Posts for Masterjun


Masterjun
He/Him
Experienced Forum User, Published Author, Site Developer, Skilled player (1971)
Joined: 10/12/2010
Posts: 1179
Location: Germany
LOL
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
Today I want to present a little proof of concept I made: An automatic movie encoding workflow running completely remotely on a different server! I will quote the information in the notebook:
1. Upload your game file and movie file 2. Edit the first cell accordingly 3. Run every cell in order. Current limitations (due to limited BizHawk command line arguments): - Can only start the encode at movie start - Can't give ffmpeg encode commands - Has to manually enter encode length - Can't fast-forward
It's using a Google Colab notebook. The link is here: https://colab.research.google.com/drive/17kqGyOQfsnj-fmjiC_8r-Mfa8qTxMbIN Again, only a proof of concept, the biggest missing feature is currently the ffmpeg command, making the thing not very useful at all. But hey, I just wanted to share that this is technically possible.
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
Yo Noxxa, thanks for doing all the things and stuff over the years! :D (Especially the awards lol.)
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
Since there was discussion about Youtube Tags, I thought I'd mention something. Over the years the tags themselves became less and less important. Not only were the tags eventually hidden for viewers (they used to be displayed under the description). But now the tags are also hidden for uploaders, so you have to actually search for them in the UI if you want to add them.
Youtube itself wrote:
Tags can be useful if the content of your video is commonly misspelled. Otherwise, tags play a minimal role in your video's discovery.
I'd not be surprised if Youtube decided to remove tags completely. So in terms of discussions, it's worth considering removing the tags field entirely, and instead rethink their purpose and the approach to be taken here.
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
So I'm also going against the trend here. This run is technically impressive. However, it severely lacks focus. The branch name isn't so much "glitchfest/playaround", but instead more like "all glitches showcase" which fits more to a documentation type video instead of a superplay. Back in March 2018 (4 years ago) when the first 20 minute WIP of a run like this was released, I wrote this:
Masterjun: wow, that movie is slower paced than I would have liked Masterjun: it works if the plan is to quickly proceed to bowser, but I can't see it getting 96 exits and staying entertaining Masterjun: (considering we're looking at a 3 hour TAS then)
Reading the other feedback in this thread, it's possible to see the issue. "Very nice run, but I'm not going to watch it all, haha" is a big red flag for a movie that aims to be entertaining. Instead of choosing the best glitches that make for an exciting experience, this run follows a bucket list of levels and areas where a unique glitch can happen and tries to cover up the time travelling there by repeating the same glitches over and over again. The fact that this run has to pause the game every time a graphics glitch happens shows how much it wants to make sure it shows off every possible glitch, instead of showing off the entertaining ones. While I think a run like this between 1-2 hours can be entertaining, it needs to be a lot more polished. To keep the viewer on their toes, it needs to not show off everything all from the start. It needs to be an gradual increase in craziness. Compare with [1438] SNES International Superstar Soccer Deluxe "playaround" by Marcokarty in 15:24.38. That run doesn't even show off it can play outside the field until 50% into the run. And afterwards it plays around with the score count for the first time. So here my thoughts while watching it and some suggestions. Don't spend a lot of time in the same level. Seeing the same scenery and hearing the same music can quickly become boring. After six minutes you are done with 2 levels, where in the speedrun in the same time 9 levels were finished. The mosaic glitch really didn't need to go on for 14 minutes now, did it? Watching that part wasn't fun for my eyes. Please stop with the mushroom in reverse spam. We get it, you can mash the Select button at 30Hz, but the sound effect is really annoying after the 20th time. There are a lot of places where you glitch into a floor or a wall and then just display that fact for a while. It's fine a few times, but at some point the mere fact you can clip into things isn't exciting anymore. With nothing else to show this was really repetitive. For a viewer "I don't know what's going on" can only be entertaining for so long. For example at 47:20 where Mario turns invisible. Something is going on, but this goes on for 50 seconds until finally something happens that can be understood. Repeating levels just because they have more glitches to show is just a plain poor decision. This goes to show the goal was more showing off every possibility instead of making something with a nice flow. But what might have been the most disappointing part of this TAS was the lack of a bowser fight. The climax of every game, the final boss, was just skipped in favor of a game end glitch type ending. The complete lack of ending music due to the glitch was just sad. In general it seems people enjoyed parts of this run, but didn't want to watch the whole thing. So I'd say if this does get accepted I can see it being obsoleted with a run that ironically has less content, but more focused and polished. For the question "Did you like watching this movie?" I started off with Yes indeed I like it, then a Meh it drags a bit I hope you have something up your sleeve for the next 2 hours, and then a No this is too long.
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 always dislike when lists do this. It might sound reasonable and easy to get used to, but this more heavily impacts people who aren't too familiar with the English language. Having to know exactly which words at the start of a game title to leave out when searching through the list is just troublesome or even annoying. I'd ask to please just leave the title and sort position as 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
Patashu wrote:
was it strictly necessary to flip the game 180 degrees in the ghost house? I kind of assumed that since you had a cape you could do anything.
I mean I have total control so it's not like the cape increased my movement pool. I did mention in the submission text how a big problem was aligning all the individual Mario tiles. A cape flight has even more tiles and more animation frames. So my decision was to flip the world instead of fixing 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
If the site detects nonsense it will default to TimeSpan.MaxValue. There is even an intentionally designed test for exactly this value here.
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 count 948 moves in this submission. Heh, my TAS attempt finished the game in 881 moves. But I want to see someone manage to manipulate RNG enough to get the minimum, 519 moves.
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
Haha, this was too much fun, wow. My solution only has to check the fibonacci numbers downwards starting from the biggest one smaller than the number we aim for. I can't really write it down mathematically, but my three key insights were these:
  1. An equation for a number in position j of a sequence starting with a and b can be written as F(j-1)*a + F(j)*b where F is the fibonacci sequence. This means checking from above going down you can't miss an optimal solution.
  2. Subsequent numbers in fibonacci sequences are always coprime.
  3. The modular multiplicative inverse of a fibonacci number to the modulus of its following fibonacci number is always another fibonacci number.
I'm not even sure if all these are needed or if this is a good solution, but I like it.
Language: Python

def check(n): # generate fibonacci sequence, stop when bigger than n fib = [0,1] while fib[-1]<n: fib.append(fib[-1]+fib[-2]) # start at biggest fibonacci number smaller than n and iterate down for j in range(len(fib)-2,1,-1): res = (fib[j-1-(j%2)]*n)%fib[j] # magic if fib[j-1]*res<=n: break # a solution was found a = int(res) b = int((n-(res*fib[j-1]))/fib[j]) if b==0: # we dont allow 0 solutions, even if I enjoy those b=a j-=2 assert(n==a*fib[j-1]+b*fib[j]) # it really is a solution seq = [a,b] for i in range(j-1): seq.append(seq[i]+seq[i+1]) return seq
>>> check(7)
[2, 1, 3, 4, 7]
>>> check(81)
[3, 2, 5, 7, 12, 19, 31, 50, 81]
>>> check(179)
[11, 7, 18, 25, 43, 68, 111, 179]
>>> check(102334154)
[9349, 9349, 18698, 28047, 46745, 74792, 121537, 196329, 317866,
514195, 832061, 1346256, 2178317, 3524573, 5702890, 9227463,
14930353, 24157816, 39088169, 63245985, 102334154]
>>> check(102334155)
[1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, 610, 987, 1597,
2584, 4181, 6765, 10946, 17711, 28657, 46368, 75025, 121393,
196418, 317811, 514229, 832040, 1346269, 2178309, 3524578,
5702887, 9227465, 14930352, 24157817, 39088169, 63245986, 102334155]
If 0 were allowed, the sequence starting with 9349, 9349, ..., could be rewritten as starting with 9349, 0, 9349, 9349, ..., easily adding two extra sequence entries! :D
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
Juarez wrote:
Can't submit movie - An general error occured while processing the movie file
It looks like the site has an issue with parsing .gmv files compressed into .zip files. A quick solution because your file is small enough: You can submit it in an uncompressed .zip file.
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
Count me out. I personally wouldn't want this to be done to my old movies. If there's a superior movie existing then digging out and publishing an old movie feels like pointing with your finger and saying look at this worse run, it's obsoleted! I do think authors might feel something like pride about their runs, so taking pity on now suboptimal movies feels... wrong. At least to me.
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
The "Posts since last visit" feature was in a broken state for a while now, and the recent Activity Indicators addition did nothing to change the behavior of that page. There was a discussion a while back about just removing the page once we had better-working alternatives, but apparently nobody got around around to actually removing the button. I'd say this broken feature is unlikely to change until the hardware is upgraded.
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
Yes, tracking the posts themselves is a whole different feature. It requires a larger amount of tracking, and for that the hardware we have is not prepared yet. I do hope this new feature is not that "less useful" than the old site.
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
Reminder that there are several different backup options already available. You should make use of them.
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:
I came across a post that asked the "correct" value of 8 / 2 * 4
This reminds me of one of the first things my teacher said to our class when we were taught fractions: "From now on you are not allowed to use / anymore. Only use ― now." After this we would be reminded of it each time we would make that "mistake", and eventually a simple reminder turned into a real mistake like a wrong calculation and we would lose points. And indeed, this resolves the ambiguity, as this: 8 ― * 2 4 is simply not the same as this: 8 ―― 4 * 2 In retrospect, that's a neat solution here. You can't really explain the intricacies of mathematics about divisions being a side-effect of multiplications having inverses coming from the multiplicative identity or something like that. Division makes intuitive sense because you have objects, and then you give equal amounts of objects to multiple people. That's a practical application of dividing, and how it's taught from the very beginning. It'd be very difficult to tell people division doesn't exist. Making people avoid writing something ambiguous, that's a lot easier. But of course, this division-rewriting idea doesn't solve the problem of what people think a "correct" value of 8 / 2 * 4 is. Whenever people personally asked me what I think the correct solution is (which apparently people care about), I always answer that the term is ill-formed. Then I give an ambiguity in something more familiar: language. "John saw the man on the mountain with a telescope." (Who was on the mountain, who had the telescope, etc.). I say that it's the same in this math situation. People who are more comfortable with ambiguities in language because well humans invented it and humans aren't perfect, but want math to be this rigorous ideal because math just exists. Those people forget humans also invented math notation.
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
That doesn't contradict anything I've said though.
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 not staff, but I posted in here! I stand by everything I've said! Considering the incident, I think it would be foolish to migrate to Discord and drop IRC support. Most of the posts I made here were asking people to give proper reasons why Discord server maintenance would be worth it.
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:
The Posts since last visit option on the forum does not work for me. It leads to a page saying No posts since last visit. This is obviously incorrect, since the main forum page says that the last post in Workbench was 1 hour ago, and I last logged into the forum more than 1 hour ago.
Just for reference, we now have a Latest Posts page, in case the last visit doesn't work again.
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, so the 5 minute run only restricted itself to avoid the escape key, since this was the only way to trigger the gamestate manipulation. I think now that there's a different way to trigger it, it calls into question whether it should be allowed by itself or not. I personally prefer the default category (branchless) to avoid the glitch, because it feels memory-corruption'esque. It doesn't seem like what viewers would expect. The other runs all still physically reach the locations of the crewmates and the end location, albeit with weird teleportations and going through screens, but those features are themselves part of the game. It feels like a category with restrictions like in this run can quickly become a second game end glitch visually. I'm not the one to decide where to put the distinction between game end glitch branch and default branch, but my opinion is that the gamestate manipulation falls into the game end glitch area.
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
So, I'm not quite sure what's going on here, the submission text is all over the place with versions and patches and what glitches were actually used and which ones are patched. From what I gathered, this submission uses the glitch that advances the game-state like in the current game end glitch movie [2611] Windows VVVVVV "game end glitch" by Masterjun in 00:45.33, but in a less powerful way because it's somewhat patched now? Then an argument could be made that this submission shouldn't obsolete the current run of the default category, since the current one avoids using that glitch (as it would be the game end glitch run otherwise). In that case it feels like this run is an inferior version of the published game end glitch run. But as I said I didn't quite get everything from the submission text, so maybe the author can explain where this run fits in between the existing branches.
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 approve. Now I'm not a staff member, but I experienced several instances where I requested a simple thing, or suggested a fix, or even made a poll about an issue. When they got to Nach it was always either a "make a thread about it" and then an answer saying the whole system needs to be reworked for it. But it usually never got to Nach in the first place. Frustrating is the correct term. That was 7 years ago, and it seems like the behavior hasn't changed. So I'm glad changes happen. Things are worth discussing again.
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
The top and bottom offset comes from a console quirk. Here is a quote from a hardware register page:
Note that many games will set their vertical scroll values to -1 rather than 0. This is bacause the SNES loads OBJ data for each scanline during the previous scanline. The very first line, though, wouldn't have any OBJ data loaded! So the SNES doesn't actually output scanline 0, although it does everything to render it. These games want the first line of their tilemap to be the first line output, so they set their VOFS registers in this manner. Note that an interlace screen needs -2 rather than -1 to properly correct for the missing line 0 (and an emulator would need to add 2 instead of 1 to account for this).
In other words, if your game doesn't account for this offset, you will miss the first scanline and have an extra one on the bottom.
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
The borders in Final Fantasy IV are not an emulator glitch or emulator display decision. They are intended and deliberately set by the game. The game makes use of the Window hardware registers in the range $2123 to $212F. I suppose you can debug the game and check where the values are set. One example is $15:C968 where $2126 is set to 0x01 (left border), and 5 bytes later $2127 is set to 0xFE (right border), but there are more places.
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
Image doesn't load, it's an infinite redirect.
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)