Skip to main content

Search CooLWJ

选择搜索范围,输入关键词后按回车。

Language

Technology & everyday life

Development, AI tools, local voices and reading.

Latest articles

7–12 / 25 articles

Local voices

What Does “rang” Mean in the Henan Dialect? It Even Appears in the 《左传》 (Zuo Commentary)

Which Chinese character should represent “rang”, the Henan dialect word for scolding children? This word specifically refers to adults criticizing or telling off mischievous children. The author considered two candidate characters: Candidate one:“让”. In Classical Chinese, this character really did mean “to rebuke.” In the Zuo Commentary, “公使让之” means that the duke sent someone to rebuke him. Mandarin has lost this meaning, but Henan dialect has preserved it. Candidate two:“嚷”. In northern dialects, “嚷” itself can mean to scold. The progression from speaking loudly to berating someone loudly also makes sense. There is one small question, though: in the author's hometown, “rang” is pronounced in the first tone. But neither “让” as in “谦让” (to yield modestly) nor “嚷” is pronounced in the first tone locally. So which character it actually is remains unsettled. The two words may also have merged as their pronunciations grew closer. The dialect still lives on in everyday speech, but its original written character was lost long ago.

AI & tools

Testing Mainstream Large AI Models on Calculating Stock Returns

I used common Chinese and international AI models available on the market to help calculate my stock profit. The Chinese models included DeepSeek, Alibaba's Tongyi Qianwen, iFlytek Spark, ByteDance's Doubao, Moonshot AI's Kimi, Tencent Yuanbao, Baidu's ERNIE Bot, Zhipu Qingyan, Baichuan AI, 360 Zhinao, Tiangong AI, and SenseTime's SenseChat; the international models included GPT-4o, Claude-3-Opus, Gemini-2.0-Pro, and Grok. Given the same input, only DeepSeek calculated correctly.

AI & tools

Model Distillation and Model Pruning

Model distillation and pruning are two important techniques for compressing neural network models. They can effectively reduce a model's parameter count and computational complexity, improving its operating efficiency and performance. Although both model distillation and model pruning are neural network compression techniques, their methods and objectives differ slightly. Model distillation focuses more on using a teacher model's predictive behavior to train a student model, while model pruning focuses more on removing unnecessary parameters and connections to compress the model.

Development

Can Being Too Long Be Wrong?—An Unusual Journey to Fix Precision Loss in Backend Long IDs

In the era of separate frontends and backends, our lives are full of countless opportunities and challenges—including those maddening bugs that pop up out of nowhere. Today, we are going to discuss a classic problem that leaves countless developers unsure whether to laugh or cry: backend Long-type IDs that are too long cause precision loss on the frontend. This problem really is a “root of all evil”—why does JavaScript have to safely handle only numbers within Number.MAX_SAFE_INTEGER (that is, 9007199254740991)? If you have ever racked your brains over this problem, don’t worry: you are not alone. Today, we will not only dissect this “problem of the century” with humor, but also walk you through solving it at its root, so your code no longer “loses its precision.”

Development

The Golden Trio of Message Queues:RabbitMQ, RocketMQ, and Kafka in a Comprehensive Showdown—Which Is the Best Choice?

RabbitMQ: Based on AMQP, it supports multiple messaging protocols and offers flexible routing. It suits scenarios requiring complex routing and protocol support, but its relatively low throughput makes it more suitable for small to medium-scale messaging. RocketMQ: An open-source distributed messaging middleware from Alibaba, supporting high throughput and low latency. It is suitable for financial-grade transactional messaging and real-time big data processing. It supports ordered, scheduled, and transactional messages, making it suitable for large-scale distributed systems. Kafka: A high-throughput message queue with persistence, suitable for handling large-scale data streams and real-time data streams, and widely used in big data ecosystems. Its partitioning and replication mechanisms ensure high availability and data consistency, making it particularly suitable for log collection and stream processing.

Choose your theme

Your theme applies across the site.