Posts for Masterjun


Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
Original Title: What is wrong with the site management? Where are my awards? Why am I not credited when others are? Honestly what is going on? Why is everything so slow?
Is this a joke? One year ago, there was a TASer of the Decade 201X award poll, and the top 3 winners were meant to get an award. I was one of the winners but to this day I have not received any kind of icon. ? Huh? What's going on here? One year ago? And still nothing? I'm sure this must be some kind of mistake. But it's a big one. And it's growing. Hey just for reference, I also won an Exotic TAS award this year, and again, still no icon. It gets worse. It's not like everyone simultaneously forgot about it. No, changes are being discussed, and then nothing happens. This site about superhuman speed is balanced out in the supersloth management. It should reject itself as it's slower than its RTA counterpart. What kind of treatment is this? I can't show off my awards like this! Spawning a new branch of Super Mario World runs in 2012, I brought down the fastest time by 7:21 minutes. Then in 2013 I completely changed strategies and improved myself by 56 seconds, and in 2014 I saved 57 seconds with yet another strategy change. And then, in 2015 someone else optimized the same strategy by 10 frames and in 2019 another 8 frames. And those frames were enough to overshadow my credit: Last time I checked Mas[...] was not my username. I'm not planning to change it. But I see, so those are the more important authors. It's even easier to see the hierarchy when I'm completely gone:
  • Who makes the decision which of the authors deserve to be mentioned fully?
  • Which of them deserve to be mentioned partly?
  • And which authors are less important and can be omitted in the title?
Who do I ask if I want to exchange some of Doomsday31415's important nickname characters to show off more of mine? Now here's a crazy idea: If the whole author section can't fit into the title, replace it completely with [...], keeping it fair, imagine that! Just show the authors in the first line of the description. But you know what the problem with this suggestion is? It's different from how it currently is. This means it might be considered, but there will never be a change. Sorry. This page is phpBB © 2001, 2005. That's how modern it gets. What is HTTPS? Think about it from my point of view, I can't show off my awards, and I can't even be mentioned equally like other co-authors. With this in mind, I don't know if I can keep creating the high quality content I want to. And then it's not worth it to continue TASing. It might be time to stop this hobby. Management is slow. Changes don't happen ever. Nothing is worth a discussion. Does anyone think it's time for a big change? If you ask me: Yes, it definitely is.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
Alyosha wrote:
How much more difficult would it be to make it work without cheats? Or is it eve possible?
Now that it's been proven possible in theory, and now that there's a route, it's actually not too impossible. Sadly it will still require a lot of redo's of everything, so the bigger problem would be the motivation. After all, you're doing the whole run, and then hope that it works out in the end. I'm guessing like at least 20 runs until one works. (This one took me 2 months to make.) I certainly wouldn't do it. I only make the April Fools submissions.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
ah yes, only one enemy can act at a time, so decides to spend his time running into the instant death walls, oh no :)
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
DaSmileKat wrote:
I believe the 0.78 second SMB3 TAS has known improvements.
Wrong.
Potato wrote:
[4032] NES Super Mario Bros. 3 "game end glitch" by Masterjun & ais523 in 00:00.78 for its incredibly short length
Note, while the movie only consists of 47 video frames, it actually has 2201 input frames, so its complexity can be compared to a 36 second TAS. Of course not all of the inputs are equally important but it's something to keep in mind.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Post subject: counterpoint
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
Yeah this seems like quite complicated compared to simply selecting the FFmpeg writer in Bizhawk, selecting [Custom], and then entering your commands, for example -vf scale=iw*4:ih*4 -crf 18 -sws_flags neighbor -pix_fmt yuv420p -b:a 384k -f mp4 for 4x scale and basically visually lossless.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
p4wn3r wrote:
come up with an algorithm, better than plain brute force, to find two single digit 2x2 matrices whose product is obtained by concatenating the base 10 digits of each element.
I tried to solve this by not brute forcing all 8 numbers, but instead brute forcing only 4 numbers to get all 4-tuples (w,x,y,z) which solve the upper left number of the result. You test all 104=10,000 possibilities to find the 153 4-tuples. Then you can see that the same restrictions apply again for the lower right number, so exactly these 4-tuples appear again with flipped positions. At this point you can just test all 153*153=23,409 candidates for the remaining 2 restrictions. I found 101 solutions. Then I figured I'd test the same approach with 2-digit number multiplication resulting in 4-digit numbers (with leading 0s). After searching the 1004=100,000,000 possibilities, I found 24,209 4-tuples, which means another 586,075,681 tests for solutions. After a minute of running it, I found 36494 solutions. One example is {{10,19},{23,44}}*{{14,38},{46,82}}:
I decided to try and optimize it a bit further. I saw that the remaining 2 restrictions were the same as well. They had a separate set of 4-tuples, which could be searched for during the first loop. So I built the two tables in a way where I could use two specific values to lookup solutions for the other two. I go through all 4-tuples for (a,b,e,g), then I use (e,g) to lookup all possible (c,d), then I use (c,d) to look up all possible (f,h), and finally I lookup (f,h) to see if any single one is equal to (a,b). If it is, then a solution was found. In the 2-digit to 4-digit case, this reduces the tests needed in the second loop from 586,075,681 to 11,188,789, and takes about 10 seconds.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
Memory wrote:
...No I'm pretty sure it was rejected because it resulted in a broken loop. Sure that might be a result of jumping to a different point in credits, but that itself wasn't the reason.
Yeah I worded that badly. But it wasn't important to what I wanted to clear up anyway. I hope my point was clear that a jump to a different point in the credits can both be accepted or rejected, depending on the specific case.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
EZGames69 wrote:
Plenty of Game End Glitches have existed where the location of the credits they warp to have been different, super mario world is one example. This: http://tasvideos.org/1945M.html is how it used to look, where it warped directly to the end screen, compared to how it is now: http://tasvideos.org/3989M.html
If you're already linking my runs, you should not forget the rejected one, rejected because the jump to the credits location was different. Edit: Sorry. My point wasn't clear here, read my next reply below for elaboration.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
Invariel wrote:
That wasn't a request, Dacicus. Our team, the sixth team to be officially registered, is called "Team 5", and that's our logo.
LOL I love this.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
You forgot the stun timer (sprite table at address $1540). The timer needs to be non-zero, and then hit zero just after spitting it out.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
I can't believe you decided to submit this spliced run. A wonderful run Malleo! I love how far the cooperation between TASers, glitch hunters, and RTAers has come in this game. I enjoyed this TAS. Thanks for making it! :D
Btw, trying to turn that static into a proper image with RNG manipulation is probably a lost cause if you're aiming for something big. If you have an RNG with 232 states, then you can arbitrarily control (i.e. creating any combination of) 32 bits, so 32 pixels at best. Adding a single pixel will leave you with only 50% possible images, another one to 25%, etc. Of course, there are a lot of sets of x pixels on screen, so the probability is multiplied again, and you have a lot of frames, so add those together. But if you look closely, the static is actually a smaller pattern repeated, and a lot of frames are mostly transparent. And most importantly, the exponential growth of the unlikeliness is much bigger than the linear growth if having multiple areas or frames. In short, a 8x8 binary image is 64 bits, so you have a 232/264 chance of your image being possible to manipulate. Which is ~0.00000002%. The takeaway is: Keep in mind the rough limit of 32 pixels with a 232 state RNG.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
I was just about to edit this part in. Yes indeed it is a bit tougher than other mice. I did not notice this before reading the reviews myself. I would not say the wheel click is useless and making the mouse a returnable item. I also wouldn't say it's a "relatively herculean task". But I do say that it's noticeable enough for me to say "Yeah, it's tougher". (But it's not enough for me personally to take a risk and try out a completely new mouse once this one breaks.)
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
I'm using the Logitech M500 Corded Mouse, which seems to fit most of the criteria. (I'm strictly talking about the M500, not the M500s.) - Fits large hands nicely, should have a nice size to it. It's not as small as all these really small ones. - Has some weight to it, none of this lightweight junk that you can barely tell you're moving it. Yes. - Has a body, so you can actually wrap your hand around it. Yes. - Has a decent wheel, scrolls nice and smoothly and can click. Yes. The wheel is great. There is a mechanical button which unlocks the wheel allowing for infinite scroll (no need to ever use it, and it doesn't get in the way). When locked the scrolling is precise. - Wired, USB. Yes. - At least two buttons besides the wheel. Two buttons. Or do the "tilt scroll" buttons (left/right) count as well? In any case, the specification says 7 buttons. - 1000 DPI or higher. 1000 DPI. - No craziness around the sides, dozens of buttons. No craziness. - Primarily for right hand use, but should be able to be used left handed as well. It's not a symmetrical mouse, so this point is not an easy yes. I just tried using it with my left hand. If you try only putting two fingers on top, your ring finger will accidentally hit the side buttons. Putting the ring finger next to the middle finger on top of the button works fine. - Price is $50 or less (really shouldn't be more than $15, but junk these days...) Yes. - Good quality, will last several years without issues. Amazon tells me I bought it in January 2017, and I've been using it since then. I've noticed no difference except for all the text on the mouse fading. (I plan to buy this one again once my current one stops working properly.) (Also, I absolutely love how a bunch of one-star reviews on amazon are people not being careful with the package and cutting the cord off. I guess that's a thing to be wary about?)
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
Warp wrote:
What's the expected average number of times you'll need to toss the coin in order to get 10 of the same side consecutively?
I once learned this one method one can use to avoid having to collapse an infinite series when calculating expected values. First, the basics of expected values. Let's toss a die with numbers 1 to 6. You calculate the expected value by multiplying the probabilities with their outcome: E = (1/6)*1 + (1/6)*2 + (1/6)*3 + (1/6)*4 + (1/6)*5 + (1/6)*6 = 3.5 Now let's toss a coin and ask for the expected number of tosses until the first Heads. There is a 1/2 probability it's the first one, so 1 toss. And then there is a 1/2 probability it's something higher than 1. E = (1/2)*1 + (1/2)*X You could take that X and expand it into another split case, like X = (1/2)*2 + (1/2)*Y (we multiply the first probability with 2 since we're now at the second toss). Of course, this leads to an infinite series. But, here comes the trick. Instead of expanding it into more and more split cases, use a self reference somehow. E = (1/2)*1 + (1/2)*(E+1) If the first toss wasn't Heads, then we're back at the start with the same expected value, except we already have one toss. It's then simple to calculate that E=2, the expected number of tosses until the first Heads. Let's use a simplified version of the question. Instead of 10 consecutive tosses, we ask for the expected number of tosses until you have 5 consecutive ones. I split the total expected value up into Heads or Tails separately. E_h := the expected value for number of additional tosses until 5 consecutive Heads, after already tossing one Heads. And E_t respectively. (Due to Heads and Tails both having a chance of 0.5, it's intuitive that E_h = E_t) Then it's easy to see that our total E is: E = (1/2)*(E_h+1) + (1/2)*(E_t+1) For E_h (or E_t) we now use the trick above.
We start with one H tossed:
          HT              HHT            HHHT             HHHHT          HHHHH
E_h = (1/2)*(E_t+1) + (1/4)*(E_t+2) + (1/8)*(E_t+3) + (1/16)*(E_t+4) + (1/16)*4
(Note how E_h references E_t, and vice versa, but we can write them both out, insert them, and see that they're indeed the same, I left this out. This also means that it's just E = E_h+1) E_h = (15/16)*E_h + (30/16) E_h = 30 E = 31 It's easy to see how that long sum for E_h gets longer with a higher toss requirement, but it can be simplified. Checking the answer for numbers 2,3,4,5 gives 2,6,14,30, so you might guess it's 2^n - 2. E = 2^n - 2 + 1 E = 2^n - 1 So for n=10 it is indeed 1023 tosses.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
1) You can find the code for the 3 supported palette conversions in BizHawk.Emulation.Cores/Consoles/Nintendo/SNES/SnesColors.cs. 2) The SNES Graphics Debugger has a "User Backdrop" checkbox which allows you to set a background color for the debugger display. Alternatively the backdrop color for backgrounds is the 16-bit value at CGRAM address 0, and the Fixed Color Data at register $2132 is probably also important.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
Dacicus wrote:
Why TC version 4.7.7, though?
It had more optional applications to choose from.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
Whoa! I will say this has been an awesome decade for TASing. All the new emulators, actual console verifications, interesting thread discussions, and the higher entertainment standards have made it really fun to be a TASer. To be voted one of the TASers of the decade on this site is a big honor. I'm really happy! Thank you, I appreciate this.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
Project files aren't in the registry. They are the .prj files in the lsnes config folder (on Windows, that should be %appdata%/lsnes). Those are the project files lsnes tries to enumerate. They're just text files so you can open them to check their name and main directory. So just delete the ones you don't want (and you might want to delete their main directory as well).
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
I don't know a whole lot about Celeste mechanics, so I don't really know what the suggested screenshot is supposed to show other than being quite close to spikes. Seeing how the red berries are a big part of the run and how neat they look when they follow you, I decided to suggest two screenshots of my own: Nice movie!
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
When it says "compute the greatest multiple of 98765432 that's less than or equal to 123456789", it actually means an integer multiple. In other words, the division is turned into multiple integer divisions, which are easy because they compute the result digit by digit, thus only allowing for 1 of 10 possible digits per step. With big divisors, it makes sense to write down the first 9 multiples:
*1  98765432
*2 197530864
*3 296296296
*4 395061728
*5 493827160
*6 592592592
*7 691358024
*8 790123456
*9 888888888
With this, you no longer need to "compute" the greatest usable multiple, instead you simply check by looking at the numbers. So now you can do the long division, which now requires nothing more than comparisons and subtractions:
123456789:98765432 = 1.2499...
 98765432     <------/ ||||
 --------              ||||
 24691357              ||||
 197530864    <--------/|||
 ---------              |||
  49382706              |||
  395061728   <---------/||
  ---------              ||
   98765332              ||
   888888888  <----------/|
   ---------              |
    98764432              |
    888888888 <-----------/
    ...
Of course, the subtraction result can never be 98765432 or over (otherwise you chose the wrong multiple). So at some point the subtraction result repeats, meaning you found the repeating part of the decimal.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
Graphics Debugger is only available with the BSNES core. Enable it under: Config -> Cores -> SNES -> BSNES. After rebooting the core, it should appear in SNES -> Graphics Debugger.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
I agree with the question. How the hell is anyone enjoying the game?
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
InfamousKnight wrote:
She only has 20 dollar bills in a safe. She counts them like this: 4 20s according to her makes 100. As she counts this way, it ends up being 2180. So she counts 4 20 dollar bills and puts in her calculator 100 instead of 80 which was the mistake made. And when she was done, she ended up with 2180 under the belief that 4 20 dollar bills make 100 dollars. Realizing that 4 20 dollar bills doesn't make 100 dollars, what does she actually have?
I asked for the solution to this puzzle, so here it is: <InfamousKnight> The answer is 1810 <Masterjun> we need an explanation about how only 20 dollar bills can add up to $1810, that's quite the interesting puzzle indeed <InfamousKnight> Oh, that was actually an about guess. Its probably more like 1820 <Masterjun> so another explanation needed is how to get from the information in the puzzle to that answer <Masterjun> like, why isn't the answer 1800 or 1840 <InfamousKnight> I dont know :/
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)