Kezhi's Blog

随着网络基础设施的建设和完善，现在已经有了很多远程游戏的解决方案。其主要分为两种类型：局域网串流（steam串流，PlayStation Portal）和互联网云游戏（云原神，网易云游戏平台）。前者利用局域网下的高带宽和低延时实现同一局域网下的设备之间完成高速串流，解决了设备放在房间，人去另外一个房间（或者躺在床上）玩游戏的需求。后者利用压缩编码和延时控制技术，实现在互联网下利用中低带宽进行远程游戏，解决了在低性能终端设备上随时随地玩高性能需求游戏的问题。

然而，这两种解决方案也有其固有问题。局域网串流要求设备都处于同一局域网下，这对于串流的空间范围就有了限制。而云游戏存在竞争资源，账号隐私等问题。那么能不能有一个中间方案，扩大局域网串流的空间范围，同时避免云游戏存在的平台问题呢。我选择的解决方案是使用RDP在内网进行远程游戏。

With the construction and improvement of network infrastructure, there are now many solutions for remote gaming. They are mainly divided into two types: LAN streaming (Steam Link, PlayStation Portal) and Internet cloud gaming (Genshin Impact Cloud, NetEase Cloud Gaming Platform). The former uses high bandwidth and low latency in the LAN to achieve high-speed streaming between devices in the same LAN, solving the need to play games in a different room (or lying in bed) when the device is in another room. The latter uses compression encoding and latency control technology to enable remote gaming over the Internet using medium to low bandwidth, solving the problem of playing high-performance demanding games anytime, anywhere on low-performance terminal devices.

However, these two solutions also have their inherent problems. LAN streaming requires devices to be on the same LAN, limiting the spatial range of streaming. Cloud gaming faces issues such as competition for resources and account privacy. So, is there a middle ground solution that expands the spatial range of LAN streaming while avoiding the platform issues of cloud gaming? The solution I choose is to use RDP for remote gaming within the LAN.

这是一个简单的故事，只有两位主人公。暂且称这两位主人公为小鱼（男）和小萍（女）吧。

This is a simple story with only two protagonists. Let's call these two protagonists Xiao Yu (male) and Xiao Ping (female) for now.

之前写过一篇Zotero6插件推荐和下载的文章 Zotero插件推荐及下载 | Kezhi's Blog。随着Zotero开发组多年的不懈努力，现在终于迎来了Zotero7。相比于Zotero6，Zotero7基于更新的firefox内核，支持64位，具有更好的性能表现和更美观的界面。但是由于内核的大版本更新，过去Zotero6的插件不再兼容。故再撰写一篇文章用来推荐Zotero7的插件，并提供下载。

I have written an article recommending and downloading Zotero6 plugins before. With the unremitting efforts of the Zotero development team over the years, Zotero7 has finally arrived. Compared to Zotero6, Zotero7 is based on the updated firefox kernel, supports 64-bit, has better performance and a more beautiful interface. However, due to the major version update of the kernel, the plugins for Zotero6 are no longer compatible. Therefore, I am writing another article to recommend Zotero7 plugins and provide downloads.

从 ChatGPT 问世带给人们惊艳的文本生成能力表现之后，各家大模型如雨后春笋般逐渐冒出。开发者们也基于大语言模型的能力，开发出来很多让大模型来赋能生活和工作的功能。然而，大语言模型究其本质，也只是一个文本概率预测模型，而非真正的通用人工智能。尽管它在很多方便表现出很强的智能能力，在一些任务上极大提升了人们的效率。但一味地信任它给出的结果也可能会工作和生活带来副作用。目前也有很多对大语言模型回答幻觉和事实性的研究。本文不从研究的角度出发，仅以个人日常生活中对大模型使用的取舍出发，给出一点使用当前水平大语言模型的一些思考。

After the amazing text generation capabilities brought by the emergence of ChatGPT, various large models have gradually emerged like mushrooms after the rain. Developers have also developed many functions based on the capabilities of large language models to empower life and work. However, at its core, a large language model is just a text probability prediction model, not a true general artificial intelligence. Although it demonstrates strong intelligence in many aspects and greatly improves efficiency in some tasks, blindly trusting its results may also bring side effects to work and life. Currently, there is also a lot of research on the illusion and factual nature of large language model responses. This article does not start from a research perspective, but only discusses the considerations of using large models in daily life.

原本是不想写 latex 本地环境配置的文章的，因为网上已经有很多相关的博客和文章。但是师妹在配置 vscode 时复制了网上错误的配置导致 bib 半天编译不过，并难以定位问题。因此，写下本篇配置说明来维护一个正确的配置，避免未来再出现类似的问题。

Originally, I didn't want to write an article about configuring the local LaTeX environment because there are already many related blogs and articles online. However, my junior sister made a mistake by copying the wrong configuration from the internet when setting up vscode, which caused the bibliography to fail to compile for half a day and was difficult to locate the problem. Therefore, I wrote this configuration guide to maintain a correct configuration and avoid similar issues in the future.