I added brunovalads as a coauthor since: (1) I modified the input in TAStudio directly from the published BK2, and (2) This submission wouldn't exist without brunovalads's original submission.
There aren't too many improvements, but I went with different routes (both overworld and puzzle panels) whenever it doesn't lose time. The improvements are:
I have no name's post (nice username by the way) with a better solution to one of the panels. 40 frames saved.
Different route in the upper left section having less lag. 1 frame saved.
Approaching a panel head-on in the upper right section. 1 frame saved.
Different route in the upper right section. 1 frame saved.
Secondary goal: Maximize the number of 90-degree turns.
To make things more interesting, I made a secondary goal aiming to maximize the number of 90-degree turns in the route (without losing time).
What counts as a 90-degree turn? You can see it tracked in the comparison encode, but a 90-degree turn is defined as: Immediate (within 3 frames) change from horizontal movement to vertical movement OR immediate change from vertical movement to horizontal movement. So:
90-degree turns apply to both the overworld as well as the puzzle panels.
In the overworld, moving from a standstill, or turning on the spot to face a panel, does not count.
Turning the tracer into the "finish" spot counts.
Without losing time, I managed to finish with 315 turns, compared with the previous submission's 262.
P.S. I wasn't going to do this initially, but it bugs me to see such an obvious and easy-to-implement solution go unimplemented. No one else bothered to take advantage of such an easy opportunity for credit, and it's been almost a month. So I figured I would take the credit myself. Thanks everyone! (But thanks especially to I have no name for reporting this improvement.)
Edit: Added I have no name as a coauthor.
Samsara: I'd crack a joke here, but I'm feeling pretty wit.less at the moment. Judging.
Samsara: All the fun of tracing lines and making turns without the confusion of waking up in a daze next to a bottle of your own piss. Accepting as an improvement to the published run.
Joined: 4/7/2015
Posts: 331
Location: Porto Alegre, RS, Brazil
Oh nice! I never imagined it being improvable, i should've tried different combinations in all puzzles.
Thanks I have no name for the suggestion and FractalFusion for working on it.
My suggestion is adding I have no name as a co-author too, since his proposed solution had the greatest impact in this improvement.
Games are basically math with a visual representation of this math, that's why I make the scripts, to re-see games as math.
My things:
YouTube, GitHub, Pastebin, Twitter
Sounds good. I will add I have no name as a co-author.
I was surprised that you apparently never saw I have no name's post. I thought you saw it already and chose to ignore it.
By the way, there is another weird thing that got me looking into this game: Prior to your submission, I didn't know anything about The Witness (I may have heard about it 5 years ago but never paid attention.) So when I saw your TAS and then I have no name's post, I started seeing "improvements" in the lower section's puzzles (the ones where you separate differently-colored dots) that didn't actually exist, because I thought I knew the rules but actually didn't. These false improvements were actually what got me to download TheWit.nes and try them out myself. So I guess ignorance actually does pay off sometimes.
Joined: 4/7/2015
Posts: 331
Location: Porto Alegre, RS, Brazil
Yea I didn't saw lol, i don't check my emails in months and the updates in the discord server get buried fast sometimes, so i'm glad you did.
Great game, I always recommend.
Games are basically math with a visual representation of this math, that's why I make the scripts, to re-see games as math.
My things:
YouTube, GitHub, Pastebin, Twitter