My personal code and solution to the Synacor Challenge from 2012 OSCON.

Overview

Synacor OSCON Challenge Solution (2012)

This repository contains my code and solution to solve the Synacor OSCON 2012 Challenge.

If you are interested in checking out or trying the challenge for yourself, it can be found online still here:

https://challenge.synacor.com/

Notes

Firstly, please understand this is an old challenge. I am not the first to solve it, not even close, and this was solely done because a friend suggested it to me on Discord this past week. I never saw the challenge before and since it involved implementation of a VM, it was something I was interested in checking out since it has been a topic I've been involved in recently.

Next, the challenge is still online and fully functional. Because of that, it is important to note that if you do sign up and decide to try the challenge, the information in this solution will work but the flags (codes) you need will be different. The challenge generates unique flags for each player. (The challenge.bin data file is unique to each player.) If you try to use my flags, you will get an error.

Lastly, I used this challenge as a means to continue with my progress of learning Python. So please excuse the messy code and probably poor / old means of which I did some things. I'm sure there are much better ways to code various things I made, but I am still fairly new to Python.

Repository Information

You can read my full solution here: Full Solution

In order to solve the challenge, the main task you are given is to implement a virtual machine that can emulate the given opcodes found within the challenge arch-spec file. To handle this part of the challenge, and assisting with other parts, I wrote the virtual machine and a disassembler for the binary data file in Python.

Throughout the challenge, once the VM is functional, there are puzzles to be solved. The three puzzles all required their own implementation of code to be solved. Two of the puzzles I was able to solve in Python, however the other was too slow to implement in Python alone. Instead, I opt'd to use C++ for that one instead. (I made a Python implementation using ghetto threading, but it's ugly and slow so not worth including.)

The first puzzle is within the Ruins area of the game. My solver for that can be found here:

The next puzzle, which required the C++ implementation to not be ungodly slow, is for the teleporter item puzzle. That can be found here:

The final puzzle, in the Vault area, can be solved with my solution here:

Other files included in the repo are:

Challenge Information

== Synacor Challenge ==
In this challenge, your job is to use this architecture spec to create a
virtual machine capable of running the included binary.  Along the way,
you will find codes; submit these to the challenge website to track
your progress.  Good luck!


== architecture ==
- three storage regions
  - memory with 15-bit address space storing 16-bit values
  - eight registers
  - an unbounded stack which holds individual 16-bit values
- all numbers are unsigned integers 0..32767 (15-bit)
- all math is modulo 32768; 32758 + 15 => 5

== binary format ==
- each number is stored as a 16-bit little-endian pair (low byte, high byte)
- numbers 0..32767 mean a literal value
- numbers 32768..32775 instead mean registers 0..7
- numbers 32776..65535 are invalid
- programs are loaded into memory starting at address 0
- address 0 is the first 16-bit value, address 1 is the second 16-bit value, etc

== execution ==
- After an operation is executed, the next instruction to read is immediately after the last argument of the current operation.  If a jump was performed, the next operation is instead the exact destination of the jump.
- Encountering a register as an operation argument should be taken as reading from the register or setting into the register as appropriate.

== hints ==
- Start with operations 0, 19, and 21.
- Here's a code for the challenge website: fNCoeXxLEawt
- The program "9,32768,32769,4,19,32768" occupies six memory addresses and should:
  - Store into register 0 the sum of 4 and the value contained in register 1.
  - Output to the terminal the character with the ascii code contained in register 0.

== opcode listing ==
halt: 0
  stop execution and terminate the program
set: 1 a b
  set register <a> to the value of <b>
push: 2 a
  push <a> onto the stack
pop: 3 a
  remove the top element from the stack and write it into <a>; empty stack = error
eq: 4 a b c
  set <a> to 1 if <b> is equal to <c>; set it to 0 otherwise
gt: 5 a b c
  set <a> to 1 if <b> is greater than <c>; set it to 0 otherwise
jmp: 6 a
  jump to <a>
jt: 7 a b
  if <a> is nonzero, jump to <b>
jf: 8 a b
  if <a> is zero, jump to <b>
add: 9 a b c
  assign into <a> the sum of <b> and <c> (modulo 32768)
mult: 10 a b c
  store into <a> the product of <b> and <c> (modulo 32768)
mod: 11 a b c
  store into <a> the remainder of <b> divided by <c>
and: 12 a b c
  stores into <a> the bitwise and of <b> and <c>
or: 13 a b c
  stores into <a> the bitwise or of <b> and <c>
not: 14 a b
  stores 15-bit bitwise inverse of <b> in <a>
rmem: 15 a b
  read memory at address <b> and write it to <a>
wmem: 16 a b
  write the value from <b> into memory at address <a>
call: 17 a
  write the address of the next instruction to the stack and jump to <a>
ret: 18
  remove the top element from the stack and jump to it; empty stack = halt
out: 19 a
  write the character represented by ascii code <a> to the terminal
in: 20 a
  read a character from the terminal and write its ascii code to <a>; it can be assumed that once input starts, it will continue until a newline is encountered; this means that you can safely read whole lines from the keyboard and trust that they will be fully read
noop: 21
  no operation
Owner
:rainbow: Self-taught programmer / reverse engineer. Game hacker / modder. Looking for support for any of my projects? Check my homepage.
A criticism of a recent paper on buggy image downsampling methods in popular image processing and deep learning libraries.

A criticism of a recent paper on buggy image downsampling methods in popular image processing and deep learning libraries.

70 Jul 12, 2022
Transparent Transformer Segmentation

Transparent Transformer Segmentation Introduction This repository contains the data and code for IJCAI 2021 paper Segmenting transparent object in the

谢恩泽 140 Jan 02, 2023
Towers of Babel: Combining Images, Language, and 3D Geometry for Learning Multimodal Vision. ICCV 2021.

Towers of Babel: Combining Images, Language, and 3D Geometry for Learning Multimodal Vision Download links and PyTorch implementation of "Towers of Ba

Blakey Wu 40 Dec 14, 2022
Code for "Continuous-Time Meta-Learning with Forward Mode Differentiation" (ICLR 2022)

Continuous-Time Meta-Learning with Forward Mode Differentiation ICLR 2022 (Spotlight) - Installation - Example - Citation This repository contains the

Tristan Deleu 25 Oct 20, 2022
ONNX Command-Line Toolbox

ONNX Command Line Toolbox Aims to improve your experience of investigating ONNX models. Use it like onnx infershape /path/to/model.onnx. (See the usag

黎明灰烬 (王振华 Zhenhua WANG) 23 Nov 13, 2022
The code repository for "RCNet: Reverse Feature Pyramid and Cross-scale Shift Network for Object Detection" (ACM MM'21)

RCNet: Reverse Feature Pyramid and Cross-scale Shift Network for Object Detection (ACM MM'21) By Zhuofan Zong, Qianggang Cao, Biao Leng Introduction F

TempleX 9 Jul 30, 2022
POT : Python Optimal Transport

POT: Python Optimal Transport This open source Python library provide several solvers for optimization problems related to Optimal Transport for signa

Python Optimal Transport 1.7k Dec 31, 2022
[ICCV'21] UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for Multi-View Reconstruction

UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for Multi-View Reconstruction Project Page | Paper | Supplementary | Video This reposit

331 Dec 28, 2022
A flexible and extensible framework for gait recognition.

A flexible and extensible framework for gait recognition. You can focus on designing your own models and comparing with state-of-the-arts easily with the help of OpenGait.

Shiqi Yu 335 Dec 22, 2022
Plugin adapted from Ultralytics to bring YOLOv5 into Napari

napari-yolov5 Plugin adapted from Ultralytics to bring YOLOv5 into Napari. Training and detection can be done using the GUI. Training dataset must be

2 May 05, 2022
Computational modelling of ray propagation through optical elements using the principles of geometric optics (Ray Tracer)

Computational modelling of ray propagation through optical elements using the principles of geometric optics (Ray Tracer) Introduction By applying the

Son Gyo Jung 1 Jul 09, 2022
The repo of Feedback Networks, CVPR17

Feedback Networks http://feedbacknet.stanford.edu/ Paper: Feedback Networks, CVPR 2017. Amir R. Zamir*,Te-Lin Wu*, Lin Sun, William B. Shen, Bertram E

Stanford Vision and Learning Lab 87 Nov 19, 2022
MASA-SR: Matching Acceleration and Spatial Adaptation for Reference-Based Image Super-Resolution (CVPR2021)

MASA-SR Official PyTorch implementation of our CVPR2021 paper MASA-SR: Matching Acceleration and Spatial Adaptation for Reference-Based Image Super-Re

DV Lab 126 Dec 20, 2022
Conflict-aware Inference of Python Compatible Runtime Environments with Domain Knowledge Graph, ICSE 2022

PyCRE Conflict-aware Inference of Python Compatible Runtime Environments with Domain Knowledge Graph, ICSE 2022 Dependencies This project is developed

<a href=[email protected]"> 7 May 06, 2022
This is 2nd term discrete maths project done by UCU students that uses backtracking to solve various problems.

Backtracking Project Sponsors This is a project made by UCU students: Olha Liuba - crossword solver implementation Hanna Yershova - sudoku solver impl

Dasha 4 Oct 17, 2021
Unsupervised Pre-training for Person Re-identification (LUPerson)

LUPerson Unsupervised Pre-training for Person Re-identification (LUPerson). The repository is for our CVPR2021 paper Unsupervised Pre-training for Per

143 Dec 24, 2022
Implementing yolov4 target detection and tracking based on nao robot

Implementing yolov4 target detection and tracking based on nao robot

6 Apr 19, 2022
Code of our paper "Contrastive Object-level Pre-training with Spatial Noise Curriculum Learning"

CCOP Code of our paper Contrastive Object-level Pre-training with Spatial Noise Curriculum Learning Requirement Install OpenSelfSup Install Detectron2

Chenhongyi Yang 21 Dec 13, 2022
A Benchmark For Measuring Systematic Generalization of Multi-Hierarchical Reasoning

Orchard Dataset This repository contains the code used for generating the Orchard Dataset, as seen in the Multi-Hierarchical Reasoning in Sequences: S

Bill Pung 1 Jun 05, 2022
Lama-cleaner: Image inpainting tool powered by LaMa

Lama-cleaner: Image inpainting tool powered by LaMa

Qing 5.8k Jan 05, 2023