absolutely an interesting point, and over the years quite a few submitters expected it to be the case, intuitively (and it wasn't). Level of required optimization may be worth another thread.
My natural feeling is accepting things that both look sloppy and are sloppy, would feel anticlimactic to a lot of people who watch TASes. But if a run looks clean and quick, and optimization has been attempted at all, we may try being more lax.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
To clarify, I don't remember "submitting many movies quickly" being used as a reason to limit people. It happened a couple times, and we just ask them not to do that. The things I have personally revoked submission privs over are:
1) repeatedly breaking some fundamental movie rules (at the time), even after having been informed about it, and warned about getting limited if the trend continues, and
2) repeatedly submitting visibly suboptimal movies (that were also proven to be suboptimal by actual tests), ignoring advises and feedback, and continuing the trend despite the warnings.
There hasn't been a defined system on having your privs restored, but it has usually been promised if a better movie is created and sent to staff to check. I was once or twice sent a movie, but it was either still easily improvable or non-trivial to check how optimal it is (2-player mame-rr movies). Then I came up with a suggestion for limited users to try improving an existing publication, which then would make it obvious their skill has improved, and the privs would be restored.
It's a fair criticism that this system doesn't really encourage people to improve, because I don't remember restoring anyone's submission privs. And it's also true that nobody really gives feedback to userfiles, especially for games they don't know, so submissions are the only way for those users to get any kind of feedback at all. Another fair point is that for malicious activity, it doesn't make sense to limit (and I don't remember if we ever did), because the user would just get banned.
Site atmosphere has changed dramatically over the last few years, so maybe there are better things to do when people repeatedly submit questionable stuff. Maybe dropping the rank entirely from the site is too soon, because we may happen to need it later, but I won't complain if currently limited users get their privs back.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
OtakuTAS wrote:
feos wrote:
Wait is it 12 hours or 128GB or 12 hours and 128GB?
Or. And it goes by the minimum, whichever you hit first.
A really grainy 12 hour video that's 1 MB will trigger a rejection, and a 10 second 130GB video will trigger a rejection.
That's an "and" then. It has to be shorter than 12 hours AND smaller than 128GB.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
Wait is it 12 hours or 128GB or 12 hours and 128GB?
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
OtakuTAS wrote:
Since "anything" is accepted these days is there even a point to the voting system or discussion (aka "vote and discuss") anymore?
Seems like game choice and entertainment are no longer considered, so it's just optimization I presume. Not trying to be sassy and it's hard to set tone through a keyboard, so a disclaimer that I'm genuinely curious.
I do think this is worthy of acceptance due to the effort put in to making the bot.
We're thinking of changing it to something like submission rating, similar to pub rating, with ratings carrying over to the pub. Then it would be able to mean whatever you want to use it for, technical quality, or entertainment, or both, or something else.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
Samsara wrote:
The quoted list looks good to me, although I'd even go as far as saying that avoiding glitches should be a universally standard category. It feels a little arbitrary to say that it's standard in some cases but not all of them.
Major skip glitch is usually big enough to allow some kind of community consensus, like "yeah we all agree that this is the glitch that we want to have both branches of, with and without it, because both feel common, traditional, and clear". For other kinds of glitches, which are usually smaller, and there's more of them, it's harder to draw a consensus-based line on which ones to avoid exactly. Unless you mean avoiding all glitches (which does feel standard eligible to me).
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
So I'm judging this together with #8110: Spikestuff's Flash Arsenal in 01:31.36 and #8111: Spikestuff's Flash Sniper Assassin 3 in 02:12.80, and the problem with checking how FlashPlayer handles offscreen mouse activity is libTAS doesn't work with mouse in FlashPlayer very well. I asked keylie if it's feasible to fix, but didn't get a reply yet, and other, even simpler Flash games, seem to have the same problem with libTAS+FP.
So far the argument has been that resizing the window and sending OOB mouse activity is not an intended feature for Flash games in their normal environment (web browser), but you can't really ignore the fact that the official launcher allows resizing (and possibly OOB mouse action too). So resizing it on the fly is still too standard to be banned, much like pressing cardinal directions on a NES controller. Sure it's very hard on an official controller. But there are also official controllers where it's easy. Not any less official: https://en.wikipedia.org/wiki/Power_Pad
Same thing with OOB mouse input: if it's digitally possible and the game processes it fine, it's similar to SNES controller buttons present only in digital form, but not on the physical controller itself.
So.
I'm leaning towards simply fully allowing window resizing and OOB mouse input for Flash TASes.
If it results in a game breaking glitch, fine, we then can have a movie that uses it, and a movie that avoids it, both in Standard.
If it results in smaller glitches, we can have a movie that utilizes them in Standard, and a more appealing movie that avoids them - in Alt.
If the difference is hard to notice, it's just a speed-entertainment trade-off, and up to the author, and can be a part of any branch.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Experienced Forum User, Published Author, Site Admin, Skilled player
(1238)
Joined: 4/17/2010
Posts: 11305
Location: RU
Did some updates, too tired to add the rest.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Waiting for these two to finish their dialogue and run out:
My initial impression was that off-screen shooting only skips the waiting time, and the change is not visible to the viewer anyway. But there's also an argument that the wait is skipped by shooting, in a shooting game, so it's a direct part of normal gameplay.
So the main question is, do we make this a new branch or do we obsolete [4923] Flash Sniper Assassin 3 "inbound mouse" by Spikestuff in 02:37.50? If new branch, then what's the defining criterion? There's no visible difference so it can't go to Alternative, but maybe results of the off-screen shooting can be considered a major skip glitch, hence making both movies go to Standard alongside each other? If the argument is that glitchless can be a new standard branch, both movies abuse a glitch that skips some levels due to fast shooting (tho I'd personally be interested in a glitchless branch).