Experienced Forum User, Judge, Published Author, Skilled player
(1278)
Joined: 9/12/2016
Posts: 1645
Location: Italy
I've noticed that both Groudon and Kyogre require 18 hits to be beaten the second time you meet them. Does the same apply for Rayquaza?
By the way, I found it very entertaining! Great job.
Experienced Forum User, Judge, Published Author, Skilled player
(1278)
Joined: 9/12/2016
Posts: 1645
Location: Italy
tormented wrote:
You're using a bad ROM dump (as evidenced by the lack of green checkmark in the lower left). I tried "Karnov (U) [!].nes" from my collection with this movie just now, and encountered none of these issues.
That's weird, because I used the good dump for playing back the movie, and it synced fine. That's strange to think that a movie made with a bad ROM that doesn't work on an emulator can produce a movie file that also syncs with the a ROM that works on both emulator.
Experienced Forum User, Judge, Published Author, Skilled player
(1278)
Joined: 9/12/2016
Posts: 1645
Location: Italy
Double_Pickaxe_YT wrote:
ThunderAxe31 wrote:
Please post your new movie file in the userfile space, don't make a new submission.
Why tho
Because the submission form should be used only for movies that have a chance of getting accepted. As explained by Samsara in the judgement note she wrote for you, SMB1 is a very hard game to TAS, and making an acceptable TAS for this game requires a lot of experience. In order to avoid getting another rejection too soon, I warmly recommend you to upload your new TAS in the userfile space, and then ask for feedback in the SMB1 thread or in the Discord server. We will answer your questions and help you to improve, until you will be ready.
Experienced Forum User, Judge, Published Author, Skilled player
(1278)
Joined: 9/12/2016
Posts: 1645
Location: Italy
I found this movie as repetitive in some aspects, and I found the non-repetitive aspects to be just chaotic. Because of that, I couldn't observe the role played by the TASing optimization for this game. It leaves me a taste of boredom and confusion. No vote from me, sorry.
Experienced Forum User, Judge, Published Author, Skilled player
(1278)
Joined: 9/12/2016
Posts: 1645
Location: Italy
Hi there, I have a question. I'm watching side-by-side the RTA WR by hirexen and this submission movie, and I've noticed that the RTA run is faster in the second half of stage 3... Is there a known reason for this?
Experienced Forum User, Judge, Published Author, Skilled player
(1278)
Joined: 9/12/2016
Posts: 1645
Location: Italy
WarHippy wrote:
arkiandruski wrote:
... we should probably wait until the good ending run is finished before passing judgement on this one.
Why? It sounds like you're grasping at straws instead of having a solid basis for acceptance.
Let's not begin arguing on what sounds like what. The judging process should always take in account as many aspects as possible, so if we've learned that a good ending run may be out soon, then it's only natural wanting to wait just a bit for it, in order to have the chance of making some comparisons. Additionally, this is necessary when assessing the entertaining response of the audience, as it's always better to be able to actually see the opinions of the people.
Experienced Forum User, Judge, Published Author, Skilled player
(1278)
Joined: 9/12/2016
Posts: 1645
Location: Italy
evilas wrote:
This is interesting, so basically if I'm understanding it right, the 4 ways to complete the game I mentioned could in theory be considered 4 different branches? I can see how that could be the case, with publication or lack thereof being dependent on how distinct the runs are from each other.
Well, technically yes, we consider them as different goals, but we can allow only one of them for publication, as we try to avoid redundant movie contents as much as possible, so only the fastest of those 4 would be acceptable. See from the Judge Guidelines:
Sum-up
Quantity is not quality.
Keep the number of different branches per game minimal. A run for a proposed new branch for a game should offer compelling differences relative to previously published runs of that game.
Avoid making decisions that undermine this guideline (or other guidelines) now or in the future. For example, don't publish an arbitrarily rule-restricted movie just because there are too few movies for that game; doing so may lead to impossible-to-solve "why A but not B" debates later.
Hard work should have a reasonable chance of being published. More on this below.
Satisfy the audience's expectations.
evilas wrote:
In that case yes, I would suggest, at least for now, the "Defeat Ganon" label for the old publication.
Experienced Forum User, Judge, Published Author, Skilled player
(1278)
Joined: 9/12/2016
Posts: 1645
Location: Italy
evilas wrote:
ThunderAxe31 wrote:
For the case of this game, it's very simple: just beat the game without using arbitrary code execution, as it's the currently known way of skipping more than half of the game.
I hope this clears it a bit, but if you still have doubts, feel free to point them out.
But the thing about it is, this is just not true! The old TAS, without stale reference manipulation, absolutely beat the game with VERY little of the game beaten. It used an extremely major skip to warp directly from the end of the first level to the final boss, and was labelled as such! It just was also left branchless, whereas this is branched!
Sorry, I completely forgot that the labelless movie also performs major skips. Indeed, in this case we can distinguish the two run goals only by looking at which one skips right to the credits via a glitch.
evilas wrote:
ThunderAxe31 wrote:
Instead, we leave the label blank in order to denote that that movie beats the game without showcasing any glitch in particular, and/or follows a restriction that makes it closer to the most common way of playing the game.
Again, this isn't how you're implementing it! The other one is branchless just because of historical reasons and I don't think it should be. If you want to let people know what the movie is about, I think it should be branched as "defeat Ganon", to clarify that this is what people are signing up for.
Sorry again, I was just explaining what are the policies in general. I'm not following the OoT scene closely, so I wasn't implying that the labelless movie is the most common way to play the game. If you think that it isn't the case here, then we can discuss for adding a proper label to that movie.
evilas wrote:
There's 4 ways a TAS that doesn't ban stale reference manipulation can beat the game quickly:
1) use ACE to warp to the credits
2) use ACE to warp to the final boss
3) use non-ACE methods to warp to the credits
4) use non-ACE methods to warp to the final boss
Which one(s) would you consider branchless?
Mind that obsoletion management and movie labelling are two separate issues. The judges define the requirements for obsoleting a movie, while the publishers define the label that explains the movie contents.
To answer you question more extensively, see from the Publisher Guidelines:
There are cases where some branch is unique in not setting any unique conditions. There is nothing special to put in its label, since it explicitly avoids external goals, and the game doesn't have any internal conditions. We refer to such runs as the trunk and leave their labels blank. Remember that this is not related to "any%" or "default goal", because these things are complicated, relative, and not always clear. Also note how "trunk" and "branches" are metaphors for a concept in the real world: they describe a tree. We try to structure runs of the same game in a similar manner.
Thanks, I forgot to mention that. The movie tags make it much simpler to understand it for the casual watchers.
evilas wrote:
Why would that be? Why not call the fastest category "any%" and the categories that do more interesting stuff by actual category names?
If GEG is treated as any%, then you can easily publish the other one as "no GEG", nothing else changes. The only difference is which one you consider "default".
EDIT: why not have category names for all of them? Why not call Bloob and Grunz's run "no SRM"?
It looks like my post was buried and remained unnoticed. Please, try reading my post and see if it answers.
But I'll try to explain it again: we actually don't label branches, but we label each movie publication independently, because we want to make it clear to casual watchers about what are the movie contents. Instead, for denoting the difference of the branch goal we use movie tags, from which you can tell which new submission could obsolete which movie. In most of the cases these tags are "uses major skip glitch" vs "forgoes using major skip glitch", like for the movie in this case. Additionally, the fastest-completion movie is denoted with the relative flag:
For this reason, we usually don't use a label that consists in "no that glitch", because that doesn't tell what the movie contains. Instead, we leave the label blank in order to denote that that movie beats the game without showcasing any glitch in particular, and/or follows a restriction that makes it closer to the most common way of playing the game.
Regarding the "any%" naming, we don't use it because it doesn't tell what kind of movie contents are featured. Also, when a new faster branch is introduced due to a newly discovered glitch, the "any%" naming would become misleading for the previous movie, as the fastest way to beat the game would effectively have changed. We can't rely on a label whose meaning may change depending on external factors. Additionally, the "any%" meaning is technically inapplicable for most games, as most games don't have any kind of in-game percentage. We use the "fastest-completion" wording instead, and we denote it with the relative movie flag , not with labels.
evilas wrote:
If the dividing line is whether or not you use ACE, then GEG is not a good branch name for this run imo, considering you can do a credits warp that looks just as "game ending glitch"-y without using ACE?
The dividing line is using major skip glitches or not, which is an objective yardstick that can't cause any confusion. Any movie that skips more than 50% of the supposedly mandatory game contents, will be denoted with that movie tag: http://tasvideos.org/Movies-C3041Y.html
Experienced Forum User, Judge, Published Author, Skilled player
(1278)
Joined: 9/12/2016
Posts: 1645
Location: Italy
The main purpose of the site is to showcase entertaining tool-assisted gameplay. For that reason, we manage the publication branches in a way to feature diverse gameplay. At the same time, we avoid having redundant movie contents in published movies, by obsoleting movies that are too much similar to each other.
With that said, we also give importance to competition, so of course we also manage obsoletions by following a logical reasoning. In this specific case, we differentiate between "game end glitch" and labelless by the presence of arbitrary code usage. In simple words: any future submission that beats the game as fast as possible we be competing for obsoleting the "game end glitch" publication, while instead submissions that beat the game without using glitches that are known to allow to skip more than half of the game (any sort of game end glitch) will be competing for obsoleting the labelless publication.
Now, about the labelling system. This involves just the publication naming of movies, so it doesn't affect the judgement process, which I just explained above. In fact, we use labels in order to explain in a short and clear way what's the movie contents about. The label "game end glitch" explain clearly that in this movie the game is beaten by triggering the game ending routing via a glitch. It doesn't matter if it's done thorough a simple glitch or via a complex arbitrary code execution setup. What it matter is that more than half of the game has been skipped. Instead, for movies that don't skip more than half of the game through glitches, we leave the movie as labelless, as that denotes a sort of gameplay that it's closer to the most common way of playing the game. We never use "any%" because it just tells that you're beating the game as fast as possible, without indicating what the movie contents are about. What's worse, is that if at some point a new branch is creted due to a new glitch discovery, the branch previously known as "any%" would technically not be fastest completion anymore, leading to confusion.
Reassuring:
1. We technically don't label branches. We label each published movie independently.
2. We manage obsoletion by looking at how much of the game is skipped, usually "beat the game as fast as possible" vs "beat the game while forgoing using a specific glitch or a set of glitches". If a movie uses a too much arbitrary or confusing ruleset, we simply reject it, as we want TASers to know exactly what they have to do for obsoleting a non-fastest completion movie. For the case of this game, it's very simple: just beat the game without using arbitrary code execution, as it's the currently known way of skipping more than half of the game.
I hope this clears it a bit, but if you still have doubts, feel free to point them out.
Experienced Forum User, Judge, Published Author, Skilled player
(1278)
Joined: 9/12/2016
Posts: 1645
Location: Italy
TASVideoAgent wrote:
Engine: The skip for the first room that PiePusher did involved waiting for a spike to get high enough on the screen to get a damage boost. I found a better skip which involves getting bounced by a canon then by a cannonball that involves very little waiting.
Hi there. In the last level, K.Kruizer III Engine, you indeed saved a lot of time in the first room. However, if we look at the second room alone, you're actually slower than the currently published movie. Could you please take a look and see if you can fix it?
Experienced Forum User, Judge, Published Author, Skilled player
(1278)
Joined: 9/12/2016
Posts: 1645
Location: Italy
Nach wrote:
The CPU computation in a CGB is supposed to be the same as a DMG, the place where they differ is in the sound and video generation components. For a game that doesn't try to read that back, it should in theory play exactly the same, minus audio or visual differences. However if the game does read it back, or use it as part of timing, which in turn feeds into an RNG or for other computations, the game play is different. I've heard stories how bosses suddenly became much easier or harder. This is where the differences can be subtle.
We already allow any different revision of the same game, especially when one revision has specific bugs that allow to beat the game faster. We also already allow different hardware revisions, for example the different PSX BIOS versions, for which in some aspects may feature even deeper differences than GB vs GBC. You insist that GB games haven't been designed for the GBC platform, but that also applies for any other platform game that was developed before introducing hardware revisions. Is there any reason why we should treat the Game Boy differently?
feos wrote:
If glitches that are caused by newer mode can't be easily noticed and don't hinder gameplay, the newer mode is allowed for the sake of console verification.
It's not just for the sake of console verification. In my opinion, GB-in-GBC should be also considered as a stylistic choice for improving aesthetic of a movie, as well as be allowed to taking advantage of hardware differences that may allow for a more optimized movie (apart from the BIOS screen duration difference).
Experienced Forum User, Judge, Published Author, Skilled player
(1278)
Joined: 9/12/2016
Posts: 1645
Location: Italy
After the judgement precedents set with Devil Island and Link's Awakening, we have general agreement that there is no requirement to wipe the SRAM with in-game functionalities. On the other hand, I still consider it as a good practice. But on the other other hand, in this submission the game is properly saved already once, before performing the save file corruption, so it's technically unnecessary. Still, I want to leave freedom of choice to the author, and still allow it as a stylistic choice.
CasualPokePlayer, would you like to replace the submitted movie with a version that skips the memory clearing, or leave the movie file unchanged?
Experienced Forum User, Judge, Published Author, Skilled player
(1278)
Joined: 9/12/2016
Posts: 1645
Location: Italy
Pardon me if I'm attempting to give a different answer after that two other people already did it (very informative posts btw), but I feel that feos' question was trying to figure out a slightly different issue.
feos wrote:
I have a different question. For GB games that are supposedly not supported explicitly by GBC, it still somehow assigns colors in a meaningful way, it's not an utterly random mess. How is this determined on the hardware level? It's not full-color like actual GBC games are, but still looks sensible. Are there also GB games that have completely wrong colors assigned in the GBC mode?
The different coloration is assigned on the basis of which is the VRAM region that holds that graphic, and the GB's Object Palette setting (more info below). All GB and GBC graphics consist of three graphical layers, from bottom to top:
1. The normal background (also known as Background)
2. The window background (also known as Window)
3. The sprites (also known as Object Attribute Memory or OAM)
Now, the GBC assigns three different palettes to a GB game, depending of the following:
1. The first palette is assigned to both Background and Window
2. The second palette is assigned to sprites that are set to use the Object Palette #0 (OBP0)
2. The third palette is assigned to sprites that are set to use the Object Palette #1 (OBP1)
I know it may look weird, but even if the original GB doesn't have colors at all, it still allows two diferent grayshade palettes for the sprites. Since all sprites always have one transparent color out of the four grayshades available, it effectively limits the graphical possibilities. In order to circumvent this limitation, the GB gives to the programmers the ability to freely decide which color use for transparency, and have two different sprite palettes available for simultaneous use.
So in the end, yeah, I agre the colors picked by the GBC are "not an utterly random mess", but are still technically random, unless the game used falls into the list linked by Nach above.
Anyway, in a worst case scenario, the user is still able to manually pick the desired palette combo during the GBC BIOS sequence, for the better or the worse results. However, it doesn't help for cases of games that use the sprites as graphics to blend with the background layers, since it sometimes makes them stand out in an inappropriate way, hence the scanarios you posted previously:
It also doesn't help for cases where the sprites were originally supposed to be kept hidden among the background layer, as explained by Nach here:
Nach wrote:
There's another element also that in a way seems like cheating. The DMG only had 4 colors or so to choose from. Some games would hide enemies or breakable areas in the background/wall/floor, and you wouldn't notice it unless they attacked you, or you attacked it, as they blended together. When a CGB plays a DMG game, it automatically uses more than 4 colors, and purposely uses different colors for what is truly background and what is a sprite. Suddenly all those enemies or breakable areas (sprites) that were hiding suddenly stand out (from the background), and what it supposed to be a surprise is readily obvious. You can see enemies in advance that are supposed to ambush, you can know where to strike without any guess work. It dumbs down various games in certain areas.
For these reasons, I'd like to pick this chance to reiterate my proposal: since the GBinGBC mode may make some games potentially uglier to look than GB, I think it should be allowed only for Moons & Stars.
For this reason, we usually don't use a label that consists in "no that glitch", because that doesn't tell what the movie contains. Instead, we leave the label blank in order to denote that that movie beats the game without showcasing any glitch in particular, and/or follows a restriction that makes it closer to the most common way of playing the game.
Regarding the "any%" naming, we don't use it because it doesn't tell what kind of movie contents are featured. Also, when a new faster branch is introduced due to a newly discovered glitch, the "any%" naming would become misleading for the previous movie, as the fastest way to beat the game would effectively have changed. We can't rely on a label whose meaning may change depending on external factors. Additionally, the "any%" meaning is technically inapplicable for most games, as most games don't have any kind of in-game percentage. We use the "fastest-completion" wording instead, and we denote it with the relative movie flag 
It also doesn't help for cases where the sprites were originally supposed to be kept hidden among the background layer, as explained by Nach here: