Algorithms vs SottoPet

Did you pass the interview algorithm question? Learning algorithm questions takes too much time? Not your problem! Before this, no algorithmic learning method could be so simple and efficient. It only takes three steps to become an algorithm master: The first step is to read the question; The second step is to play the animation, refer to the subtitles, and understand the ideas and codes; The third step is to read, analyze and summarize; Why do you make an algorithm animation diagram app? Like many programmers, I have to prepare a lot of time to brush up the algorithm every time I change jobs. Very good, in the process of brushing the questions, you must first read the solution. I also read a lot of problem solutions from big guys on the Internet, and I especially like to read algorithm analysis with moving pictures or schematic diagrams, which is faster than reading text. In the process, several problems were also found, such as: the animation is not interactive and cannot be paused; the data is dead and cannot be changed; it is not convenient to study on the mobile phone, etc. I saw some good animations at the beginning, and I had a strong urge to implement them in the app with native code several times. This time I finally made up my mind and made 70 animations in one go. I understand that the positioning of this app is a tool that can help us quickly understand algorithm ideas and codes, help memory, and improve learning efficiency. But here you can't submit, you can't execute the code, and the code still needs to be typed on the computer to make it feel better. Why is it not arranged in the order of the leetcode number? At present, the algorithm questions are classified into three major blocks: - linked list, string, array, binary tree; - Backtracking, greedy, dynamic programming, divide and conquer; - Sort and classify separately. Under each category, the difficulty is divided into easy, medium and hard. First classify data structures and algorithms separately, and learn according to knowledge points. For example, you can concentrate on learning binary tree structures or dynamic programming algorithms, so that concentrated saturation training can help improve learning efficiency. The data structure comes first, and the algorithm follows, which is also arranged in the order in which we learn theoretical knowledge. In terms of algorithms, four types are currently arranged: backtracking, greedy, dynamic programming, and divide-and-conquer. When I learned data structures and algorithms many years ago, I first came into contact with various sorting algorithms. At that time, I felt that sorting was too difficult, and I was very impressed, so I made a separate sorting classification for beginners. Thinking on how to combine animation and code and other interactive issues - In the process of learning algorithms by myself, I often understand the ideas, but I can't understand the codes, and many codes in the text explanations have no comments. Seeing this is a big headache. Therefore, when thinking about App interaction, the code highlighting function is specially designed, which means that each line of code will be highlighted as the animation progresses. In this way, the animation is executing, the code is highlighted, and it looks cool. - Although the animation is easy to understand, it still needs some short explanations, so the "subtitle" function is designed below the animation. Every time the corresponding step is executed, the corresponding explanation will be displayed. - Animation playback and reset, each algorithm page has a playback function, so how to prepare test cases? To simplify the concept of test cases, a "reset" button is designed, which means that the test cases are random. When you want to change a set of data, just click reset.

SottoPet is a companion app for the coding agents running on your Mac. It does not work on its own: you need SottoPet for Mac (macOS 12 or later, Apple Silicon), signed in to the same account, and you pair the two by scanning a QR code shown on the Mac. ■ Read the history Every Claude Code and Codex conversation on the paired Mac, newest first. Each one carries a short topic and its current state — working, waiting, done. Open one and you can read it from the beginning, images included. ■ Reply Conversations you started from this app accept further messages. When an agent has stopped to ask for permission or to put a question to you, you answer by tapping one of its options. ■ Start work Pick one of the folders your Mac already trusts, choose Claude or Codex and a model, and send the first message. ■ Watch the limits How much of your Claude and Codex rate limits is spent, next to the speech allowance on your SottoPet account. Each figure says when it arrived, and one that has not been refreshed for a while is marked as old. ■ What this app deliberately cannot do It cannot send an arbitrary command or a filesystem path to your Mac — the format the two sides speak has no field of that shape. Work can only start in a folder the Mac already trusts, and trust is granted on the Mac alone. Conversations started from the phone run read-only. Pairing does not take effect until someone at the Mac confirms it. Either side can undo it. Accounts are created at sotto.pet. The interface is available in English, Japanese, Korean, Simplified Chinese and Spanish.