Search all stories搜索全部新闻/
Daily每日

Friday, October 9, 20262026年10月9日 · 星期五

Daily Best 50 + Trending 61每日最佳 50 + 热门 61 · Updated Oct 9, 2026 · 6:23 PM PDT更新于 2026年10月9日 18:23(PDT) · 111 stories条
Click a tag to filter点标签筛选
carette.xyz
01

Lobbying is corruption游说就是腐败carette.xyz

Lobbying concentrates influence among moneyed interests, which can distort policy and undermine fair representation.游说把影响力集中在有钱势力手中,侵蚀平等的政治代表性。

Lobbying concentrates influence among moneyed interests, which can distort policy and undermine fair representation. The piece argues that this dynamic amounts to corruption and calls for stronger transparency and anti-corruption reforms.游说把影响力集中在有钱势力手中,侵蚀平等的政治代表性。文章指出这一现象促成腐败并扭曲政策,应加强透明度与反腐改革。

Politics16.2414204
community.bitwarden.com
02

Bitwarden Dual License ModelBitwarden 双许可证模式community.bitwarden.com

Bitwarden introduces a dual-licensing model to balance openness with commercial viability, offering an AGPL-licensed community edition alongside a paid license for enterprise deployments.Bitwarden 推出双许可证模式,在社区版采用 AGPL 的同时提供企业级付费许可,以兼顾开源透明度与商业可行性。

Bitwarden introduces a dual-licensing model to balance openness with commercial viability, offering an AGPL-licensed community edition alongside a paid license for enterprise deployments. This approach aims to keep core software open while providing a compliant path for large organizations to deploy Bitwarden at scale.Bitwarden 推出双许可证模式,在社区版采用 AGPL 的同时提供企业级付费许可,以兼顾开源透明度与商业可行性。此举旨在为大型组织的部署提供合规路径,同时确保核心开源工作继续受益于社区贡献。

Open SourceBusiness11.3177125
cphpost.dk
03

`123456' password used in Danish CPR data breach丹麦 CPR 数据泄露中使用了 123456 密码cphpost.dk

Attackers used the simple password 123456 to access Danish CPR records, exposing sensitive personal data.攻击者使用简单密码 123456 访问了丹麦 CPR 记录,暴露了敏感个人信息。

Attackers used the simple password 123456 to access Danish CPR records, exposing sensitive personal data. This incident shows how weak credentials can enable large scale government data breaches and underscores the need for stronger authentication and password hygiene across public-facing systems.攻击者使用简单密码 123456 访问了丹麦 CPR 记录,暴露了敏感个人信息。该事件凸显了弱密码如何使政府数据大规模泄露成为可能,也强调在面向公众的系统中需要更强的身份认证和密码管理。

SecurityPrivacyPolitics5.7313172
github.com
04

Talorys – A self-hosted personal AI agent on Cloudflare's free tierTalorys:在 Cloudflare 免费层自托管的个人 AI 代理github.com

Talorys is a self-hosted personal AI assistant that runs on Cloudflare's free tier, enabling edge-based AI without a traditional cloud subscription.Talorys 是一个在 Cloudflare 免费层运行的自托管个人 AI 助手,支持在边缘执行 AI 而无需传统云订阅。

Talorys is a self-hosted personal AI assistant that runs on Cloudflare's free tier, enabling edge-based AI without a traditional cloud subscription. It demonstrates how individuals can deploy private AI tools at the edge, balancing privacy, latency, and cost while avoiding vendor lock-in.Talorys 是一个在 Cloudflare 免费层运行的自托管个人 AI 助手,支持在边缘执行 AI 而无需传统云订阅。它展示了个人在边缘部署私有 AI 工具的可能性,平衡隐私、低延迟和成本,同时降低对单一厂商的依赖。

AIPrivacyCode3.214678
05

Apple/macOS silently removed from official Unix registry苹果/macOS 已从官方 Unix 注册中移除opengroup.org

Apple's macOS has been quietly removed from the Open Group's official Unix branding registry, indicating a branding shift rather than a software change.苹果的 macOS 已从开放组织的官方 Unix 品牌注册名单中移除,体现的是品牌层面的变化而非系统本身的改动。

Apple's macOS has been quietly removed from the Open Group's official Unix branding registry, indicating a branding shift rather than a software change. The move matters for developers and vendors who rely on Unix branding for portability and interoperability.苹果的 macOS 已从开放组织的官方 Unix 品牌注册名单中移除,体现的是品牌层面的变化而非系统本身的改动。此举可能影响那些依赖 Unix 品牌认知来展示兼容性和可移植性的开发者与厂商。

Politics3.2143144
06

Triple-A MinesweeperAAA级扫雷minesweeper.mikelacher.com

Triple-A Minesweeper showcases a highly polished, modern take on a classic puzzle, blending AAA production values with a timeless game.AAA级扫雷将经典益智游戏推向高水准的现代化呈现,拥有精致的美术、流畅的交互和性能优化。

Triple-A Minesweeper showcases a highly polished, modern take on a classic puzzle, blending AAA production values with a timeless game. It highlights how contemporary tooling and UX can reinvent retro titles for new audiences.AAA级扫雷将经典益智游戏推向高水准的现代化呈现,拥有精致的美术、流畅的交互和性能优化。该项目展示了当代工具链与用户体验如何为复古游戏注入新活力。

Gaming3.21225243
07

REA Reverse – Engineer AnythingREA Reverse:逆向工程万物rea.tools

REA Reverse provides a hands-on platform for reverse engineering across domains, offering challenges and tutorials to dissect how software and hardware work.REA Reverse 提供一个聚焦逆向工程的实践平台,涵盖软件与硬件领域的挑战与教程,帮助用户拆解系统工作原理。

REA Reverse provides a hands-on platform for reverse engineering across domains, offering challenges and tutorials to dissect how software and hardware work. It aims to democratize RE skills by curating practical content and community-driven exercises. This matters for defenders, researchers, and hardware hackers looking to deepen code comprehension and vulnerability analysis.REA Reverse 提供一个聚焦逆向工程的实践平台,涵盖软件与硬件领域的挑战与教程,帮助用户拆解系统工作原理。通过实操内容和社区驱动的练习,推动逆向工程技能的普及。对安全研究、漏洞分析和硬件黑客都具有现实意义。

SecurityEducationHardware3.1588256
08

Cloudflare acquires DenoCloudflare 收购 Deno,整合边缘计算生态deno.com

Cloudflare is acquiring Deno, marrying Deno’s secure, TypeScript-friendly runtime with Cloudflare’s edge network and Workers platform.Cloudflare 将收购 Deno,将 Deno 的安全、TypeScript 友好运行时与 Cloudflare 的边缘网络和 Workers 平台整合。

Cloudflare is acquiring Deno, marrying Deno’s secure, TypeScript-friendly runtime with Cloudflare’s edge network and Workers platform. The deal could accelerate edge-based development and broaden Cloudflare’s developer toolkit, though it raises questions about governance and open-source leadership in the workspace.Cloudflare 将收购 Deno,将 Deno 的安全、TypeScript 友好运行时与 Cloudflare 的边缘网络和 Workers 平台整合。此举可能加速基于边缘的开发并扩展 Cloudflare 的开发者工具包,同时也引发关于治理与开源领导力的讨论。

Open SourceBusinessProgramming2.91311676
gizmodo.com
09

YouTuber Says Cops Visited Him After He Built a Flock-Style Camera to Track Cops网红称警方在他自制的Flock风格摄像头追踪警察后上门gizmodo.com

A YouTuber says police visited him after he built a Flock-style camera to track cops.网红称警方在他自制的Flock风格摄像头用于追踪警察后上门。

A YouTuber says police visited him after he built a Flock-style camera to track cops. The incident highlights tensions between civilian surveillance tools and law enforcement. It raises questions about privacy, safety, and accountability when DIY cameras monitor public spaces.网红称警方在他自制的Flock风格摄像头用于追踪警察后上门。此事暴露了民用监控工具与执法之间的紧张关系,也引发对隐私和安全的讨论。它强调在 DIY 摄像头监控公共场域时,谁应对行为负责以及相关法律边界的问题。

Privacy2.4634342
10

Telegram Desktop vulnerability allowed any user's file to be stolenTelegram 桌面端漏洞致使任意用户文件可能被窃取beaksec.github.io

Telegram Desktop suffered a vulnerability that let an attacker access and steal a victim's files via the desktop client, potentially enabling a one-click account takeover.Telegram 桌面端存在漏洞,攻击者可通过桌面客户端访问并窃取受害者的文件,甚至实现一键式账户接管。

Telegram Desktop suffered a vulnerability that let an attacker access and steal a victim's files via the desktop client, potentially enabling a one-click account takeover. The bug highlights how trusted desktop apps can become backdoors for data exfiltration, so users should update to patched versions and developers should harden local storage and permission handling.Telegram 桌面端存在漏洞,攻击者可通过桌面客户端访问并窃取受害者的文件,甚至实现一键式账户接管。这一漏洞凸显受信任的桌面应用可能成为数据外泄的后门,用户应尽快升级打补丁,开发者需加强本地存储与权限管理。

SecurityPrivacy2.2322164
iminafleeting.com
11

Sorry, I'm in a meeting我在开会iminafleeting.com

A succinct look at how being marked as 'in a meeting' has become a common status signal in modern work culture.一句话概括地揭示在现代职场中‘我在开会’已成为常见的状态信号。

A succinct look at how being marked as 'in a meeting' has become a common status signal in modern work culture.一句话概括地揭示在现代职场中‘我在开会’已成为常见的状态信号。

Culture1.7974268
wiki.cateat.fish
12

Computers Cannot Make Decisions电脑无法做出决定wiki.cateat.fish

Computers cannot make decisions on their own; genuine decision-making requires human judgment, accountability, and context.计算机无法独立做出决策;真正的决策需要人类的判断、责任和情境。

Computers cannot make decisions on their own; genuine decision-making requires human judgment, accountability, and context. The piece argues for designing AI as decision-support tools rather than autonomous deciders, emphasizing governance and safety in automated systems.计算机无法独立做出决策;真正的决策需要人类的判断、责任和情境。文章主张将人工智能设计为决策辅助工具,并强调在自动化系统中的治理与安全性。

AI1.6175148
13

Our $445M Series D我们完成4.45亿美元的D轮融资oxide.computer

Oxide Computer announced a $445M Series D round, fueling its cloud-native hardware platform designed for disaggregated data centers.Oxide Computer 宣布完成4.45亿美元的 D 轮融资,为其面向云原生硬件平台的去中心化数据中心设计提供资金。

Oxide Computer announced a $445M Series D round, fueling its cloud-native hardware platform designed for disaggregated data centers. The tranche signals sustained investor interest in software-defined infrastructure and enterprise-grade edge/cloud deployments, while boosting R&D and go-to-market efforts.Oxide Computer 宣布完成4.45亿美元的 D 轮融资,为其面向云原生硬件平台的去中心化数据中心设计提供资金。此轮融资反映投资者对软件定义基础设施与企业级边缘/云部署的持续兴趣,同时促进研发与市场拓展。

StartupsBusinessProgrammingHardwareDevOps1.5678309
geohot.github.io
14

Lobbying游说geohot.github.io

Lobbying often shapes policy through concentrated influence, risking policy capture by special interests.游说往往通过集中的影响力来塑造政策,存在被特定利益集团控制政策的风险。

Lobbying often shapes policy through concentrated influence, risking policy capture by special interests.游说往往通过集中的影响力来塑造政策,存在被特定利益集团控制政策的风险。文章讨论其伦理问题以及遏制不当影响的潜在改革方向,聚焦科技政策领域。

Politics1.417667
reuters.com
15

US imposes sanctions on ICC hours after former judge wins Nobel Peace Prize美国在前任法官获诺贝尔奖后数小时对 ICC 实施制裁reuters.com

Hours after a former ICC judge was awarded the Nobel Peace Prize, the US announced sanctions on the International Criminal Court.在前 ICC 法官获得诺贝尔和平奖后数小时,美国宣布对国际刑事法院实施制裁。

Hours after a former ICC judge was awarded the Nobel Peace Prize, the US announced sanctions on the International Criminal Court. The move underscores ongoing tensions between Washington and the ICC over investigations into powerful interests, potentially chilling cooperation and complicating future accountability efforts.在前 ICC 法官获得诺贝尔和平奖后数小时,美国宣布对国际刑事法院实施制裁。此举凸显华盛顿与 ICC 之间持续的紧张关系,可能削弱合作并使未来追责工作变得更为复杂。

PoliticsLegal1.2501503
16

Typesafe AI raises $870M at $7.5BTypeSafe AI 获8.7亿美元融资,估值75亿美元typesafe.ai

TypeSafe AI raised $870 million at a $7.5 billion valuation, underscoring investor enthusiasm for safety-focused AI tooling.TypeSafe AI 完成 8.7 亿美元融资,估值达 75 亿美元,凸显市场对具备安全性与可控性的 AI 工具的强烈兴趣。

TypeSafe AI raised $870 million at a $7.5 billion valuation, underscoring investor enthusiasm for safety-focused AI tooling. The funding will likely accelerate product development, governance features, and enterprise adoption of safer AI systems.TypeSafe AI 完成 8.7 亿美元融资,估值达 75 亿美元,凸显市场对具备安全性与可控性的 AI 工具的强烈兴趣。资金将推动产品迭代、治理能力和企业级应用的落地。

AIStartups1.2416334
17

Show HN: Carrier-Explode: iPhone, Pixel and Galaxy carrier settings decodedShow HN:Carrier-Explode:解码 iPhone、Pixel、Galaxy 的运营商设置carrierexplode.com

A project decodes the carrier settings used by major mobile platforms, showing how iPhone, Pixel, and Galaxy handle provisioning differently.一个项目解码 iPhone、Pixel、Galaxy 在运营商设置上的差异,揭示三大平台在 provisioning 方面的不同处理方式。

A project decodes the carrier settings used by major mobile platforms, showing how iPhone, Pixel, and Galaxy handle provisioning differently. The work helps developers and power users understand network behavior, update management, and potential privacy implications of carrier configurations.一个项目解码 iPhone、Pixel、Galaxy 在运营商设置上的差异,揭示三大平台在 provisioning 方面的不同处理方式。此工作有助开发者与高阶用户理解网络行为、更新管理,以及运营商配置带来的隐私影响。

Show HNProgrammingPrivacy1.238245
18

No Man Is an Island没有人是孤岛borretti.me

The piece argues that individual limits are overcome through collaboration and networks, emphasizing interdependence in modern work and life.本文主张通过协作与网络来克服个人能力的局限,强调现代工作与生活中的相互依赖。

The piece argues that individual limits are overcome through collaboration and networks, emphasizing interdependence in modern work and life.本文主张通过协作与网络来克服个人能力的局限,强调现代工作与生活中的相互依赖。

Culture1.1304195
19

Nobel Peace Prize for 2026 to Navanethem Pillay2026 年诺贝尔和平奖授予 Navanethem 'NAVI' Pillaynobelprize.org

Navanethem Pillay has been named the recipient of the 2026 Nobel Peace Prize, recognizing her decades of advocacy for human rights and international justice.纳瓦涅坦·皮莱被授予2026年诺贝尔和平奖,表彰她多年来在维护人权和推动国际司法方面的努力。

Navanethem Pillay has been named the recipient of the 2026 Nobel Peace Prize, recognizing her decades of advocacy for human rights and international justice. The award underscores the ongoing emphasis on human rights as a foundation of global governance and diplomacy.纳瓦涅坦·皮莱被授予2026年诺贝尔和平奖,表彰她多年来在维护人权和推动国际司法方面的努力。此项荣誉凸显人权在全球治理与外交中的持续重要性。

Politics0.9475244
20

Whistle: Speech to Text in 16.9 MBWhistle:16.9 MB 的语音转文本cactuscompute.com

Whistle introduces a speech-to-text model that runs at just 16.9 MB, delivering surprisingly good transcription quality for lightweight use.Whistle 推出仅16.9 MB的语音转文字模型,在重量级外的轻量化场景也能实现相当不错的转写效果。

Whistle introduces a speech-to-text model that runs at just 16.9 MB, delivering surprisingly good transcription quality for lightweight use. It exemplifies the trend toward tiny, efficient AI models suitable for edge devices and offline scenarios.Whistle 推出仅16.9 MB的语音转文字模型,在重量级外的轻量化场景也能实现相当不错的转写效果。体现了小型高效 AI 模型在边缘设备与离线场景中的发展趋势。

AI0.8941181
21

Anthropic AI model submits false tip on unsolved Philly murder, police sayAnthropic AI模型向费城未解凶杀案提交虚假线索nbcphiladelphia.com

A generative AI model submitted a false tip to Philadelphia police about an unsolved murder, highlighting the risk of deploying AI in real-time law-enforcement contexts without proper safeguards.生成式 AI 模型向费城警方提供了一个虚假线索,暴露了在执法场景中直接使用 AI 输出的风险。

A generative AI model submitted a false tip to Philadelphia police about an unsolved murder, highlighting the risk of deploying AI in real-time law-enforcement contexts without proper safeguards. The incident emphasizes the need for human review, audit trails, and robust guardrails in safety-critical AI deployments.生成式 AI 模型向费城警方提供了一个虚假线索,暴露了在执法场景中直接使用 AI 输出的风险。事件强调在安全关键场景下需要人工审查、可追溯性和强健的安全边界。

AISecurity0.8195139
github.com
22

Show HN: Let your AI agents paint big arrows, boxes and text on your screen让 AI 代理在屏幕上绘制大箭头、方框和文本github.com

A lightweight demo shows AI agents annotating the UI with large arrows, boxes, and labels to explain their plans in real time.一个轻量级演示展示 AI 代理在屏幕上绘制大箭头、方框和文本标签,以实时解释其计划。

A lightweight demo shows AI agents annotating the UI with large arrows, boxes, and labels to explain their plans in real time. Such overlays boost interpretability and debugging, helping humans understand where the agent is headed and why. The approach is easy to prototype and could be used in demos, tutorials, or human-in-the-loop workflows.一个轻量级演示展示 AI 代理在屏幕上绘制大箭头、方框和文本标签,以实时解释其计划。此类覆盖能提升可解释性与调试效果,帮助人类理解代理的意图与下一步动作。该方法易于原型化,适用于演示、教程和人机协作场景。

Show HNAICode0.8404184
python.org
23

Python 3.15Python 3.15 版本发布python.org

Python 3.15 launches with performance improvements, better error messages, and standard library enhancements.Python 3.15 正式发布,带来性能提升、错误信息改进与标准库增强。

Python 3.15 launches with performance improvements, better error messages, and standard library enhancements. The release continues Python’s modernization, while marking deprecated features that developers should begin migrating away from.Python 3.15 正式发布,带来性能提升、错误信息改进与标准库增强。该版本延续了 Python 的现代化进程,同时标注了一些将被弃用的特性,开发者应尽早迁移。

Programming0.7299105
24

OpenAI fires three safety researchers for "mishandling research information"OpenAI 因“不当处理研究信息”解雇三名安全研究员techcrunch.com

OpenAI reportedly fired three safety researchers over alleged mishandling of research information, sparking disputes about the facts and the implications for safety work at the company.据报道,OpenAI 因疑似不当处理研究信息解雇了三名安全研究人员,引发关于事实与对公司安全工作的影响的争论。

OpenAI reportedly fired three safety researchers over alleged mishandling of research information, sparking disputes about the facts and the implications for safety work at the company. Critics warn the moves could chill internal safety research and discourse.据报道,OpenAI 因疑似不当处理研究信息解雇了三名安全研究人员,引发关于事实与对公司安全工作的影响的争论。批评者担心此举可能抑制内部安全研究与讨论。

BusinessAI0.6349217
25

Microsoft-Decision-1, our model for fast decision-making微软推出 Decision-1:用于快速决策的模型commandline.microsoft.com

Microsoft introduces Decision-1, a model designed to speed up decision-making in workflows, potentially serving as a backbone for AI-powered automation and governance tooling.微软推出 Decision-1,旨在提升工作流中的决策速度,可能成为 AI 驱动自动化与治理工具的核心。

Microsoft introduces Decision-1, a model designed to speed up decision-making in workflows, potentially serving as a backbone for AI-powered automation and governance tooling. The development signals a push toward faster, more reliable AI-assisted decisions.微软推出 Decision-1,旨在提升工作流中的决策速度,可能成为 AI 驱动自动化与治理工具的核心。此举体现了向更快、更可靠的 AI 辅助决策迈进的趋势。

AI0.619565
animationobsessive.substack.com
26

'Wallace and Gromit,' 90% Alone《华莱士与格罗米特》:90% 的工作由独自完成animationobsessive.substack.com

The piece argues that modern stop-motion animation projects can be largely solo endeavors, leveraging automation and modular workflows to produce a feature-like result with limited collaboration.该文探讨在现代定格动画创作中,创作过程往往由单人主导,通过自动化与模块化工作流程实现接近大片水准的产出,强调个人在创作中的核心作用。

The piece argues that modern stop-motion animation projects can be largely solo endeavors, leveraging automation and modular workflows to produce a feature-like result with limited collaboration.该文探讨在现代定格动画创作中,创作过程往往由单人主导,通过自动化与模块化工作流程实现接近大片水准的产出,强调个人在创作中的核心作用。

EntertainmentDesign0.626141
27

Yandex Takes a Second Data Center Hit in 48 Hours48小时内雅andex 再次遭遇数据中心攻击united24media.com

Yandex reports a second data center outage within 48 hours, impacting its services and user traffic.Yandex 在48小时内再次遭遇数据中心中断,影响服务与用户流量。

Yandex reports a second data center outage within 48 hours, impacting its services and user traffic. The incident highlights resilience challenges for large-scale platforms in volatile environments and underscores the importance of DR planning, geo-redundancy, and rapid recovery.Yandex 在48小时内再次遭遇数据中心中断,影响服务与用户流量。此事凸显在高风险环境下大规模平台的韧性挑战,强调灾备规划、地理冗余和快速恢复的重要性。

BusinessDevOps0.5227319
28

Scam American companies are using to manipulate ingredient lists美国公司用来操纵配料表的骗局twitter.com

Discussion of deceptive labeling practices used by some American companies to manipulate ingredient lists, potentially misleading consumers about product contents.讨论部分美国公司操控配料表以误导消费者的欺骗性做法,涉及成分披露与健康声称的混淆。

Discussion of deceptive labeling practices used by some American companies to manipulate ingredient lists, potentially misleading consumers about product contents. The note underscores the need for clearer labeling, regulatory enforcement, and consumer vigilance.讨论部分美国公司操控配料表以误导消费者的欺骗性做法,涉及成分披露与健康声称的混淆。强调需要更清晰的标签、监管执行和消费者警觉。

BusinessLegalPolitics0.5177204
29

Germany transforms former coal mines into Europe's largest lake landscape德国将废弃的煤矿改造成欧洲最大的湖泊景观euronews.com

Germany is turning old open-pit coal mines into Europe’s largest lake landscape, a massive ecological and tourism project.德国正在把废弃的露天煤矿改造成欧洲最大的湖泊景观,这是一次规模巨大的生态修复与旅游开发计划,体现环境治理与区域经济多元化的长期努力。

Germany is turning old open-pit coal mines into Europe’s largest lake landscape, a massive ecological and tourism project. The transformation demonstrates long-term environmental reclamation, water management, and regional economic diversification.德国正在把废弃的露天煤矿改造成欧洲最大的湖泊景观,这是一次规模巨大的生态修复与旅游开发计划,体现环境治理与区域经济多元化的长期努力。

ScienceBusiness0.5206115
30

Theranos.worldTheranos.worldtheranos.world

A project revisits Theranos and its legacy, highlighting the gap between hype and reality in biotech storytelling.此项目回顾 Theranos 及其影响,揭示生物科技领域里炒作与现实的落差。

A project revisits Theranos and its legacy, highlighting the gap between hype and reality in biotech storytelling. It offers cautionary insights for founders and investors about due diligence and governance.此项目回顾 Theranos 及其影响,揭示生物科技领域里炒作与现实的落差。为创始人和投资者提供关于尽职调查与治理方面的警示与启示。

BusinessLegal0.5559204
terrytao.wordpress.com
31

What mathematicians should know about the Lean Theorem Prover: reliability & AILean 定理证明器:可靠性与 AI 的关系terrytao.wordpress.com

Lean's theorem prover raises questions about reliability and AI-assisted proof, urging mathematicians to understand its limits and strengths.Lean 定理证明器引发了关于可靠性与 AI 辅助证明的问题,敦促数学家理解其局限与优势。

Lean's theorem prover raises questions about reliability and AI-assisted proof, urging mathematicians to understand its limits and strengths.Lean 定理证明器引发了关于可靠性与 AI 辅助证明的问题,敦促数学家理解其局限与优势。文中讨论了 AI 在帮助发现证明时如何保持严格性。

ProgrammingAI0.517045
earthquake.usgs.gov
32

M7.6 Earthquake in Panama巴拿马发生7.6级地震earthquake.usgs.gov

A powerful M7.6 earthquake struck the Panama region, triggering widespread shaking and emergency alerts.巴拿马地区发生7.6级强震,震感波及周边地区并引发应急警报。

A powerful M7.6 earthquake struck the Panama region, triggering widespread shaking and emergency alerts. Seismologists are coordinating rapid damage assessment, highlighting the importance of resilient infrastructure and early warning in densely populated zones.巴拿马地区发生7.6级强震,震感波及周边地区并引发应急警报。地震学家正协同开展快速的损害评估,凸显在人口密集区域提升基础设施韧性与早期预警的重要性。

ScienceDevOps0.516038
33

Show HN: The rarest tech books and docs you've probably never readShow HN:你可能从未读过的最罕见科技书籍与文档readrare.com

A Show HN featuring a catalog of rare tech books and docs many readers won't have encountered.Show HN 收录了一批极罕见的科技书籍与文献,许多读者可能从未接触。

A Show HN featuring a catalog of rare tech books and docs many readers won't have encountered. The collection offers historical perspective and technical depth for researchers and curious engineers.Show HN 收录了一批极罕见的科技书籍与文献,许多读者可能从未接触。该收藏为研究人员与好奇的工程师提供了宝贵的历史视角和技术深度。

Show HNEducation0.515459
34

Keyboard differences between Windows and MacsWindows 与 Mac 键盘差异unsung.aresluna.org

Explores the practical keyboard layout and shortcut differences between Windows and macOS, including Ctrl vs Cmd, Alt vs Option, and function keys.对比 Windows 与 macOS 的键盘布局与快捷键差异,重点包括 Ctrl 与 Cmd、Alt 与 Option,以及功能键的不同。

Explores the practical keyboard layout and shortcut differences between Windows and macOS, including Ctrl vs Cmd, Alt vs Option, and function keys. The piece explains how these variances affect cross-platform workflows, tooling, and onboarding for developers and power users.对比 Windows 与 macOS 的键盘布局与快捷键差异,重点包括 Ctrl 与 Cmd、Alt 与 Option,以及功能键的不同。分析这对跨平台工作流、开发工具和新手上手的影响,以及如何在设计中考虑这些差异。

Programming0.4336281
35

Imposing Sanctions on the International Criminal Court对国际刑事法院实施制裁state.gov

The United States imposes sanctions on the International Criminal Court, signaling a hard-line stance on ICC investigations.美国对国际刑事法院实施制裁,显示对 ICC 调查行动的强硬立场。

The United States imposes sanctions on the International Criminal Court, signaling a hard-line stance on ICC investigations. The move could limit cross-border cooperation and intensify geopolitical friction around accountability for war crimes, highlighting how politics shapes international justice mechanisms.美国对国际刑事法院实施制裁,显示对 ICC 调查行动的强硬立场。此举可能限制跨境合作并加剧在追责战争罪方面的地缘政治摩擦,凸显政治因素对国际司法机制的影响。

PoliticsLegal0.4178228
36

ADHD as a circadian rhythm disorder: evidence and implications for chronotherapy (2025)ADHD 被视为昼夜节律障碍:证据与时序治疗的启示(2025)frontiersin.org

New evidence positions ADHD as a circadian rhythm disorder, with implications for chronotherapy strategies.新证据将 ADHD 定位为昼夜节律障碍,可能改变时序治疗策略。

New evidence positions ADHD as a circadian rhythm disorder, with implications for chronotherapy strategies. Timing medications and light exposure could help align symptoms and improve outcomes, signaling a shift toward circadian-aware care.新证据将 ADHD 定位为昼夜节律障碍,可能改变时序治疗策略。通过优化药物时机和光照暴露,或能提升治疗效果,推动以昼夜节律为核心的治疗模式。

HealthScience0.4390261
blog.glyph.im
37

Programming Isn't Special编程并不特别blog.glyph.im

The author argues programming is not a magical discipline; success stems from disciplined practice, clear thinking, and code craftsmanship, not novelty.作者主张编程并非神奇艺术,成功来自于系统化训练、清晰思考与代码工匠精神,而非追逐新奇。

The author argues programming is not a magical discipline; success stems from disciplined practice, clear thinking, and code craftsmanship, not novelty. It emphasizes common-sense engineering and the value of learning the boring bits.作者主张编程并非神奇艺术,成功来自于系统化训练、清晰思考与代码工匠精神,而非追逐新奇。强调踏实的工程实践与打牢基础的重要性。

Programming0.4243248
washingtonpost.com
38

Iranian campaign planted fake articles in real U.S. publications using ChatGPT伊朗行动利用 ChatGPT 在美国主流出版物投放伪文章washingtonpost.com

Iranian campaigns used ChatGPT to generate fake articles and seeded them into real U.S.伊朗行动利用 ChatGPT 生成伪文章并投放至美国主流出版物,试图影响公众舆论。

Iranian campaigns used ChatGPT to generate fake articles and seeded them into real U.S. publications, attempting to sway public opinion. The incident demonstrates AI-generated misinformation’s risk to media integrity and the need for stronger verification, attribution, and detection in publishing platforms.伊朗行动利用 ChatGPT 生成伪文章并投放至美国主流出版物,试图影响公众舆论。此事凸显 AI 生成的错误信息对媒体公信力的风险,促使出版平台加强核验、溯源与检测能力。

PoliticsSecurityAI0.4187171
39

OpenAI annualised revenues $20B less than previously signalledOpenAI 年化收入比此前预计低出200亿美元cnbc.com

OpenAI reported annualized revenue that came in about $20B below what investors had signaled, signaling a sharper slowdown or slower-than-expected monetization of its model.OpenAI 的年化收入约比此前透露的数字低出约200亿美元。

OpenAI reported annualized revenue that came in about $20B below what investors had signaled, signaling a sharper slowdown or slower-than-expected monetization of its model. The shortfall highlights how even top-tier AI platforms are still iterating on pricing, enterprise adoption, and cost-to-revenue dynamics as competition and infrastructure costs mount.OpenAI 的年化收入约比此前透露的数字低出约200亿美元。该修正透露出其商业化路径的不确定性,以及企业客户的实际 uptake 以及竞争与基础设施成本对盈利的影响。

BusinessFinanceAIDevOps0.4427300
40

Can you use autoregressive diffusion to generate market data?能否用自回归扩散生成市场数据?blog.janestreet.com

Explores using autoregressive diffusion models to generate synthetic market data for backtesting and research.评估使用自回归扩散模型生成市场数据以用于回测和研究的可行性,讨论在保留真实统计性质与避免数据泄露之间的权衡,以及需面对的市场动态匹配挑战。

Explores using autoregressive diffusion models to generate synthetic market data for backtesting and research. The piece weighs practicality, such as preserving realistic statistics and avoiding data leakage, against challenges like matching complex market dynamics.评估使用自回归扩散模型生成市场数据以用于回测和研究的可行性,讨论在保留真实统计性质与避免数据泄露之间的权衡,以及需面对的市场动态匹配挑战。

AIDataFinance0.415343
41

Pointing AI at archives found a forgotten meteorite, lost rhinos, and more将 AI 应用于档案馆的检索:发现被遗忘的陨石、失踪的犀牛等jessewaites.com

AI applied to centuries of archives surfaces discoveries like a forgotten meteorite and extinct rhinos, illustrating AI's potential to aid historians and conservation.将 AI 应用于几十到上百年的档案检索,揭示了被遗忘的陨石和消失的犀牛等线索,显示 AI 在史料研究和保护工作中的潜力。

AI applied to centuries of archives surfaces discoveries like a forgotten meteorite and extinct rhinos, illustrating AI's potential to aid historians and conservation. It demonstrates how search, OCR, and reasoning can unlock overlooked material.将 AI 应用于几十到上百年的档案检索,揭示了被遗忘的陨石和消失的犀牛等线索,显示 AI 在史料研究和保护工作中的潜力。该案例展示了文本检索、光学字符识别与推理等技术如何挖掘被长期忽视的材料。

AIScience0.317588
quake-srp.pages.dev
42

Show HN: Quake ported to safe Rust, playable in browserShow HN:用安全 Rust 重写的 Quake 可在浏览器中游玩quake-srp.pages.dev

A browser-playable Quake port written in safe Rust demonstrates that high-performance, low-level games can run in the browser without unsafe code.用安全 Rust 编写、可在浏览器中游玩的 Quake 端口,展示无需不安全代码也能在浏览器实现高性能低级别游戏。

A browser-playable Quake port written in safe Rust demonstrates that high-performance, low-level games can run in the browser without unsafe code. The project highlights Rust's WASM toolchain maturity and the potential for safer, accessible retro-game demonstrations in web apps.用安全 Rust 编写、可在浏览器中游玩的 Quake 端口,展示无需不安全代码也能在浏览器实现高性能低级别游戏。项目体现了 Rust 的 WebAssembly 工具链成熟度,以及在网页应用中展示复古游戏的安全、可访问性潜力。

Show HNProgrammingWeb0.3231176
github.com
43

MXC - a sandboxed code execution systemMXC:一个沙箱化的代码执行系统github.com

MXC provides a sandboxed, policy-controlled code execution environment designed to run arbitrary code safely in the cloud.MXC 提供一个沙箱化、受策略控制的代码执行环境,旨在安全地在云端运行不受信任的代码。

MXC provides a sandboxed, policy-controlled code execution environment designed to run arbitrary code safely in the cloud. It supports isolating processes, controlling resources, and auditing behavior, enabling safer experimentation and evaluation of untrusted code.MXC 提供一个沙箱化、受策略控制的代码执行环境,旨在安全地在云端运行不受信任的代码。它通过进程隔离、资源限制和行为审计等机制,便于在云端进行安全的代码评估与试验。

ProgrammingSecurityCode0.319283
44

Bevy 0.20Bevy 0.20bevy.org

Bevy 0.20 brings updates to the Rust game engine, including improvements to the ECS, rendering, and tooling ecosystem.Bevy 0.20 版本带来对 Rust 游戏引擎的更新,提升 ECS、渲染管线和开发工具链等方面的性能与稳定性。

Bevy 0.20 brings updates to the Rust game engine, including improvements to the ECS, rendering, and tooling ecosystem. The release typically emphasizes performance, stability, and expanding support for both 2D and 3D games.Bevy 0.20 版本带来对 Rust 游戏引擎的更新,提升 ECS、渲染管线和开发工具链等方面的性能与稳定性。新版本持续扩展生态,帮助开发者构建 2D/3D 游戏。

ProgrammingGaming0.325068
45

OpenAI, the Partition Principle, and MathematicsOpenAI、分区原理与数学karagila.org

This post discusses how partition principles—methods for breaking problems into independent parts—relate to OpenAI's approaches and the mathematics underpinning modern AI systems.本文探讨如何将分区原理应用于 AI 的数学基础与推理方法,及其与 OpenAI 的研究取向之间的关联。

This post discusses how partition principles—methods for breaking problems into independent parts—relate to OpenAI's approaches and the mathematics underpinning modern AI systems. It argues that problem decomposition can inform scalable, provable AI design. The piece connects theoretical foundations with practical AI architectures.本文探讨如何将分区原理应用于 AI 的数学基础与推理方法,及其与 OpenAI 的研究取向之间的关联。通过将复杂问题拆分为独立子问题,相关观点试图揭示可扩展且可证明的 AI 体系的设计思路。

AIScience0.2205307
ken.arneson.name
46

The value of not getting to the point (2015)不急于点题的价值(2015)ken.arneson.name

The piece argues there is value in delaying the main point to build context, illustrate nuance, and guide readers through a reasoning process.这篇文章主张在表达中延迟进入核心结论,先构建背景、展示细节,从而帮助读者建立理解。

The piece argues there is value in delaying the main point to build context, illustrate nuance, and guide readers through a reasoning process. It suggests that rushing to conclusions can undermine understanding and retention, while a slower, story-like approach can improve persuasion and recall.这篇文章主张在表达中延迟进入核心结论,先构建背景、展示细节,从而帮助读者建立理解。快速跳到结论可能削弱说服力和记忆,而更慢、带叙事性的写法往往更易于让读者理清思路并记住要点。

Culture0.223181
47

What should we tell our students?我们应该对学生说些什么?terrytao.wordpress.com

Terry Tao reflects on the messages educators should convey to students, balancing curiosity, problem‑solving mindset, and realistic expectations.陶哲轩在博文中讨论教师应向学生传达的核心信息,强调培养好奇心、解决问题的思维方式以及务实的期望。

Terry Tao reflects on the messages educators should convey to students, balancing curiosity, problem‑solving mindset, and realistic expectations. The note argues for cultivating deep understanding and resilience over chasing quick results.陶哲轩在博文中讨论教师应向学生传达的核心信息,强调培养好奇心、解决问题的思维方式以及务实的期望。主张重视扎实理解与韧性,而非追求短期成就。

Education0.2158240
48

Show HN: Making a flexible "neon" t-shirt with LED filamentsShow HN: 用 LED 细丝做柔性“霓虹”T 恤scottbezek.blogspot.com

DIY project creating a flexible neon-like T-shirt using LED filaments.一个 DIY 项目,使用 LED 细丝制作柔性霓虹效果的 T 恤。

DIY project creating a flexible neon-like T-shirt using LED filaments. It covers hardware, power management, and wearability, illustrating practical wearable electronics.一个 DIY 项目,使用 LED 细丝制作柔性霓虹效果的 T 恤。介绍硬件选型、供电与穿戴性等要点,展示了可穿戴电子的实用性。

Show HNHardware0.221937
49

AI-ready biological data: $1.8B global commitmentAI就绪生物数据:全球18亿美元承诺biohub.org

A global collaboration commits $1.8B to create AI-ready biological data, standardizing and sharing datasets to accelerate biotech research.该计划投入18亿美元,致力于打造可用于AI训练的生物数据,并推动数据标准化与共享。

A global collaboration commits $1.8B to create AI-ready biological data, standardizing and sharing datasets to accelerate biotech research. It aims to speed up model training, improve reproducibility, and advance responsible AI in the life sciences.该计划投入18亿美元,致力于打造可用于AI训练的生物数据,并推动数据标准化与共享。此举旨在加速模型训练、提升可重复性,并推动生命科学领域的负责任 AI 实践。

AIScience0.114620
50

Step 5 Preview, a 1M-context MoE from StepFun, shows up on OpenRouterStep 5 预览:StepFun 的1M上下文MoE在OpenRouter上现身openrouter.ai

StepFun's Step 5 Preview unveils a Mixture-of-Experts model with a 1 million token context, highlighting ultra-long-context capabilities and modular routing.StepFun 的 Step 5 预览展示了具备1百万上下文长度的混合专家(MoE)模型,凸显超长上下文和模块化路由能力。

StepFun's Step 5 Preview unveils a Mixture-of-Experts model with a 1 million token context, highlighting ultra-long-context capabilities and modular routing. The release on OpenRouter expands access to cutting-edge MoE tech for researchers and developers building AI tools.StepFun 的 Step 5 预览展示了具备1百万上下文长度的混合专家(MoE)模型,凸显超长上下文和模块化路由能力。该版本在 OpenRouter 上发布,进一步让研究者和开发者更容易使用前沿的 MoE 技术。

AI0.114438
51

11 of 23 Core Open Source Projects Run on 1 or 2 People23个核心开源项目中有11个由1~2人维护linuxstans.com

A look at core open source projects shows that nearly half are sustained by very small teams.核心开源项目中近一半由极小团队维持,凸显清晰的维护分工、自动化以及模块化设计在维持关键基础设施方面的重要性。

A look at core open source projects shows that nearly half are sustained by very small teams. Lean maintenance tends to rely on clear maintainership, automation, and modular design to keep critical infrastructure alive. This underscores that large organizations aren’t the only path to robust OSS ecosystems.核心开源项目中近一半由极小团队维持,凸显清晰的维护分工、自动化以及模块化设计在维持关键基础设施方面的重要性。说明大型组织并非唯一的稳健开源生态之路。

Open SourceProgrammingDevOps2.42311
52

Ideas aren't getting harder to find (2022)想法并不难找(2022)experimental-history.com

A reflection arguing that ideas are not scarce in modern times; success depends on execution, collaboration, and identifying the right problems.反思当下的想法并不稀缺,成就取决于执行、协作与发现合适的问题。

A reflection arguing that ideas are not scarce in modern times; success depends on execution, collaboration, and identifying the right problems. It challenges the notion that innovation is inherently rare and points to overlooked opportunities and cross-disciplinary blends.反思当下的想法并不稀缺,成就取决于执行、协作与发现合适的问题。挑战“创新稀缺”的观点,指出被忽视的机会与跨领域融合的潜力。

Culture2.211954
53

Anthropic's AI gave Philadelphia police a fake tip about an unsolved homicideAnthropic 的 AI 向费城警察提供了一个虚假线索theverge.com

Another report on an AI-generated tip sent to Philadelphia police about an unsolved homicide, reinforcing concerns about safety-critical deployments and the need for robust guardrails and human oversight.再次报道 AI 生成的线索被发送给费城警察,涉及未解凶杀案,强化对安全关键部署的担忧以及对健全边界和人工监督的需求。

Another report on an AI-generated tip sent to Philadelphia police about an unsolved homicide, reinforcing concerns about safety-critical deployments and the need for robust guardrails and human oversight.再次报道 AI 生成的线索被发送给费城警察,涉及未解凶杀案,强化对安全关键部署的担忧以及对健全边界和人工监督的需求。

AISecurity1.4130
apnews.com
54

Trump administration is suspending Microsoft from a green card program特朗普政府暂停微软参与绿卡计划apnews.com

Microsoft was suspended from a green-card program by the current administration, a move reflecting ongoing immigration policy shifts affecting tech recruiting.政府暂停微软参与将永久居民身份的绿卡计划,反映对技术人才移民路径的收紧趋势。

Microsoft was suspended from a green-card program by the current administration, a move reflecting ongoing immigration policy shifts affecting tech recruiting. The decision could disrupt hiring plans for large tech employers and underscores policy-risk in skilled-immigration programs.政府暂停微软参与将永久居民身份的绿卡计划,反映对技术人才移民路径的收紧趋势。此举可能扰乱大型科技公司的招聘计划,凸显科技领域移民政策的风险与博弈。

PoliticsBusiness1.48871510
55

A statement on the Tor Project's relationship with Mullvad关于 Tor 项目与 Mullvad 的关系声明blog.torproject.org

Tor Project clarifies its relationship with Mullvad, addressing funding, governance, and independence concerns.Tor 项目就与 Mullvad 的关系发布声明,涉及资金、治理与独立性等事项。

Tor Project clarifies its relationship with Mullvad, addressing funding, governance, and independence concerns. The note reassures users about how the collaboration impacts anonymity and trust.Tor 项目就与 Mullvad 的关系发布声明,涉及资金、治理与独立性等事项。该说明帮助用户理解这层合作对匿名性与信任的影响。

PrivacyOpen Source1.4115265
56

You might want to try being less creative也许你应该少一点创意blog.bawolf.com

The piece argues that forcing creativity can hinder practical outcomes; embracing constraints and disciplined approaches can yield clearer, more reliable results in software and product work.文章认为过分追求创意可能损害实际落地的效果;在约束条件下运用更克制、系统化的方法,往往能产出更清晰、可靠的软件与产品。

The piece argues that forcing creativity can hinder practical outcomes; embracing constraints and disciplined approaches can yield clearer, more reliable results in software and product work. It challenges the notion that more imagination always equals better software.文章认为过分追求创意可能损害实际落地的效果;在约束条件下运用更克制、系统化的方法,往往能产出更清晰、可靠的软件与产品。它质疑“越创越好”的观念,并强调务实的开发节奏。

Culture1.36433
news.ycombinator.com
57

Tell HN: I've been paying for a rural Tanzanian's education for 10 yearsTell HN:我已资助坦桑尼亚农村学生教育十年news.ycombinator.com

I've funded a rural Tanzanian student's education for 10 years, covering school fees and supplies.分享者表示自己资助一名坦桑尼亚农村学生教育十年,涵盖学费与学习用品等支出。

I've funded a rural Tanzanian student's education for 10 years, covering school fees and supplies.分享者表示自己资助一名坦桑尼亚农村学生教育十年,涵盖学费与学习用品等支出。该故事凸显长期资助的持续影响,并提出技术界同侪可复制或支持的低摩擦、可扩展的慈善模式。

Education1.1714220
58

Rewriting Prime Agent in Rust用 Rust 重写 Prime Agentprimeintellect.ai

Porting Prime Agent to Rust aims to improve safety and performance, leveraging Rust’s memory safety and ecosystem.将 Prime Agent 重写为 Rust 版本旨在提升安全性与性能,利用 Rust 的内存安全特性与生态系统。

Porting Prime Agent to Rust aims to improve safety and performance, leveraging Rust’s memory safety and ecosystem. The rewrite can yield more predictable behavior and better tooling, though it may require re-implementing features and verifying compatibility with existing workflows.将 Prime Agent 重写为 Rust 版本旨在提升安全性与性能,利用 Rust 的内存安全特性与生态系统。重写有望带来更可预测的行为和更优秀的工具链,但也可能需要重新实现部分功能并确保与现有工作流的兼容性。

Programming1.1163
59

Tanker attacks in Strait of Hormuz surge to wartime high, Iran tries choking oil霍尔木兹海峡油轮袭击激增至战时高位,伊朗试图阻断石油cnbc.com

Tanker attacks in the Strait of Hormuz surge to wartime highs as Iran tries choking oil flows.霍尔木兹海峡的油轮袭击激增至接近战争水平,伊朗试图切断石油运输。

Tanker attacks in the Strait of Hormuz surge to wartime highs as Iran tries choking oil flows. This raises risk for shipping, insurance, and global energy markets, underscoring how geopolitical flashpoints threaten supply chains.霍尔木兹海峡的油轮袭击激增至接近战争水平,伊朗试图切断石油运输。此举提高了航运、保险与全球能源市场的风险,突显地缘政治冲突对全球供应链的冲击。

EnergyBusiness1.02819
60

Why isn't the industry freaking out about DeepSeek 4.1 Flash?为什么业界对 DeepSeek 4.1 闪存不惊慌?dgt.is

The post questions the market's reaction to DeepSeek 4.1 Flash, perhaps arguing that the technology's impact on performance and storage economics is uncertain or anticipated.作者质疑 DeepSeek 4.1 闪存的行业影响为何未引发广泛震动,认为对性能与存储经济性的实际影响还有不确定之处。

The post questions the market's reaction to DeepSeek 4.1 Flash, perhaps arguing that the technology's impact on performance and storage economics is uncertain or anticipated. It discusses how memory-class devices and flash innovations intersect with software workloads and ecosystem readiness.作者质疑 DeepSeek 4.1 闪存的行业影响为何未引发广泛震动,认为对性能与存储经济性的实际影响还有不确定之处。文中探讨了内存类设备、闪存创新与软件工作负载、生态系统成熟度之间的关系。

Hardware0.91083954
htmx.org
61

Yes, andYes, and(是的,而且)htmx.org

The essay argues for additive collaboration in software, embracing 'Yes, and' as a mindset rather than hesitation.本文主张在软件开发中采用“Yes, and”作为协作思维,强调在不否定对方想法的前提下叠加改进。

The essay argues for additive collaboration in software, embracing 'Yes, and' as a mindset rather than hesitation. Using HTMX as a practical lens, it shows how to build incrementally while preserving compatibility.本文主张在软件开发中采用“Yes, and”作为协作思维,强调在不否定对方想法的前提下叠加改进。以 HTMX 为例,展示如何在保持向后兼容的同时实现逐步扩展。

Programming0.9719283
62

The Hetzner Cloud network stack – history and technical overviewHetzner Cloud 网络栈:历史与技术概览hetzner.com

Hetzner's technical blog offers a historical walkthrough and a technical overview of its cloud network stack, from early components to the current architecture.Hetzner 的技术博客对其云网络栈进行了从早期组件到当前架构的历史回顾和技术概览。

Hetzner's technical blog offers a historical walkthrough and a technical overview of its cloud network stack, from early components to the current architecture. It covers routing, switching, security, and performance considerations, highlighting design trade-offs and lessons learned for running scalable cloud networks.Hetzner 的技术博客对其云网络栈进行了从早期组件到当前架构的历史回顾和技术概览。文章聚焦路由、交换、安全与性能等要点,揭示了在构建可扩展云网络时的取舍与经验教训。

DevOpsSecurity0.810326
63

Why are coding agents so dumb?为什么代码代理这么蠢?mtlynch.io

The piece analyzes why current coding assistants struggle with real-world coding tasks due to brittle reasoning and misinterpretation of intent.文章分析了当前代码代理在真实编程任务中的困难,原因包括推理脆弱、对意图的误解等。

The piece analyzes why current coding assistants struggle with real-world coding tasks due to brittle reasoning and misinterpretation of intent. It advocates stronger evaluation, better prompts, and human-in-the-loop safeguards.文章分析了当前代码代理在真实编程任务中的困难,原因包括推理脆弱、对意图的误解等。主张改进评估、提升提示质量,并在必要时引入人机协作。

AIProgramming0.77572
atarimuseum.nl
64

Atari FalconAtari Falcon 计算机简史atarimuseum.nl

Atari Falcon is a late-ST era computer known for multimedia capabilities and a built-in DSP.Atari Falcon 是 ST 系列晚期的一台家用电脑,突出多媒体能力与内置 DSP。

Atari Falcon is a late-ST era computer known for multimedia capabilities and a built-in DSP.Atari Falcon 是 ST 系列晚期的一台家用电脑,突出多媒体能力与内置 DSP。文章回顾其规格、历史意义及在日益竞争的市场中的困境,展现家用计算机历史中的关键节点。

Hardware0.7112
65

Man discovers his parents' coffee machine used 1TB of data in 10 days男子发现父母的咖啡机在十天内消耗了1TB数据dexerto.com

Man discovers his parents' coffee machine used 1TB of data in 10 days一名男子发现父母的咖啡机在十天内消耗了1TB数据,揭示物联网设备背后的数据意外和隐私趣闻。

Man discovers his parents' coffee machine used 1TB of data in 10 days一名男子发现父母的咖啡机在十天内消耗了1TB数据,揭示物联网设备背后的数据意外和隐私趣闻。

Hardware0.6920559
mathstodon.xyz
66

“Math 2.0” will need to value mathematical progress more holistically“Math 2.0”将以更全面的标准评估数学进步mathstodon.xyz

Math 2.0 advocates rethinking how mathematical progress is measured, moving beyond traditional exams.Math 2.0主张重新定义数学进步的衡量标准,超越传统考试。

Math 2.0 advocates rethinking how mathematical progress is measured, moving beyond traditional exams. It emphasizes holistic indicators such as conceptual understanding, problem-solving, collaboration, and real-world impact, which could reshape curricula and assessment practices.Math 2.0主张重新定义数学进步的衡量标准,超越传统考试。它强调概念理解、问题解决、协作与实际应用等全面指标,认为这将影响课程设计与评估方式。

Education0.6598636
lwn.net
67

Compiling Rust to readable C with Eurydice用 Eurydice 将 Rust 编译成可读的 C 代码lwn.net

Eurydice demonstrates transforming Rust code into readable C, aiding debugging and cross-toolchain portability.Eurydice 展示将 Rust 代码转换成可读的 C,便于调试和跨工具链移植。

Eurydice demonstrates transforming Rust code into readable C, aiding debugging and cross-toolchain portability. This approach can help audit Rust projects and bridge interop gaps with C ecosystems, though it adds translation overhead and potential style differences.Eurydice 展示将 Rust 代码转换成可读的 C,便于调试和跨工具链移植。该方法有助于对 Rust 项目进行审核并促进与 C 生态的互操作,但会增加翻译开销与代码风格差异。

Programming0.682
68

Sharing AI progress in mathematics在数学领域分享人工智能进展openai.com

OpenAI highlights progress in AI-assisted mathematics, including automated reasoning and conjecture generation.OpenAI 展示了 AI 在数学中的进展,包括自动推理与猜想生成。

OpenAI highlights progress in AI-assisted mathematics, including automated reasoning and conjecture generation. The piece underscores collaboration with mathematicians and ongoing work to evaluate the reliability and safety of AI-assisted proofs.OpenAI 展示了 AI 在数学中的进展,包括自动推理与猜想生成。文章强调与数学界的协作,以及在评估 AI 辅助证明的可靠性与安全性方面的持续努力。

AI0.513331518
github.com
69

Once: Cache CLI commandsOnce:缓存 CLI 命令github.com

Once is a CLI utility that caches your frequently used shell commands for quick reuse.Once 是一个命令行工具,用于缓存常用命令以便快速重用。

Once is a CLI utility that caches your frequently used shell commands for quick reuse. It speeds up repetitive tasks and improves reproducibility across sessions and machines. The tool targets developers and sysadmins who want smoother, repeatable workflows.Once 是一个命令行工具,用于缓存常用命令以便快速重用。它可加速重复任务并提升跨会话、跨机器的可重复性,适合开发者与系统管理员构建更顺畅的工作流程。

ProgrammingCode0.58837
vale.rocks
70

Beauty in DVD MenusDVD 菜单之美vale.rocks

This piece celebrates the aesthetic design of DVD menus, showing how deliberate typography, motion, and navigation create a lasting UI craft.文章聚焦 DVD 菜单的设计美学,展示排版、动效与导航如何营造独特的用户体验。

This piece celebrates the aesthetic design of DVD menus, showing how deliberate typography, motion, and navigation create a lasting UI craft. It connects retro media interfaces to modern UX thinking, arguing that design quality can still inspire current products.文章聚焦 DVD 菜单的设计美学,展示排版、动效与导航如何营造独特的用户体验。将旧有媒体界面的设计价值与现代 UX 联系起来,强调优秀设计对当代产品的启发作用。

Design0.5329181
techdirt.com
71

Platforms' Violent Content Rules Are About to Meet The Pentagon's Firing Squad平台暴力内容规则将面临五角大楼的严格审查techdirt.com

Platforms may face tighter pressure to police violent content as the Pentagon signals stricter enforcement or collaboration.五角大楼释放更严格的执法信号,平台的暴力内容治理将面临更大压力。

Platforms may face tighter pressure to police violent content as the Pentagon signals stricter enforcement or collaboration. The policy dynamics could reshape moderation, content removal, and safety commitments across social media and apps.五角大楼释放更严格的执法信号,平台的暴力内容治理将面临更大压力。此一政策趋势可能重塑社媒与应用的内容审核、移除以及安全承诺。

Politics0.52410
72

open-slopware – Alternatives to FOSS projects choosing to use LLMs/AIopen-slopware:在选择使用 LLM/AI 的 FOSS 项目中的替代方案codeberg.org

open-slopware explores options for FOSS projects that adopt LLMs/AI, weighing openness, licensing, and governance implications.open-slopware 探讨在采用大型语言模型/人工智能的 FOSS 项目中可以采用的替代路径,权衡开放性、许可与治理的影响。

open-slopware explores options for FOSS projects that adopt LLMs/AI, weighing openness, licensing, and governance implications. It highlights how AI integration can complicate traditional OSS philosophies.open-slopware 探讨在采用大型语言模型/人工智能的 FOSS 项目中可以采用的替代路径,权衡开放性、许可与治理的影响。强调 AI 集成如何给传统开源理念带来新的挑战。

Open SourceAICode0.4176
antonfrolov.substack.com
73

I hired an illustrator to draw my house. Now it's my Home Assistant dashboard我请插画师画了我的房子,如今它成了我的 Home Assistant 仪表板antonfrolov.substack.com

A playful project turns a house illustration into a Home Assistant dashboard.这是一个有趣的尝试:把房子插画直接做成 Home Assistant 的仪表板。

A playful project turns a house illustration into a Home Assistant dashboard. It blends art with automation, illustrating how visual design can improve smart-home usability and engagement.这是一个有趣的尝试:把房子插画直接做成 Home Assistant 的仪表板。通过将艺术设计引入智能家居界面,展示了美观与易用性并重的积极效果。

Design0.4886199
74

OpenAI withdraws three mathematical resultsOpenAI 撤回了三项数学结果twitter.com

OpenAI withdraws three mathematical results, sparking discussion about reproducibility, peer verification, and the speed of AI research.OpenAI 撤回了三项数学结果,引发关于可重复性、同行验证与 AI 研究节奏的讨论。

OpenAI withdraws three mathematical results, sparking discussion about reproducibility, peer verification, and the speed of AI research.OpenAI 撤回了三项数学结果,引发关于可重复性、同行验证与 AI 研究节奏的讨论。

AIScience0.4336584
lwn.net
75

Reducing undefined behavior in the C language在 C 语言中降低未定义行为lwn.net

Outlines practical approaches to reduce undefined behavior in C, including compiler options, sanitizers, and safer coding practices.概述在 C 语言中减少未定义行为的实用方法,包括编译选项、动态/静态分析工具以及更安全的编码习惯。

Outlines practical approaches to reduce undefined behavior in C, including compiler options, sanitizers, and safer coding practices. The piece emphasizes how UB lurks in subtle corner cases and how tooling can curb it.概述在 C 语言中减少未定义行为的实用方法,包括编译选项、动态/静态分析工具以及更安全的编码习惯。强调 UB 常藏在细微边界之处,工具链如何帮助降低风险。

Programming0.3120130
76

Orkut.comOrkut.com:往日的社区社交网络orkut.com

Orkut, one of the early social networks, helped popularize community-based groups before Facebook dominated the space.Orkut 是早期社交网络之一,在 Facebook 主导前推动了基于社区的群组文化。

Orkut, one of the early social networks, helped popularize community-based groups before Facebook dominated the space. The revival or continued availability invites nostalgia and lessons on how online communities scale, moderate content, and sustain engagement.Orkut 是早期社交网络之一,在 Facebook 主导前推动了基于社区的群组文化。如今的回归或持续可用让人怀旧,同时也提醒我们社区如何扩展、如何进行内容治理以及如何维持活跃度的经验。

Web0.3216158
77

100+ reactions to 100+ solutions对100多个解决方案的100多条回应proofsandprompts.com

A roundup of 100+ reactions to 100+ solutions highlights how communities discuss and critique ideas, with insights into what resonates, what misses, and how feedback patterns shape problem-solving.汇总了对100多个解决方案的100多条反应,揭示社区在讨论与评判中的共振点、短板,以及反馈模式如何影响问题解决的走向。

A roundup of 100+ reactions to 100+ solutions highlights how communities discuss and critique ideas, with insights into what resonates, what misses, and how feedback patterns shape problem-solving.汇总了对100多个解决方案的100多条反应,揭示社区在讨论与评判中的共振点、短板,以及反馈模式如何影响问题解决的走向。

Culture0.34335
scottaaronson.blog
78

The Mathocalypse数学灾变scottaaronson.blog

Scott Aaronson argues that AI and automated reasoning are transforming mathematical practice, challenging old assumptions about proofs and discovery.Scott Aaronson 认为人工智能和自动推理正在改变数学研究的实践,挑战关于证明与发现的传统观念,重新定义何为有意义的数学成果,并强调人类直觉的重要性。

Scott Aaronson argues that AI and automated reasoning are transforming mathematical practice, challenging old assumptions about proofs and discovery. The piece reframes what counts as a mathematical result and why human intuition matters.Scott Aaronson 认为人工智能和自动推理正在改变数学研究的实践,挑战关于证明与发现的传统观念,重新定义何为有意义的数学成果,并强调人类直觉的重要性。

AIScience0.3388403
reddit.com
79

I think I found a planet nobody knew existed. I used Claude Code to find it我用 Claude Code 找到了一个没人知道的行星reddit.com

Using Claude Code, the author claims to have found a planet nobody knew existed, illustrating how AI-assisted coding can aid scientific discovery.用 Claude Code 找到了一颗此前无人知晓的行星,展示了 AI 辅助编程在科学发现中的潜力。

Using Claude Code, the author claims to have found a planet nobody knew existed, illustrating how AI-assisted coding can aid scientific discovery.用 Claude Code 找到了一颗此前无人知晓的行星,展示了 AI 辅助编程在科学发现中的潜力。

AIScience0.3261127
80

OTel-Native by Design – Building Products That Export to Any Observability StackOTel 原生设计:打造可导出到任意观测栈的产品opentelemetry.io

Design products with OpenTelemetry in a backend-agnostic way so telemetry can be exported to any observability stack.提出以 OpenTelemetry 的原生设计为核心,构建可对接任意观测后端的产品,不再绑定到单一栈。

Design products with OpenTelemetry in a backend-agnostic way so telemetry can be exported to any observability stack. It covers exporter architecture and the benefits of vendor-agnostic instrumentation, showing how this reduces lock-in while expanding interoperability across backends.提出以 OpenTelemetry 的原生设计为核心,构建可对接任意观测后端的产品,不再绑定到单一栈。内容聚焦导出器架构与厂商无关的观测数据语义,说明这种做法如何降低锁定、提升对不同后端的互操作性。

Open SourceDevOpsWeb0.25511
letsencrypt.org
81

Let's Encrypt: 64-Day Certificate Lifetimes By Default Coming Feb 2027Let's Encrypt:默认证书有效期将至64天,2027年2月起实施letsencrypt.org

Let's Encrypt will shorten default certificate lifetimes to 64 days starting February 2027, shortening lifespans to reduce risk of compromised keys and improve automation.Let's Encrypt 将于 2027 年 2 月起默认将证书有效期缩短至 64 天,以降低密钥被攻击的风险并提升自动化续签能力。

Let's Encrypt will shorten default certificate lifetimes to 64 days starting February 2027, shortening lifespans to reduce risk of compromised keys and improve automation. The change nudges operators toward more frequent renewal and resilience.Let's Encrypt 将于 2027 年 2 月起默认将证书有效期缩短至 64 天,以降低密钥被攻击的风险并提升自动化续签能力。此变更将促使运营者更频繁地进行证书更新以增强韧性。

Security0.2117
82

Anne Carson wins Nobel Prize in literature 2026安妮·卡森获2026年诺贝尔文学奖theguardian.com

Anne Carson wins the Nobel Prize in Literature 2026, honoring her genre-spanning work that blends poetry, classical scholarship, and translation.安妮·卡森获2026年诺贝尔文学奖,表彰她在诗歌、古典研究与翻译等领域的跨界贡献。

Anne Carson wins the Nobel Prize in Literature 2026, honoring her genre-spanning work that blends poetry, classical scholarship, and translation. The decision highlights cross-genre literary achievement and could raise broader attention to experimental forms and Canadian authors.安妮·卡森获2026年诺贝尔文学奖,表彰她在诗歌、古典研究与翻译等领域的跨界贡献。此殊荣凸显跨体裁成就的重要性,或提升对实验性写作及加拿大作家的关注。

Culture0.217344
newgtldprogram-aps.icann.org
83

New gTLD Application for .lan.lan 新 gTLD 投标newgtldprogram-aps.icann.org

ICANN reveals a new gTLD application for .lan, prompting questions about private namespaces and local naming conventions.ICANN 公布了关于 .lan 的新 gTLD 申请,引发对私有命名空间和本地命名约定的讨论。

ICANN reveals a new gTLD application for .lan, prompting questions about private namespaces and local naming conventions. The move could raise conflicts with existing DNS standards and practical local network use.ICANN 公布了关于 .lan 的新 gTLD 申请,引发对私有命名空间和本地命名约定的讨论。此举可能与现有 DNS 标准及本地网络使用发生冲突。

Legal0.2130149
84

US proposes $100k charge for international students to do post-graduate work美国拟向国际学生毕业后工作征收10万美元费用nature.com

The report discusses a US proposal to charge international students up to $100k for post-graduate work eligibility.报道指出美国拟对国际学生毕业后工作资格征收最高10万美元的费用。

The report discusses a US proposal to charge international students up to $100k for post-graduate work eligibility. If enacted, the policy could deter foreign talent, affect universities’ diversity, and reshape funding for graduate pathways.报道指出美国拟对国际学生毕业后工作资格征收最高10万美元的费用。若实施,该政策可能削弱对外国人才的吸引力、影响高校多样性,并改变研究生路径的资金结构。

PoliticsEducation0.22620
85

Tesla Rebrands Full Self-Driving as Tesla Assisted Driving in Europe特斯拉在欧洲将全自动驾驶改名为特斯拉辅助驾驶qz.com

Tesla Europe renames Full Self-Driving to Tesla Assisted Driving to align with regulatory expectations and consumer perception, signaling branding and safety messaging shifts across markets.特斯拉在欧洲将‘全自动驾驶’改名为‘特斯拉辅助驾驶’,以符合监管要求并调整市场对安全信息的理解,标志着品牌传达的变革。

Tesla Europe renames Full Self-Driving to Tesla Assisted Driving to align with regulatory expectations and consumer perception, signaling branding and safety messaging shifts across markets.特斯拉在欧洲将‘全自动驾驶’改名为‘特斯拉辅助驾驶’,以符合监管要求并调整市场对安全信息的理解,标志着品牌传达的变革。

BusinessPolitics0.21313
news.ycombinator.com
86

Ask HN: What do you run on a $5 VPS that's worth keeping online 24/7?Ask HN: 你在5美元VPS上运行的、值得24/7 在线的内容是什么?news.ycombinator.com

An Ask HN discussion on what services people run on a $5 VPS that are worth keeping online 24/7, sharing lightweight, resilient workloads and monitoring patterns.讨论聚焦在5美元 VPS 上能长期在线且值得维护的服务,分享轻量级、鲁棒的工作负载与监控方法。

An Ask HN discussion on what services people run on a $5 VPS that are worth keeping online 24/7, sharing lightweight, resilient workloads and monitoring patterns.讨论聚焦在5美元 VPS 上能长期在线且值得维护的服务,分享轻量级、鲁棒的工作负载与监控方法。

DevOps0.2383616
thequietus.com
87

US man given prison sentence for bot-farming music streams美国男子因机器人刷流音乐被判入狱thequietus.com

A man was sentenced for bot-farming streams to inflate plays on streaming platforms.一名男子因利用机器人刷流来抬高音乐平台播放量而被判刑。

A man was sentenced for bot-farming streams to inflate plays on streaming platforms. The case underscores legal risks of automation-enabled fraud and the crackdown on streaming manipulation.一名男子因利用机器人刷流来抬高音乐平台播放量而被判刑。此案凸显自动化欺诈的法律风险以及对流媒体操控的打击。

Legal0.2181277
github.com
88

DuckDB DucklakeDuckDB 的 Ducklakegithub.com

Ducklake is a DuckDB project introducing data-lake integration that enables querying across data lakes with DuckDB.Ducklake 是 DuckDB 的数据湖集成项目,使用户能够用 DuckDB 跨数据湖进行查询分析。

Ducklake is a DuckDB project introducing data-lake integration that enables querying across data lakes with DuckDB. It demonstrates how DuckDB can bridge lakehouse data sources for analytics.Ducklake 是 DuckDB 的数据湖集成项目,使用户能够用 DuckDB 跨数据湖进行查询分析。它展示了 DuckDB 在连接湖仓数据源以实现分析的潜力。

DataOpen SourceCode0.122832
89

OLED burn-in test: 30-month updateOLED烧屏测试:30个月更新techspot.com

A 30-month update on OLED burn-in tests shows ongoing image retention risks despite improvements in newer panels.对 OLED 烧屏的长期测试在30个月更新中显示,尽管新一代面板有所改进,但仍存在残留影像的风险。

A 30-month update on OLED burn-in tests shows ongoing image retention risks despite improvements in newer panels. The study underscores that static content and uniform brightness patterns still pose a risk, though real-world usage mitigates some effects. For developers and manufacturers, it emphasizes the need for software features like pixel shifting and content variety to extend panel life.对 OLED 烧屏的长期测试在30个月更新中显示,尽管新一代面板有所改进,但仍存在残留影像的风险。研究指出静态内容和均匀亮度模式易诱发烧屏,实际使用可以缓解部分问题。对于开发者与厂商,这强调了像素位移等软件特性和内容多样性以延长屏幕寿命的必要性。

Hardware0.111583
nytimes.com
90

'Breathtaking,' 'Devastating': Mathematics Reels After New OpenAI Release数学界对 OpenAI 新发布反应强烈且分化nytimes.com

'Breathtaking,' 'Devastating': Mathematics Reels After New OpenAI Release数学界对 OpenAI 最新发布引发强烈且分化的反应,既有惊叹之声,也有否定与担忧,凸显在能力边界、安全性与理论影响上的热烈辩论。

'Breathtaking,' 'Devastating': Mathematics Reels After New OpenAI Release数学界对 OpenAI 最新发布引发强烈且分化的反应,既有惊叹之声,也有否定与担忧,凸显在能力边界、安全性与理论影响上的热烈辩论。

AI0.11116
naturalsystems.io
91

ETH-68: Ethernet Audio Interface for LinuxETH-68:适用于 Linux 的以太网音频接口naturalsystems.io

A new Ethernet-based audio interface for Linux, aiming to reduce latency and enable distributed audio setups.新型以太网音频接口针对 Linux,旨在降低延迟并支持分布式音频。

A new Ethernet-based audio interface for Linux, aiming to reduce latency and enable distributed audio setups. If the project delivers, it could influence open hardware and Linux audio workflows.新型以太网音频接口针对 Linux,旨在降低延迟并支持分布式音频。若落地,将影响开源硬件和 Linux 音频工作流。

HardwareOpen Source0.1212134
github.com
92

Sub-1-Bit LLM Compression via Latent Factorization通过潜在因子分解实现 Sub-1-Bit 的大语言模型压缩github.com

LittleBit uses latent factorization to compress large language models to sub-1-bit representations, aiming to retain useful inference quality at far smaller sizes.LittleBit 通过潜在因子分解把大模型压缩到低于1比特的表示,意在在显著减小模型规模的同时维持可用的推理质量。

LittleBit uses latent factorization to compress large language models to sub-1-bit representations, aiming to retain useful inference quality at far smaller sizes. The approach could unlock running capable LLMs on edge devices and specialized accelerators, expanding deployment scenarios.LittleBit 通过潜在因子分解把大模型压缩到低于1比特的表示,意在在显著减小模型规模的同时维持可用的推理质量。该方法有望让边缘设备和定制加速器也能部署强大的 LLM,拓展应用场景。

AICode0.18624
github.com
93

Show HN: K10s – A Clickable Kubernetes TUI (Go, Bubble Tea)Show HN:K10s——一个可点击的 Kubernetes TUI(Go、Bubble Tea)github.com

K10s is a clickable terminal UI for Kubernetes built with Go and Bubble Tea, enabling point-and-click operations inside a TUI instead of pure command lines.K10s 是一个用 Go 和 Bubble Tea 构建的可点击 Kubernetes 终端界面,允许在 TUI 内进行点选操作而非仅用命令行。

K10s is a clickable terminal UI for Kubernetes built with Go and Bubble Tea, enabling point-and-click operations inside a TUI instead of pure command lines. By combining a familiar CLI domain with a UI that can be navigated via keyboard or mouse, it aims to reduce the cognitive load of managing clusters in the terminal. This project highlights how terminal interfaces can augment traditional DevOps tooling without sacrificing portability.K10s 是一个用 Go 和 Bubble Tea 构建的可点击 Kubernetes 终端界面,允许在 TUI 内进行点选操作而非仅用命令行。它将熟悉的命令行领域与可键盘/鼠标导航的 UI 结合,旨在降低终端管理集群的认知负担。该项目展示了终端工具如何在不牺牲可移植性的前提下变得更易用。

Show HNProgrammingDevOpsCode0.16549
94

Archaeologists Are Reconstructing the 'Invisible' Technologies of the Stone Age考古学家正在重建石器时代的“看不见的”技术smithsonianmag.com

Archaeologists are piecing together the hidden technologies of the Stone Age—like rope, thread, and twine—to infer how ancient people manipulated fibers for construction, hunting, and daily life.考古学家正拼接石器时代隐藏的技术线索,如绳索、线和细绳,推断古人如何利用纤维完成建造、狩猎与日常活动。

Archaeologists are piecing together the hidden technologies of the Stone Age—like rope, thread, and twine—to infer how ancient people manipulated fibers for construction, hunting, and daily life. By visualizing the steps and tools involved, researchers illuminate early material science, craft specialization, and social coordination that underpinned long-term survival.考古学家正拼接石器时代隐藏的技术线索,如绳索、线和细绳,推断古人如何利用纤维完成建造、狩猎与日常活动。通过还原纤维制作的步骤与工具,研究者揭示早期材料科学、工艺分工与社会协作对生存的支撑作用。

Science0.114758
95

Show HN: Jevman – AI decision models play Pac-ManShow HN:Jevman——AI 决策模型挑战 Pac-Manopper.ai

Jevman uses AI decision models to play Pac-Man, serving as a benchmark for how agents plan and react in a classic arcade environment.Jevman 将 AI 决策模型用于玩 Pac-Man,作为测试 AI 代理在实时游戏环境中的计划与反应能力的基准。

Jevman uses AI decision models to play Pac-Man, serving as a benchmark for how agents plan and react in a classic arcade environment. It highlights the capabilities and limits of current AI decision systems when faced with real-time gaming challenges and simple rules.Jevman 将 AI 决策模型用于玩 Pac-Man,作为测试 AI 代理在实时游戏环境中的计划与反应能力的基准。该示例揭示了现阶段 AI 决策系统在面对简单规则和即时挑战时的潜力与局限。

Show HNAIGamingWeb0.16516
96

A 5.3M-year-old deep-sea whale necropolis in the Diamantina ZoneDiamantina 区域的 530 万年前深海鲸类尸葬nature.com

Researchers describe a 5.3-million-year-old deep-sea whale necropolis in the Diamantina Zone.研究者在 Diamantina 区发现一个 530 万年前的深海鲸鱼尸葬群,显示深海条件对化石保存极有利。

Researchers describe a 5.3-million-year-old deep-sea whale necropolis in the Diamantina Zone. The site sheds light on deep-sea preservation and ancient whale ecology, enriching our understanding of early marine ecosystems.研究者在 Diamantina 区发现一个 530 万年前的深海鲸鱼尸葬群,显示深海条件对化石保存极有利。此发现为理解史前海洋生物与深海生态提供新线索。

Science0.114014
github.com
97

Show HN: Edi Life OS – self-hosted life dashboard with an MCP server for AIShow HN:Edi Life OS——具有 MCP 服务器的自托管生活仪表盘github.com

Show HN: Edi Life OS is a self-hosted life dashboard with an MCP server enabling AI features for local control.Show HN:Edi Life OS 是一个自托管的生活仪表盘,带有用于本地 AI 的 MCP 服务器。

Show HN: Edi Life OS is a self-hosted life dashboard with an MCP server enabling AI features for local control. It aggregates notes, tasks, calendars, and AI-driven insights in a self-contained stack.Show HN:Edi Life OS 是一个自托管的生活仪表盘,带有用于本地 AI 的 MCP 服务器。它将笔记、任务、日历与基于 AI 的洞察整合在一个自包含的系统中。

Show HNAIOpen SourceCode0.13512
werwolv.net
98

Getting old Macromedia Director Games to run on modern Hardware让老旧的 Macromedia Director 游戏在现代硬件上运行werwolv.net

Old Macromedia Director games can be run on modern hardware using wrappers and emulation, with practical tips for preservation.老旧的 Macromedia Director 游戏可以通过封装与仿真等方法在现代硬件上运行,文中给出实用的保留技巧与步骤。

Old Macromedia Director games can be run on modern hardware using wrappers and emulation, with practical tips for preservation. The post offers steps to keep classic titles playable.老旧的 Macromedia Director 游戏可以通过封装与仿真等方法在现代硬件上运行,文中给出实用的保留技巧与步骤。

HardwareProgramming0.1104
99

Scaling and benchmarking a critical message bus using a new indexing strategy使用新的索引策略扩展并基准测试关键消息总线blog.janestreet.com

Jane Street discusses scaling a critical message bus with a new indexing strategy, aiming to improve throughput and latency under load.Jane Street 分享了使用新索引策略对关键消息总线进行扩展与基准测试的经验,聚焦在高并发下提升吞吐与延迟表现。

Jane Street discusses scaling a critical message bus with a new indexing strategy, aiming to improve throughput and latency under load.Jane Street 分享了使用新索引策略对关键消息总线进行扩展与基准测试的经验,聚焦在高并发下提升吞吐与延迟表现。文章探讨架构取舍及构建高可用发布-订阅基础设施的经验教训。

ProgrammingDevOps0.1474
mitchellh.com
100

A Terminal Protocol for Program Status (OSC 7501)用于程序状态的终端协议(OSC 7501)mitchellh.com

Proposes a terminal-based protocol to expose program status, enabling lightweight status reporting in CLI apps.提出一种用于在终端暴露程序状态的协议,使命令行应用可以轻量级地报告健康与进度。

Proposes a terminal-based protocol to expose program status, enabling lightweight status reporting in CLI apps. OSC 7501 seeks to standardize health and progress signals for tooling and orchestration.提出一种用于在终端暴露程序状态的协议,使命令行应用可以轻量级地报告健康与进度。OSC 7501 旨在标准化状态信号,便于工具链和作业编排使用。

ProgrammingHealth0.117452
101

The dawn of the age of the exoskeleton外骨骼时代的到来theconversation.com

The piece argues exoskeletons are entering mainstream use across industries and daily life.文章认为外骨骼技术正进入主流应用,包括工业、医疗与日常生活。

The piece argues exoskeletons are entering mainstream use across industries and daily life. It discusses challenges like power, control, safety, and accessibility, and considers implications for work and human augmentation.文章认为外骨骼技术正进入主流应用,包括工业、医疗与日常生活。讨论了能源、控制、安全与普及性等挑战,并探讨对劳动、健康与人类增强的影响。

Science0.18043
102

Show HN: Pocketty – iPhone SSH terminal that pings you when an agent is blockedShow HN:Pocketty——iPhone SSH 终端,代理被阻断时会提醒你pocketty.app

Pocketty offers an iPhone SSH terminal with monitoring that notifies you when the SSH agent becomes blocked, helping maintain awareness on mobile development workflows.Pocketty 提供一款 iPhone SSH 终端,具备在代理被阻断时给出通知的监控功能,帮助移动端开发工作流保持连接可用性。

Pocketty offers an iPhone SSH terminal with monitoring that notifies you when the SSH agent becomes blocked, helping maintain awareness on mobile development workflows. It pairs a lightweight mobile client with alerting to keep connections alive.Pocketty 提供一款 iPhone SSH 终端,具备在代理被阻断时给出通知的监控功能,帮助移动端开发工作流保持连接可用性。该应用将轻量级客户端与告警机制结合,提升运维与开发效率。

Show HNProgrammingDevOps0.12813
oss.lsantos.dev
103

Show HN: Proton Drive for LinuxShow HN:Linux 版 Proton Driveoss.lsantos.dev

Linux users can now mount Proton Drive as a filesystem, enabling native access to encrypted cloud storage.Linux 用户现在可以将 Proton Drive 挂载为文件系统,直接本地访问经过加密的云存储。

Linux users can now mount Proton Drive as a filesystem, enabling native access to encrypted cloud storage. The development expands secure-by-default cloud storage options for Linux users.Linux 用户现在可以将 Proton Drive 挂载为文件系统,直接本地访问经过加密的云存储。此进展为 Linux 用户扩展了默认安全性高的云存储选项。

Show HNOpen Source0.03312
104

Show HN: OldRoll, a free vintage photo editor for the browserShow HN:OldRoll 是一款免费浏览器端复古风格照片编辑器oldroll.io

OldRoll brings vintage photo editing directly in the browser with a free, browser-based editor.OldRoll 是一款在浏览器中即可使用的免费复古风格照片编辑器,提供怀旧滤镜与编辑工具。

OldRoll brings vintage photo editing directly in the browser with a free, browser-based editor. It showcases how modern web tech can deliver retro aesthetics and solid editing features without desktop apps.OldRoll 是一款在浏览器中即可使用的免费复古风格照片编辑器,提供怀旧滤镜与编辑工具。展示了现代网页技术如何在无需桌面应用的情况下实现专业级的图像处理体验。

Show HNWeb0.0112
loc.gov
105

Receipt for sale of Jane, age 18, and her son, Henry, age 1对 18 岁 Jane 与 1 岁 Henry 的出售凭证loc.gov

Historical artifact depicting a sale of enslaved woman Jane and her infant Henry, shedding light on the human cost of slavery.历史性文献记录了对18岁奴隶女子 Jane 与其婴儿 Henry 的出售,揭示奴隶制的人身代价及其在档案中的痕迹。

Historical artifact depicting a sale of enslaved woman Jane and her infant Henry, shedding light on the human cost of slavery. The item helps illustrate the lived experiences behind archival records.历史性文献记录了对18岁奴隶女子 Jane 与其婴儿 Henry 的出售,揭示奴隶制的人身代价及其在档案中的痕迹。该物件帮助人们理解历史记录背后的真实生活。

Culture0.0110
106

Training Text-to-Image Models Without a VAE在不使用 VAE 的情况下训练文本到图像模型linum.ai

A new method trains text-to-image models without a variational autoencoder, proposing alternative encoders or training tricks to simplify pipelines and potentially improve sample quality or training efficiency.提出在不使用 VAE 的情况下训练文本到图像的模型的新方法,借助替代编码器或训练技巧简化流程,并可能提高样本质量和训练效率。

A new method trains text-to-image models without a variational autoencoder, proposing alternative encoders or training tricks to simplify pipelines and potentially improve sample quality or training efficiency.提出在不使用 VAE 的情况下训练文本到图像的模型的新方法,借助替代编码器或训练技巧简化流程,并可能提高样本质量和训练效率。

AI0.04917
github.com
107

Show HN: Rgpu – a PyTorch device whose tensors live on a remote GPUShow HN: Rgpu – 一个让张量驻留在远程GPU上的 PyTorch 设备github.com

Rgpu is a PyTorch device where tensors live on a remote GPU, enabling offloading and potential speedups for large models.Rgpu 是一个让张量驻留在远程 GPU 的 PyTorch 设备,支持将计算分发到另一台主机以进行加速。

Rgpu is a PyTorch device where tensors live on a remote GPU, enabling offloading and potential speedups for large models. It highlights tradeoffs in latency, bandwidth, and distributed computation.Rgpu 是一个让张量驻留在远程 GPU 的 PyTorch 设备,支持将计算分发到另一台主机以进行加速。它揭示了延迟、带宽与分布式计算的权衡。

Show HNAIProgrammingHardwareCode0.0482
108

The role of cat eye narrowing movements in cat–human communication (2020)猫眼眯眼动作在猫-人沟通中的作用(2020)nature.com

The study examines how cats use pupil constriction as a social cue to humans, complementing vocalizations and body language.研究发现猫在与人互动时通过瞳孔收缩等方式传达信号,辅以叫声和身态语言。

The study examines how cats use pupil constriction as a social cue to humans, complementing vocalizations and body language. Findings suggest humans respond to these cues, revealing nuanced interspecies communication and practical welfare implications for pet care.研究发现猫在与人互动时通过瞳孔收缩等方式传达信号,辅以叫声和身态语言。结果显示人类对这些信号有反应,揭示了物种间沟通的细微之处及对宠物福利的启示。

Science0.03621
109

Show HN: TerrainSR – fast, realistic heightmap upscaling modelShow HN: TerrainSR – 快速、真实的高度图放大模型huggingface.co

TerrainSR is a fast heightmap upscaling model that produces realistic terrain features, useful for games and simulations.TerrainSR 是一个快速的高度图放大模型,能够生成真实的地形特征,适用于游戏和仿真场景。

TerrainSR is a fast heightmap upscaling model that produces realistic terrain features, useful for games and simulations.TerrainSR 是一个快速的高度图放大模型,能够生成真实的地形特征,适用于游戏和仿真场景。

Show HNAICode0.0357
build2.org
110

How to Fix autoconf-style Configuration Probing如何修复 autoconf 风格的配置探测build2.org

Practical guidance for stabilizing Autoconf-style configuration probes, reducing false positives, and improving cross-system portability.给出在 Autoconf 风格配置探测中提升稳定性、降低误报、增强跨系统可移植性的实用方法,并讨论简化配置流程的替代方案。

Practical guidance for stabilizing Autoconf-style configuration probes, reducing false positives, and improving cross-system portability. The post also discusses alternatives or improvements to streamline the configuration workflow.给出在 Autoconf 风格配置探测中提升稳定性、降低误报、增强跨系统可移植性的实用方法,并讨论简化配置流程的替代方案。

Programming0.0104
researchgate.net
111

Part of a police code for describing suspects描述嫌疑人的警察代码部分researchgate.net

The study maps the linguistics of police codes used to describe suspects, shedding light on how coded language encodes assumptions, priorities, and potential biases.研究整理了用于描述嫌疑人的警察代码及其含义,揭示了代码化语言在反映假设、优先级以及潜在偏见方面的作用。

The study maps the linguistics of police codes used to describe suspects, shedding light on how coded language encodes assumptions, priorities, and potential biases. It offers a lens into how language design intersects with law enforcement practices.研究整理了用于描述嫌疑人的警察代码及其含义,揭示了代码化语言在反映假设、优先级以及潜在偏见方面的作用。提供了一个观察语言设计如何与执法实践交互的视角。

CultureLegal0.090