Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
Here are a couple thoughts, some of which I've already brought up on discord but am adding here for the record.
Regarding Game Genie/Action Replay type codes: The number of codes allowed on runs could be either unlimited or limited to what was available on the originial hardware the code is intended for.
Examples: the NES Game Genie only allowed for 3 codes at any one time. Other Systems allowed for more than three. From what I can find, newer versions of some such ROM modifying systems may have been effectively unlimited themselves (or limited to some degree of memory, but no set numerical value). I wouldn't be able to make a justification for any number limit other than either original or unlimited, as I see any other numerical limitation as arbitrary.
Another consideration is the fact that, with enough codes, a game could potentially be rewritten completely into a wholly different game. Not that I forsee somebody wanting to go to that extreme; but I bring it up as an example of how extensively such codes can be used to impact a base game.
Regarding increased content & work load: As much potential interesting content that having a showcase/demo class may yield for the site, I have some concerns regarding an influx of new showcase/demo content.
1) If there isn't at least some method of curation or limitation of what's acceptable for this class of movies, the potential exists for the sheer mass of such content to potentially overwhelm the main site publications.
2) If curation/limitation does exist, who performs this? Given the potential freedom of goals within such a movie class, it's feasible that submission numbers could dramatically increase to a point that would majorly overtax the current site staff/judges if it becomes added to their current responsibilities.
3) Back to the idea of limiting what's acceptable. Even if this aspect of the site could be maintained discretely enough from the main site that the main site publications aren't overshadowed by what's presented in the showcase/demo class, I would still have concerns over total quantity of content.
If I understand things properly, the whole reason for wanting a showcase/demo tier is to be able to show runs that unfortunately don't adhere to the main site ruleset, yet are nonetheless interesting, highly entertaining, hold some historical significace, or provide an interesting TASing experience that is simply not publishable under current rules.
I don't want to see such a movie class be so overloaded with uninteresting content that the actually interesting/entertaining/historical runs get lost in the pile.
Unfortunately, I don't really have any good suggestions on how things should be implemented. I am rather curious to see how this plays out.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
zaphod77 wrote:
5) i still disagree on the decision that running a game that was not released in ntsc territory in ntsc is a valid choice. It's not a valid choice because real people don't have that option. their real tvs back in the day can only handle their own video standards. This is also to provide parity with console releases, where PAL versions must be tased on PAL emulation, admn NTSC versions must be tased on ntsc emulations. If we are allowed to tas pal released c64 games on NTSC settings, we would also allowed to tas pal NES game son NTSC settings. The fact that many games run on both is not a valid argument.
The bolded aspect of this statement is untrue (bolding added by me). Real people had the option/ability to play a PAL released games on a NTSC C64 system (and vice-versa) with no special requirements/modifications.
I had an NTSC C64 growing up. I also had games which were only officially released in PAL territories. These games worked just fine on my NTSC system.
It's true that we shouldn't allow emulation of a PAL system at the 60hz framerate. However, while a PAL system wouldn't function with a 60hz NTSC monitor/tv, PAL software itself often runs perfectly well on the NTSC system making it playable on a 60hz tv/monitor. This was not only possible but frequently done by real people.
Parity with console releases doesn't need to be considered. Consoles often have some form of region locking mechanisms in place that prevent play of a PAL cartridge/software on the NTSC system. It's therefore impossible to play such releases on an out-of-region system without modifying the hardware/cartridge to allow it.
The C64 does not have this type of region-locking restriction, and no modification of the hardware is necessary to load out-of-region software onto the system for execution.
EDIT:
If a game was specially coded to work with a specific regions frame timing, one of three things will happen when played on an out-of-region system.
1) The game will play normally aside from slightly increased/decreased speed.
2) The game will run, but glitches will be introduced due to the out of sync code/system refresh rate.
3) The game won't run at all.
Options 2 & 3 would be the only thing preventing someone in real life from playing an out-of-region game on their own system.
EDIT 2:
The above three possibilities still occur with properly emulated systems (BizHawk included). From my play/testing, BizHawk appears to perform the same as real systems would when presented with out-of-region software.
An example is Monty On the Run; the game appears to play fine other than a glitched end-game when played in NTSC mode. It was this glitch that required the TAS to be completed in PAL settings.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
CasualPokePlayer wrote:
DrD2k9 wrote:
I'm getting a desync in 1-3 trying to playback the winning run. Is anyone else seeing this?
Specifically: just after saving the victim, I'm guessing the player is supposed to clip north through the wall, but he doesn't.
The movie syncs fine for me. You're using 2.6.2 with the BSNES core right?
Yep correct version of hawk and correct core. Currently testing 2nd place, I'll update this post if it works or not.
EDIT: 2nd place run desynced in 3-4 for me. I found it odd that both would desync. I double checked everything I could think of including: player settings, core, cheats, etc. Nothing seemed out of place. I decided to completely close and re-open bizhawk and try again. After this restart, both runs synced fine. I have no idea what the original issue was that caused the desyncs and I can't find anything different looking at settings/etc. this time around.
TL:DR Though I can't explain why they didn't before, both of the top runs now sync fine on my system.
Congrats to the winners!
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
I'm getting a desync in 1-3 trying to playback the winning run. Is anyone else seeing this?
Specifically: just after saving the victim, I'm guessing the player is supposed to clip north through the wall, but he doesn't.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
I briefly skimmed through various pages on the demo site.
I was impressed by the overall speed improvement. Pages loaded dramatically faster than they do from the current site.
I like that access to one's forum profile stuff and private message inbox are directly available from the menu atop the main page.
I like that within a profile, the submission stats now show: Published, Cancelled, and Rejected instead of Accepted, Waiting, and Gruefood.
I expect there will be other features that I will find I like as well, but don't have time to continue looking around now.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
Now that it's actually been determined why 2 player mode is ultimately slower (longer boss fights due to higher boss health) and while refuting these arguments isn't going to magically make a 2p run faster, I'm going to respond to these anyway in order to point out that none of these arguments you provided are legitimate reasons why a 2 player TAS is slower.Technickle wrote:
1. 2nd player is more of a stand-around character and doesn't provide the necessity of having them around.
This argument is frankly not true from a movement standpoint.
Even if the 2nd player does nothing combat wise, he allows for improvements in player positions and earlier scrolling of the screen. If not for the longer boss fights, this improvement alone would be reason enough to allow for faster progression through the stages.
Technickle wrote:
2. The 2nd player doesn't have a combo'ing moveset like Cornelius does and makes it harder to kill enemies and bosses quickly. The 2nd player also would have to wait for the enemies to be on screen or semi on the screen to hit them.
All 5 of the characters special music attacks are capable of hitting off-screen characters once they are spawned.
There are times in the 1p TAS that Cornelius attacks on-screen enemies with attacks other than the special music attack. These other characters being present may allow for attacking such enemies faster than Cornelius can manage.
Regarding combo attacks: Not having as efficient combo options with other characters would matter if they were the only character being played, but they aren't. What other characters may lack in combos, the fact that there is a secondary character present means that he could potentially be positioned to attack the enemy before Cornelius could himself execute the second part of a combo attack. Further the secondary character lacking such combo possibilities is moot; because if Cornelius still can do them (and that's the fastest way to clear the enemies), the secondary character may not ever need to do such a combo attack themselves in a 2p mode.
Technickle wrote:
3. Characters such as Maxwell, Angus, Chester, and Clifford are slow movement-wise, have a slow special move, and are awkward to play as.
The only part of this that is legit is that the special music attacks are slower than Cornelius. But this again is moot as the secondary character may never need to use their music special.
Regarding movement, Maxwell literally has a dash move that zips him to the far side of the screen in the same amount of time it takes Cornelius to take 1 animated step. This alone can get things moving earlier if he is used as the 2nd character. Even if a different character were used, the 2nd player starts further to the right of stage 1 than player 1 does. Thus a slower moving player would only delay things if Cornelius could outrun them by an entire screen in such a way that the slower character would then delay scrolling. This is unlikely due to the frequent forced stops of screen scrolling.
A character being awkward to play is NOT an excuse for not using them in a TAS. You have all the time in the world to work out their movement/control in a TAS setting. While a particular character may be awkward for a human to control, a TAS should still use them if they would allow for more optimal play due to having perfect precision/control in a TAS environment. (Granted, they don't appear to be more optimal; but here I'm mainly refuting the argument against using them because of their being awkward).Technickle wrote:
4. With having the benefit of 1 player, I can control the game easier, the screen with 2 players is awkward because either player 1 or player 2 needs to progress the screen first in order to go further towards the end. Some enemies don't even activate if player 1 goes further ahead of player 2, as player 2 would need to be there to do so, thus making it slower. example: Window Guards
At least in level 1, I could not replicate a situation where player 2 being in front of player 1 prevented enemy spawns/actions in such a way that it made things slower than a 1 player version.
In my testing, zipping Maxwell to the right of Cornelius immediately at the beginning of the game allowed for an earlier attack at the first wave of enemies, and nowhere through the first stage did it appear to delay any other enemy action.
Technickle wrote:
5. While not as 'entertaining' as a two-player run, this run still aims for the fastest completion status. If by some miracle I feel obligated enough to make a 2 player TAS, it won't be the same (for me at least). Yeah, the strategies to kill enemies could be neat, but, it lacks the drive and purpose of being fast rather than: "I can do no special moves and play-around with the enemies for minutes on end".
My original question about a 2 player run was not about entertainment... it was regarding if it would be faster (though it has now been confirmed to be slower due to higher boss health).
Regarding entertainment: to me, watching this game in either 1 or 2 player mode is boring because almost all one effectively sees is Cornelius' character screaming. Boss fights really aren't enough to add much entertainment value.
Still, in claiming that you're going for fastest completion (essentially meaning regardless of entertainment), you've failed to make the only true argument as to why a 2 player TAS is ultimately slower...more boss health.TL:DR None of your arguments attempting to explain why a 2p TAS would be slower actually pan out in a TAS setting, and you failed to note the only true reason that 2 Player mode is ultimately slower.
I'm not trying to be mean/rude with my comments; I'm trying to help you understand how to consider things differently and learn from those with more experience who have tried to help you in the past.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
I had another thought/question.
Did you (Technickle) ever actually test if a 1P or 2P game was ultimately the fastest method?
I know this was brought up with one of your previous submissions of this game. Your response at the time was:
Technickle wrote:
I do agree with you on your point about the 2 players being more interesting than a 1 player scenario. However, I don't want to put myself in a situation where I am not aiming for a 2 player run.
For the purpose of this submission, I only want to compare to a 1 player run if that is acceptable and or allowed in this context.
Under the current site rules as I understand them (recognizing that the rules may be changing due to recent discussions about vault expansion/change): if a 2 Player run would be faster than this submission (which, based on the previous submission's discussion, it likely would be), this submission would need to receive enough positive entertainment response to be published into the moons tier as it wouldn't be eligible for vault publication due to being a sub-optimal method of completing the game.
Given that this submission is sitting at 1 total vote (which is a NO for entertainment), there doesn't seem to be much interest in this game at all. This means it has a significant uphill battle to garner the necessary feedback for moons publication.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
Just a note for Technickle: You may want to consider taking a bit more time to optimize your runs.
Even before this submission, I had already noted the rapid succession of complete game runs posted to userfiles from keeping an eye on the "Updates" channel in discord. Based on the history there, you have uploaded 12 userfiles of this game since June 30 (just over 2 weeks); only one of these was labeled as a WIP, and the rest were apparently complete game runs.
From what Spike's chart shows, there are odd curiosities from version to version in how long any given stage takes, with numerous occurrences where you actually lose time to one of your own previous versions.
From an optimization standpoint, it makes zero sense that you have two instances where you lost time in Level 1 to a previous version of the TAS. It's the first stage of the game; there's frankly no excuse to lose time to a version you've already made unless you can prove that the time lost in that situation results in an even greater time save later on in the game.
This all results in your runs appearing as if you're not utilizing your own best input from previous versions as you create a new version of the TAS, but instead are redoing everything from scratch each time.
You have 11-12 uploads of seemingly complete redoes of game in barely over 2 weeks. While the total game time improves with each version, the inconsistencies and lost time suggest that you're only considering optimization from a total time perspective, and not considering optimization of individual sections/levels of the game.
Ultimately, all this makes it appear as though you're really rushing through these improvement attempts.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
feos wrote:
4. Turn submission polls into a 0-10 scale
5. Change the poll question back to "Should this be published?" and treat poll results according to presence of external goals, which is finally a clear borderline.
Won't whether a not a movie is published still ultimately be determined by the judge (even if allowing for more flexibility based on viewer feedback from the forum/poll)?
I like your idea of condensing all 3 current polls (workbench yes/no/meh, post publication entertainment rating, and post publication tech rating) into a single rating. However, given that "Should this be Published?" doesn't really ask for a rating that fits on a variable scale; perhaps we could instead present viewers of workbench movies with a simple poll that just asks to "Rate This Movie" on the 0-10 rating scale?
Regardless of the poll question another consideration from a rating standpoint would be a 1-5 scale as that is another format many (if not most) users would be familiar with for rating things.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
I firmly stand behind the idea of opening things up and attempting to remove/reduce any negative 'stigma' of the Vault.
I'm all for entertainment in TASes, and I also believe that fastest completion TASes should be just as celebrated as entertaining ones simply for the TASing achievements they are (regardless of their entertainment value).
Eliminating any degree of negative perception--that Vault publications are somehow less important to the TASing community, just because they may be less entertaining--is a worthwhile endeavor.
Edit:
Out of curiosity; how would this affect things like Board Games and Edutainment? Would these titles still be restricted from the site, or would the proposed changes also potentially open publication opportunity for these types of games?
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
If this thread is meant to be a place for all TASes that are known to use uninitialized RAM regardless of whether or not they've been console verified yet,
[4257] GB Donkey Kong by DrD2k9 in 54:15.44 qualifies.
If this thread is only meant for TASes that have yet to be console verified, feel free to delete this post (or tell me to delete it).
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
I'm in and teaming up with NYMX.
To anyone else who joins us, be forewarned: I may have very limited time to put toward working on this contest, but they are fun enough that I don't want to completely miss out.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
feos wrote:
Do we want to agree that getting high scores in all events means fastest completion for them?
I don't remember who's comments got me thinking along these lines, but my perspective on score based events is that they can't effectively have a "fastest completion" that doesn't also max out the score. Stopping at any point before the maximum (even if it's a new high score) is essentially acknowledging that the event can be performed to a fuller extent with a higher score and thus remains incomplete.
It can't be a fastest completion if it remains incomplete.
Consider the general concept of a high score: The whole point of high score tables/records in games is to try and improve upon whatever the current high score is. While game designers may pre-program a default high score, it can probably be safely assumed that (in most cases) the designers didn't intend for that particular high score value to be the maximum that could be achieved in-game. Especially when considering older coin-op games; the primary purpose of the high score was to drive people to keep playing (and thus keep paying) to try and best whatever the top score was. The competition created between actual people--not people and the pre-programmed score--is what drove the continued payment. This high score concept bled into games for home-based systems even though constant payments were no longer a component.
Given this concept; for a full completion run, if a TAS can be made to perform better than "simply surpassing the pre-programmed high score," it should be. Therefore, in order for a score based event to be considered complete, a maximum score should be necessary.
Accepting a score below the max but above the pre-programmed high is an arbitrary point issue. This is especially the case if the event can be sufficiently finished with a score that allows for continuing onward but remains below the high-score threshold. NES Track and Field would be a good example of this. There are scores that qualify the player to progress in the game without getting a "game over", but are still lower than the in-game "world record" high score.
If we're wanting to allow any score below the maximum to be considered for "fastest completion," then any score that the game considers good enough for progress should be allowed. This would require including games where complete failure in an event (via attempts resulting in only faults) would have to be allowed for "fastest completion" if the game allowed the player to continue beyond that point. The above mentioned Decathlon game would be in this category, it allows for complete failure of a score based event while still progressing to the end of the game. I wouldn't like the idea of accepting things that way.
Speed based events are obviously speed based and should simply be completed as rapidly as possible.
TL:DR
IMO, From a TAS perspective, a maximum score is necessary to consider a score-based event "complete" at all. Thus for a "fastest completion" run, maximum scores in score-based events are necessary to fulfill the "complete" aspect of "fastest completion."
Score based events simply don't lend themselves to being completed quickly and this will sometimes result in repetitive gameplay to achieve the max score and resulting longer total runs.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
Dacicus wrote:
According to several sources, at least one version of this game was available as a PC booter, which means that it can run from a floppy without need for an operating system. IDK if JPC-RR emulates that, but it might save some of the loading time by eliminating the need for FreeDOS.
While that may save some startup time, it wouldn't affect the gameplay aspect of the run. Given that loading times aren't considered for improvements/obsoletion, the resutling time savings would be moot. Considering it'd be less than 3 seconds of time saved, even from a publication standpoint, it'd hardly be noticed.
As a comparison, consider Commodore 64 games. Games loaded from tape or disk can take multiple minutes (sometimes over 10) just to load the game. When one of those same games is on a cartridge, it will load near instantly. In that situation using the cartridge version would be nicer from a publication standpoint because it would elimiate minutes worth of loading times even though the gameplay itself is completely unchanged.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
Entertainment
First let me say that this was obviously a very well planned and executed TAS, and I can appreciate the work that went into creating it. Kudos to the authors for the accomplishment.
Regarding entertaiment though: While it did hold my attention throughout (mainly just out of curiosity as to how things would be done), I wasn't really entertained by the on-screen activity itself. To me, this video was a slowish completion of the game which jumped when absolutely necessary and otherwise simply wasted time while slowly waiting for a kooopa/beetle/etc. to get into positoin so it could be used either as a springboard or a method to glitch into a clipped position.
Yes, the tech of the run is interessting, but visually watching Mario rapidly duck/change direction and shoot fireballs in sync with the music in order to kill the time necessary to wait for enemies to slowly do their thing just wasn't very entertaining to watch.
Personally, I find various other of the current SMB publications (including the basic any% speedrun) significantly more entertaining to watch than this run. Even the walkathon is a more entertaining watch to me than this run.
Acceptability for Publication
I'm torn on whether or not this run should be acceptable for publication.
From a technical achievement standpoint, I fully support that this run warrants some degree of recognition/archival.
However as mentioned above, I just didn't find it all that entertaining in comparison to the collection of other SMB runs currently published.
Why this is difficult for me is as follows:
I'm one within the community (in the admittedly minority group) who believes that most, if not all, games (including educational/board games/etc) should have some place to have their fastest completions be published--even if only for archival purposes--regardless of how entertaining they may or may not be. This obviously doesn't line up with current site protocol.
I also feel that significant technical accomplishements should similarly have a place for such archival; again regardless of entertainment value.
However, as the site currently requires runs other than 100% or Fastest Completion to meet entertainment criteria for publication, I don't support publication of this run from an entertainment perspective. It already appears that I'll be in the minority on this perspective as well. I just didn't find it visually entertaining enough to warrant publication along side all the other published SMB runs from an entertainment perspective.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
I tend to ignore most April Fool's submissions, but this one intrigued me enought to actually watch.
This was very impressive!
Since I can't imagine it being accepted due to the emulator limitations and the 'cheating' required to accomplish the sync....I'm pre-nominating this for Gruefood Delight before it likely (and sadly) gets rejected.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
I've run into similar issue with the Odyssey 2 core and can somewhat replicate it.
I believe this may be an issue related to savestates.
Using the Pinball ROM, I can show some images to explain a bit of what I'm encountering.
When I just load the ROM, open TAStudio and start running emulation; I get the following:
When I add input, the following occurs and I get the wall of lag:
Then after the input is in place, restarting emulation from the beginning will result in this:
Then sometimes (but not always), clicking to jump back the emulation at an earlier frame results in the big block of lag returning. In this example I clicked to jump back to frame 12:
I'm not sure if this is the same bug that Lobsterzelda is experiencing, but it sounds related.
Again, I'm guessing it's a savestate issue, but I'm not a coder to know how to investigate further. There are no error boxes/logs/reports occurring.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
Dimon12321 wrote:
Is it possible to transfer regular input with TAstudio window opened? It's not that convenient when you have to use either only mouse to mark specific cells in the input grid, or only console input and then open TAstudio to perform various manipulations on it.
Check the "Recording Mode" check box, and TAStudio will record your controller/keyboard inputs.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
Ok. I've completed multiple versions of the game starting with Winslinator's userfile that completes the menu stuff 21 frames earlier than this submission.
Here's where things stand:
This submission (slow-menu approach) has 8,590 frames of gameplay from the first frame of controlling the first ball until the final input.
The fastest run I could attain using the fast-menu approach, has 8,617 frames of gameplay from the first frame of controlling the first ball until the final input.
In other words, I was unable to complete the actual gameplay faster using the fast-menu approach.
However, that doesn't really matter, because (even with the earlier start) the best fast menu approach run finished 6 frames after this submission.
27 frames slower gameplay - 21 frames save in menu = 6 frame slower total run.
Interesting side note: I have a second slow-menu approach run that completes the gameplay portion in 8,612 frames. Meaning there are at least 2 methods using the slow-menu approach that yield faster gameplay portions than any fast-menu approach I could accomplish.TL:DR
While finding a way to do the menu stuff faster was interesting. The improvements aren't sustained through the entire run. The submission as it stands is the fastest known completion of the game.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
Winslinator wrote:
I've actually found a potential improvement to this. I uploaded an edited movie file which shows that the menuing at the beginning could be 21 frames faster by using the second controller to squeeze in more commands. Unfortunately, it desyncs after the second throw so much of it would need to be redone.
I would volunteer to help finish it myself but I am unfamiliar with the game's mechanics, nuances, and such. Hope this comes in handy!
Very cool find. I will look into this as soon as I can. I'm going to leave this submission up for now (as it's yet to even be claimed for judging). If the improvements carry through, I'll provide an updated full completion run to userfiles.
Regarding using your menu improvements: If they don't save at least 21 frames over the course of the actual gameplay, then my run (while starting actual gameplay slightly later) would still be the faster gameplay. If I'm not mistaken, the site prefers better/faster gameplay over faster menuing when there's a slight discrepancy.
But I'll make sure to comment here later with what I can find/accomplish.
Well....it's not a dumpster fire.
When I add input, the following occurs and I get the wall of lag:
Then after the input is in place, restarting emulation from the beginning will result in this:
Then sometimes (but not always), clicking to jump back the emulation at an earlier frame results in the big block of lag returning. In this example I clicked to jump back to frame 12:
I'm not sure if this is the same bug that Lobsterzelda is experiencing, but it sounds related.
Again, I'm guessing it's a savestate issue, but I'm not a coder to know how to investigate further. There are no error boxes/logs/reports occurring.