Alright, so, it seems like people agree on the proposal.

From now on, the timing definition for the beginning point of any IL will be this :

“

Timing starts upon the first frame that the screen is fully white as the stage appears onscreen, after the black loading-screen.

“

I will now check the current submissions before verifying the pending runs. But you can consider DCT ILs as operational again !

Hello there. Not good news.

We just realized there’s a big issue with our current definition of what the beginning timing point of a DCT IL is. Unfortunately, we weren’t thorough enough to catch it back when we were making it.

The rule in question : “Timing starts upon the earliest frame of the level appearing onscreen after the black loading screen.”

To be clear, this rule was taken word from word from the EoD IL rule page counterpart. I’m not sure why I just assumed it worked fine wih DCT because it doesn’t.

You see, this rule works perfectly for EoD. When you enter a stage from the world map in that game, the game puts a loading screen of pitch black color. Then, once the stage is ready to start, the screen instantly becomes white, in only one frame. This is consistent behavior. Hence, it is easy to determine when the stage IL begins, because it begins as soon as the screen changes to white… in other words, when it isn’t black anymore.

DCT is like this… on the surface. Yes, when you select a stage on the world map, there is a loading-screen that is black, and yes, it also very quickly turns into a white screen when the stage is ready to go. But if you look closely there are differences. Two ones, which I will discuss below, and we will be using @MackieGarou’s footage for this : https:// www.youtube.com/watch?v=HzZTD6_hZZs

.

Difference 1 : The game displays several frames of randomly colored dark artifact pixels, before the screen becomes white. Here are some examples : https://imgur.com/a/kMDJowv

Not only this is consistent, but the artifact pixels patterns seem very unpredictable, they don’t look the same between stages.

It is obvious that these frames aren’t proof that the IL has started. Because the game is clearly still loading during them and because the stage hasn’t loaded. We thus cannot begin the IL on them… and so, our current rule is ineffective. But that’s an easy problem to solve. We can simply disregard these frames and instead rule that the IL must specifically begins once the screen gets white. Simple enough, right ? Well...

.

Difference 2 : Unlike EoD, the loading-screen doesn’t instantly become white. Instead, it can take more than one frame to turn the screen white. In fact, it commonly does.

This means that the first frame where the game calls for white, you often see only half of the screen being covered up in white. Some examples : https://imgur.com/a/nz743My. Worse, the overall screen coverage of white pixels also seems random and variable. It usually takes about half of the screen, but as you can see from the link above, it can sometimes stay very close to the bottom of the screen or, instead, take almost but not all of it.

Can we truly say the stage has begun, when the game screen isn’t fully white yet ?

.

Lastly, I observed that, when it comes to the loading-times that come before boss stages, Difference 1 doesn't seem to be present. There aren’t any artifact frames. That said, difference 2 is still here. So it honestly doesn't matter much.

.

As a side note : while difference 1 hasn’t been shown to happen on official GBA hardware and cart, difference 2 has been captured. Thanks to @Doomtune for the footage : https:// www.youtube.com/watch?v=vbo0i3vcJGw.

The video was taken under 30 fps but we were lucky that the frame was caught by the recording. Indeed, we can clearly spot the half-white frame, at around 1:00. Here’s a direct screenshot of it : https://imgur.com/a/9gE1c9X

.

Needless to say, the current rule doesn't apply anymore and needs to be fit for DCT. The questions here are, when do we deem the stage to begin ? What do we deem to be the earliest frame of its beginning ?

And if we’re gonna use the white screen as a visual cue for the beginning of the run, we need to be careful as to how to do it.

There aren’t many good answer to this question if you ask me, all of them have some flaw… either way, my personal pick is that we time the IL exclusively when the screen is fully white. I think we can agree that it feels right to refer to this moment as the moment the stage begins. It would make things simpler, because we only have to look for one thing, and that’s the screen being fully white. It's not perfect, but it's the least flawed option that I see.

That’s what I could come up with at least. Let me know what you think or come up with.

.

I’ll leave this conversation open for 7 days before we take a decision.

(2/2 because src forsaken character limit)

So you may be wondering how much time this could save over the current strategies.

Well, we don’t have too much data yet… but what we do have, is someone who attempted both runs !

Yup. @124412 ran both categories within a week. Their standard speedrun PB was a 2:03:59 (their first ever run - which also happens to completely decimate the previous world record. Good grief !)

How fast was their run with the underflow trick ? Well... they only did one run of each, but… they did pull away by over 8 minutes : https://www.speedrun.com/klonoaheroes/runs/y4v6qrky

It is a huge time-save, but it isn’t the game shattering glitch one would expect it to be. Though maybe that isn’t surprising. A large portion of the run is about walking through each stage, killing key enemies and defeating bosses quickly. Things that were already being efficiently done before. Such that there isn't as much potential to save time in the run.

Still, it is a new way of beating the game. We might need to create a new category where this glitch isn’t allowed.

The trick itself isn’t the problem. It isn't high entry level, and you only have to do it at the beginning of the run. On its own, it doesn’t change the speedrun much.

However, the way the game is played changes drastically as a result of this exploit. Namely : the act of avoiding experience points altogether. Some people may not want to play like this.

I think a good option would be to create a separate category called something like “any% no underflow”. Or something else, if you have a better name.

As the name implies, only runs that do not use the underflow trick would be allowed.

For the run timing points, it would use the same as that of any% :

“

• Timing starts upon the first frame of the fade-down process that applies to the screen after creating a new game.

• Timing ends upon the first frame of the HUD disappearing as the game screen fades to black, after defeating the final boss.

“

As for any%, it wouldn’t change. And for the existing submissions of any% that don’t use the underflow technique : we don’t have to move them, they can stay there. Of course, we could move or submit them to the new category as well, if the runner wants it.

That said, I also feel like we don’t particularly need to do anything here. This game has so little activity that I wonder if the category distinction even matters to add. If the lack of it would even bother people. Maybe once we have enough runners...

To be honest, this is probably what I'll do if nobody else speaks.

That's it. Tell me what you think !

And many thanks to @124412 for being the first person to make such a speedrun and for submitting it to us !

(This is a two-in-one message because of src's character limit. Sorry.....)

Hey everyone. Or well, there’s mostly tumbleweed rolling around. Regardless...

If you didn’t know, this game has an glitch/exploit where you can get unlimited skill points for Klonoa (and Guntz). Here’s a short reference video that shows the process : <>

Bear in mind that I won’t go over why the glitch works in some cases and why it doesn’t in others. Not that it’s super complicated or anything, but it isn’t necessary to be brought here and I want to keep things simple for this.

To pull off this glitch, you need to go to Klonoa's stats screen. As you do, you need to press both down and right on the exact same frame, on the last frame of the black screen before the stats screen shows up. Bear in mind you can only do this the first time you ever load the stats screen. If you fail it, it becomes impossible and you need to retry it via creating a new save-file.

It’s definitely finicky, but given that it’s at the first 10 seconds of the run, it’s actually not that difficult to do. More of a tedious obstacle than anything.

That said, this glitch can un-do itself. It’s actually really easy if you don’t pay attention, so please be careful when playing around with it.

Basically, if your current total amount of attributed points with the character is equal to the character’s level : it will automatically fix itself. So, make sure you use as many as possible.

You can do this glitch with Guntz as well, although it’s only possible when he joins you in world 2.

I’m not sure on how you do it though. Please reach up to me if you do. All I know for sure is you have to press right and up on the same frame. I don’t know the timing or if entering the menu in a specific way matters for it.

Unfortunately, this glitch doesn’t work with Pango and Guntz’s second encounter in world 5. Which is sad, but at least we got Klonoa.

Naturally, this changes the playstyle and routing of the whole game. Not only does exp become useless for Klonoa, levelling up is actually dangerous to him, something to be avoided so he doesn't lose the glitch.

You only need to focus on getting gold for items and equipments. There's nowhere as much need to fight monsters to grind resources on the way there, you can mostly zoom through stages and only defeat the required enemies for keys and specific items now.

This will be the main thread for holding discussions about adding rules, categories... This is a presentation post that might be updated as time goes on with extra information. Actual discussion will occur below.

Hey there. I made a post on the Door to Phantomile leaderboard that also relates to this game.

It is about a recent update of the duckstation emulator back around december of 2025. In it, the disk speed settings were overhauled, and loading-time differences were found between it and older versions of the emulator.

You can read more about it here : https://www.speedrun.com/posts/hc5lj

I apologize for not making this post any sooner. I had informed the Discord server two weeks ago but didn’t say anything here.

This concerns both this game and Beach Volleyball speedruns.

.

It has come to my attention that the duckstation emulator has received an update around december, upon which the disk speed settings were changed. Said update appears to have had a pretty big impact on the loading-times, which on at least one game speedrun (Rayman 1 100%, their main category) has lead to a timeloss of around a dozen seconds over their main category due to slower loading times.

Although we are used to using duck 2x (quad), a much lower disk speed setting than what that community is using, it’s still very possible that our loading times may be affected by whatever is happening there.

If you don't want to potentially lose time out of nowhere (or don't want to bother with any of this), you might want to stick to versions of the emulator dating from august 2021(*) to before december 2025.

So, it would be worth checking out for differences there. I know that @amoser was planning on looking into it, but more hands wouldn't hurt. Another way you can help out is by recording a speedrun on the updated versions of duckstation. This is so we can study the run and its loading-time data by the eye. If you do, make sure the video shows the disk speed settings.

Repeating myself here, but I’m still not too sure about what to do if this emulator update and the ones to follow that share this level of performance, were to become standard going forward. After all, it’s generally better to use the most recent versions of emulation programs, primarily so for better emulation quality. For the sake of argument, if in 5 years in the future we were still stuck with old versions of these emulators, I don’t know what that’d say about our speedruns.

We’ll just have to wait and see how things turn out.

I will give you an update once we know more.

.

(*) The day we did tests and ultimately enabled duckstation speedruns for DtP. It also was allowed for KBV but at a later time, on september 2023. While versions of the emulator weren’t tested with KBV prior to that date, my opinion is that they should probably be fine for it. With DtP speedrunning, since the day duckstation was allowed, no RTA differences were noticed or felt during runs, even after updating the program over the years. So it's unlikely that at that same moment, another game had seen a sizeable difference.

(Will be mainly talking about the input challenge here, although some ideas of it can be applied to the press variant as well)

I've thought about it some more since my post, and I would agree that its sheer complexity would make it less popular by default.

Not just that, but there is one aspect that fundamentally makes this challenge basically impossible to properly TAS for an astronomical amount of games : RNG.

Since you don't get any penalty for not doing anything, a lot of the game would revolve around waiting for massive amount of time to pass so the RNG lines up perfectly right. Watch for Rolling Rocks 0.5 A generating speed for 12h presses style, except you might be forced to do it nearly all of the time in the run. And all just to maybe save a couple inputs, among the tens of thousands the game might require to beat it anyway. Ew.

(As I side note, I do not know how one would properly keep count of the input count with games where you could save time or distance by performing specific tricks that rely on FPS changing.)

.

That said, I want to believe we can establish some rules so we can move forward without getting stomped by this unmanageable parameter. Let me list what I've been able to think :

• Assumptions. You may make theoretical estimations on how long it'd take to get the RNG, how much inputs it would save. You may then fast forward the TAS and move on to the next section. I think there is merit to this idea with some games, but it's definitely less fun. Leaving aside a myriad of definition problems, it basically amounts to us doing "what ifs" about the run.

• Count the amount of frames where you don't press any buttons, and try to limit those as well. I haven't explored this option much.

• Ban long waits. As if you had a level time limit. This would have to be defined on a game by game basis. That way, you're forced to do actions regularly enough. Of course, there may also be exceptions to this rule. I think this rule would be quite effective in solving this issue. I like this one.

• Go for the second best strat. Accept the fact that optimizing a given section to its fullest may take an impossible amount of time and carry on. Because at the end of the day we still need something to come out. If it's possible to beat a given game a little faster than normal, there must be ways to beat with less inputs than normal as well. It'll be better than nothing and it'll still be interesting. I like this one as well.

.

I can understand those who may feel that the inclusions of these rules and restrictions may reduce the integrity of the challenge, make it stand out less and sound more confusing overall. I do feel some of it as well.

But at the end of it all, in spite of its issues, I still believe there is incredible potential in this challenge idea. And it's still criminal to me that it isn't well known or played with at all.

We may need a popular speedrun content creator to make a video about it so we may finally get something started. Or at the very least, if you're feeling this way, kill the idea right there and then. But we just can't leave it unaddressed, doomed to obscurity.

I’ve been thinking of a TAS challenge concept that seems completely obscure as far as I could see.

We’ve all seen our share of these “zero A press” types challenges. What I’m surprised about, however, is that the next logical step of this challenge has seemingly never been explored : the lowest number of inputs challenge.

It is the goal of beating the game (or any other game objective) while using the least amount of inputs.

There are two ways you can go around this challenge :

• Each button press counts as one input. That said, you can hold the input for as long as you want as well as release it whenever you want, and the entire action will still count as a single input (similar to how that concept can be applied in SM64 0.5 A presses)
• The more complex and mind-numbing variant : every single frame of a button being pressed, counts. Do this for every button and add their count in order to calculate the totat input count of the “play”.

After digging for a bit on TASvideos, I was able to find a page that summarizes this topic and listes all known TASes that follows it (they use different terms for this stuff, I'll be switching to them) : https://tasvideos.org/Minimum

.

At first glimpse, 5 games have received TASes for being completed with the first variant of this lowest input challenge (“pressed”, according to TASvideos). And only one for the second kind (“inputted”).

That’s really not a lot of games and TASes. To make matters worse, only two of these TASes are watchable on youtube (the SMB and SMB3 ones). All the others you have to download what I assume to be input files that you can play over their respective emulators in order to see the run… and I don't have the time to fiddle with that.

.

So... I would like to see more of these kinds of runs. I think beating games by using the least amount of buttons in its most radical interpretation is an awesome concept that should be explored (even if it would be TAS exclusive for the most part).

For one, I think it's objectively funny to be able to poke fun of a game by beating it while playing it the "least". Imagine beating a really hard game or defeating very dynamic bosses where you're supposed to move around a lot except you're trying really hard at being lazy.

I’m also fully convinced that there are games out there that can look completely wacky and hilariously stupid if you do this with them. Like not only would the run not look fast at all, but the play would look like complete and utter dogshit to anyone's eyes and antithesis to the game's spirit. And it's all until you realize it's actually the fastest known method of progression there, and it just looks like that.

But I'm also sure there would be lots of beauty and things worth talking about with the various solutions that gets found for solving this challenge.

.

Again, I'd be doing this with my primary speedrun, if only I had more time...

.

Well, let me ask you then : in any of the games you run, under this scope, what would be the most efficient way of moving from point A to B ? (I’m mainly thinking about the inputted variant, since the press variant would be easy enough to figure out for most games.)

For instance, in SMB, I would suspect that one of the most efficient way of moving around is by building speed at first and then keep doing low one-frame jumps in order to keep the inertia and cross portions of the ground input-less. Because if you think about it, the normal, intended, sane way of movement (that is holding forward) sounds terrible here. One second of holding right in this 60 Hz of a game would mean a yield of 60 inputs, just like that. Yikes. We must strive to do better.

Let me know how the thinking experience would be like for you.

.

With this… I am hoping to raise awareness about this new kind of challenge. Or what I’d rather say : my burden is now your burden.

Hello everyone. This thread has been approved by Kellar.

I am a representative of the Misfit Mascot Relay. The MMR (for short) is a team-based relay race event that focuses on, as you might have guessed, underground mascot games and franchises.

During the event, teams of runners race to complete a gauntlet of games. The team who finishes before all the others wins the event. Everything is livestreamed on Twitch, broadcasted on the BSG speedruns channel. The whole event usually makes for 10+ hours of non-stop speedruns.

It's not all competition-driven however. We have it set up so fellow runners of the same game usually play at around the same time. This way, they can all chat and vibe while racing together, bringing warm community slice of life moments to everyone's way.

As indeed, our goal with this event isn't just to bring more light to the lesser known game and franchises of the old era, but also promote the speedruns of the games themselves along with their communities all alike.

As it's unfortunately quite common for speedrun communities to be isolated and not interact much with their neighbors... and we feel that there's a lot of missed opportunity for fun when it comes to that !

I strongly suggest taking a look at our official youtube channel so you get the right idea, in there you can find VODs of all of our previous editions : https://www.youtube.com/@MisfitMascotRelay/videos

Here's also a speedrun.com article we wrote more than a year ago, for the second edition of the event (about a year and a half ago) : https://www.speedrun.com/news/gq80lo80-the-misfit-mascot-relay-2nd-edition

As for today... we are currently preparing the Redux 2025 edition of the MMR, which will pan out during the 11-12th of october. This specific event is focused on running lesser-known games of the 2000-2010.

As it turns out... BLACK was selected and voted to be part of the event !

And with this, I am reaching up to you in order to find players who would be interested in participating in the MMR !

We would need at least three players, so we can fetch 3 teams. We can technically go up to 4 teams if there's enough people for it, but that's not our objective. And just so you know, we are open to players of different skill ranges !

I've heard that you guys tend to talk to each other through this site, and while that's fair, I wonder if if would be fine if we were to hop in on discord, for this, if just once ? For that, we could chat on Wagner Brasil's discord server, where a link can be found on the "Rules & Categories" thread.

After all, after a certain point, it'll become needed to join the official discord server of the MMR for technical event reasons. So it might be a good idea to get used to it by talking about this on Discord. But let me know what you think anyway.

Thank you so much for your attention either way, and I hope to hear more about you - and this game which I've never heard of before !

Hi. The title says it all. (picture of error message : https://imgur.com/a/aliq11L)

Has anyone else faced this problem ? It prevents me from updating one of the leaderboards I moderate. I haven't tested it extensively but I probably assume it's happening everywhere else ?

I was first editing variable A, only to receive this message, preventing me from saving. I then decided to add variable B who would essentially do what I wanted, and src allowed me to add it. However, after realizing I made a mistake, I went to edit variable B, to which I got the error message again preventing me from saving. Finally, I archived variable B and proceeded to create a variable C that was correctly set up this time... only to become unable to actually create it because yup you guessed right.

I have no idea what a value url stands for... if it isn't the one literally present in the browser link of the page, when you look at the "edit variable" page of an existing one ?

Change done. At least now, we have temporary rules in place about the use of external music.

We can have a complete discussion later.

Hey guys. Thread topic in the title.

.

The act of using external music – that is outside of the game - when doing runs is something that’s definitely uncommon in the speedrunning scene.

Most speedrun communities disallow it, either through a rule, or in the unsaid.

The leading reason usually being, it goes against the integrity of the game and speedrun.

.

But some communities do. Like the Sly Cooper community, with the first game of the series. Players there are allowed to play with music. The current world record does.

Well, this isn’t entirely exact. These players are still required to record the game locally without the music.

.

Alternatively, it would be very plausible to imagine certain games having it.

Games that don’t have any music on them, may allow players to use some as background audio. I don’t know of any at the top of my head, but they must exist.

Technically, in games where you can turn off their music, you could imagine communities allowing external music to be played.

As of today though, we have no community consensus about it. It was never discussed.

.

More importantly, there are no rules about it, which is a transparency problem.

Runners could accidentally submit speedruns with external music in them, not knowing whether it’s allowed or not. As a matter of fact, it’s happened.

The question of allowing or denying the use of external music in speedruns isn’t an easy one to answer. Because it’s a clash between run integrity and runner convenience.

While it is true that for the Klonoa games, thus far, we have been prioritizing run integrity the most, I think this issue can’t objectively be answered. It comes down to opinions.

.

For now, a temporary stance needs to be taken so that at least the runners can be informed.

I would suggest taking the safe route for the time being, by adding the following rule to every Klonoa games :

“

You may turn off the game music in your runs, but the sound effects must be kept on and be audible. You may also play with external music during your runs, but if you do, in your submission, you must provide a link of a version of the run without the external music.

“

(Note that as of right now, the only game in which you can turn off the music is KPRS.)

The rule would be changed as soon as the community collectively decides on something.

This thread will serve as a reminder of this topic.

Small update, Steam's FPS overlay (for PC only) received an overhaul, it's now finally possible to make the counter bigger :

https://steamcommunity.com/games/593110/announcements/detail/500576552635859540

Reply and help much appreciated. I can confirm your solution is working.

The fact that this has been an issue for years is truly disconcerting.

Thread content to be created soon...

Name says it all. It's been consistently happening for three days. (picture https://imgur.com/a/fmGs367)

It happens regardless of the series it's being done at. Normal forum threads such as this one and game boards work fine.

Am I the only one here ?

Oh yea,,,, uhhh... the Switch 2 is a thing now.

.

We're gonna most likely hold off on allowing submissions on it for until we can see speedrun footage of it for both games.

.

Here's sort of a basic runthrough of potential problems we could encounter with the platform :

.

• On one hand, the Switch 2 emulates Switch 1 games. Meaning that theoretically, there could be gameplay inconsistencies between it and other consoles. But I'm personally not too worried about it. I don't think it will matter for much.

And either way, it's an official way to play the game, so it would be unfair to oppose to it. Especially because it seems that KPRS runs way better on the new system. Switch had always been the worst platform to play the game on, so this is great news.

.

• On the other hand, I've heard the Switch 2 was capable of 120 FPS. And on the official website, it says that it's only for "supported games"... but they don't give a definition as to what exactly that is.

Since to my knowledge Switch 1 games would never go past 60 FPS, and all 120 FPS games would logically be Switch 2 exclusive, it seems like KPRS would be locked to 60 fps. But I'm not familiar with how you can configure the FPS on the system and I could be very wrong.

If playing the original games on 120 FPS is possible... then I'm not sure what we should do. We've always strived to run KPRS at 60 fps because it's what the game was designed around, and also because it ensures the gameplay and physics are how they should be like, especially when it comes to the DtP remaster. So my immediate opinion would be to also enforce 60 FPS play for switch 2.

But that asks two new questions : can you actually force the FPS to be 60 ? If it can't be fixed, if you're forced to play with whatever FPS the system gives you... then we have no choice but to accept it. At least it will give us an easy way out of this issue.

Now, if there's a way to fix the FPS to a specific value... would there then be a way to display the FPS counter on Switch 2, during gameplay, to show it ? Because if not, I'm not sure what we're gonna do then. I wouldn't know how to determine FPS through pure frame-by-frame analysis alone... does that mean we'd still be coerced into allowing Switch 2 KPRS runs to roam free, playing with whatever FPS the players feel like playing with ? Since we can't monitor it ?

Anyway, that's what comes off the tip of my head. Have you guys anything to add, have I missed anything important ?

Hey, I hope I'm not repeating a previous topic with this, I tried to make sure that wouldn't be the case but please be gentle with me if I did.

.

I would like to discuss about the idea of allowing a user to hide certain PBs/categories, in order to make their front page look more concise and relevant.

You would be able to select the PB/category you've participated in, label them as "irrelevant", and src would not display them on the user profile(*).

.

More broadly spoken, when user 2 scrolls through user 1's page, they may not be able to see the "irrelevant" PBs/categories user 1 chose to hide.

(*)However, player 2 may still be able to show them back, using the newly introduced "show irrelevant" slider (much like the "show obsolete" "show misc" that are in usage).

This feature only affects the basic, unfiltered view of the user profile, i.e what you see when you immediately click on someone's profile. You could always see those times back.

.

Let's say that you run game that has many different categories.

While you feel proud of your performance over certain categories, there are many more you don't care about as much.

But on your profile, you invariably see every single PBs, which is annoying if you wanted to only show the relevant ones.

I know someone who feels that way, so this isn't a fantasy story.

.

Their only solution, right now, is deleting all the submissions of the given category so they don't display. Yes, including WR times of unpopulated categories. That's not what we want.

.

I know that it is, in a way, a backwards approach to allow people to highlight their runs. But again, the lack of any reasonable solution here remains regrettable. Src has cared a lot about adding more customization options to the profile of users, so this idea would be a fair continuation of that.

.

Anyway, what do you guys think ?

I forgot to mention that the ILs currently only include the stages...

But you can read about why that is in through the link. It's better than you think.