表妹高三专业文科 400 分 让她自学 ai 技能做互联网/自媒体业务然后去应聘可能吗，比如深圳、江浙，我不太看好上大专，大专真有必要上马？ – V2EX 首页 注册 登录 V2EX = way to explore V2EX 是一个关于分享和探索的地方 现在注册 已注册用户请 登录 V2EX › 问与答 表妹高三专业文科 400 分 让她自学 ai 技能做互联网/自媒体业务然后去应聘可能吗，比如深圳、江浙，我不太看好上大专，大专真有必要上马？ fromKansasCity · 6 小时 40 分钟前 · 1154 次点击 33 条回复 • 2026-02-16 22:34:55 +08:00 1 xuld 6 小时 33 分钟前 想多了。能上大学就上。学历不重要。 2 w2er 6 小时 22 分钟前 1 听一楼的 就是亲妹也不要不看好她继续上学 3 hefish 6 小时 17 分钟前 总比在外面浪好。 专科虽差，也是一个小社会，慢慢接触。 4 Hyschtaxjh 6 小时 14 分钟前 via iPhone 大專也比高中學歷強。。 5 renmu 6 小时 12 分钟前 via Android 2 我的建议是不要建议，不然未来就是都怪表哥不让我上学，该死的原生家庭 6 xloger 6 小时 10 分钟前 必然是有的读书比没得读好啊，现在社会环境多恶劣，多读几年缓冲成长一下总是好的，尤其是女生。 7 dna1982 6 小时 5 分钟前 该上还是得上，不过注意打扮打扮，早点下手给自己找个理想老公。 大学这几年估计是她遇见金龟的唯一机会了。 8 ffgrinder 5 小时 59 分钟前 我的建议是不要建议，不然未来就是都怪表哥不让我上学，该死的原生家庭 9 oldlamp 5 小时 37 分钟前 哪怕是民办本科，也是本科，考公考编考研，都比专科强一些。 10 hammy 5 小时 16 分钟前 现在 400 多，高考可能发挥好一点，上个本科也行。上不去本科，选个好专科也足够分数了。或者各个省还有征集志愿。哪怕去民办本科，毕业也有考公的资格。 表格劝妹不读书，我不理解。 11 irisdev 4 小时 40 分钟前 你表妹和她父母鸟你吗。。。 12 chenliangngng 4 小时 34 分钟前 但读书，莫问 13 fromKansasCity OP 4 小时 19 分钟前 我破本科啊 没觉得学历有啥用。。。我满脑子想着让他掌握信息搜集学习能力然后 all in Ai/aigc 做业务才是现在的出路。 大专每分每秒都是浪费时间。。。花钱买一张废纸然后被学校卖掉 14 Biggoldfish 4 小时 19 分钟前 说句不好听的要有这自学能力不至于就考 400 分 15 WuDiHaiTai 4 小时 18 分钟前 1 @ fromKansasCity 你看现在社会招聘还有招高中学历的吗？ 16 EchoWhale 4 小时 13 分钟前 via iPhone 关你啥事？ 17 EchoWhale 4 小时 12 分钟前 via iPhone 读书有没有都让他自己决定，以后后悔了怪你一辈子 18 Muniesa 4 小时 11 分钟前 via Android 有学历的时候自然觉得学历没啥用，没有的时候寸步难行 19 saltedfishgames 4 小时 11 分钟前 via Android @ fromKansasCity 家里有资源托举还是有渠道保就业？都没有的话肯定先拿下学历这张底牌啊，哪怕是个民办本科那也是本还有机会考事业编或者公务员。没学历真的没有保底，混的更惨 20 leegradyllljjjj 3 小时 58 分钟前 via iPhone 小朋友，好好毒树 21 WuDiHaiTai 3 小时 57 分钟前 2 没话讲，我的建议是不要建议。op 对大专的有色眼镜也太大了，感觉是没进过社会的。 你希望你的亲人以后去蜜雪冰城捣奶茶挣 3000 块，还是希望她去办公室坐着领 3000 块？ 学历从来都不是 100%的护身符，我的理解一直认为学历是入场券。去 BOSS 直聘搜审核/标注/拉框，看看是不是绝大部分都是大专要求，我甚至还见过只要本科以上的公司。标注难吗？我觉得别说大专门槛，给初中学历，认识中国字的就能干。但你没大专学历，人家就是不要你，这叫大专浪费时间吗？ 技术好 OK ，我实习同事是中专学历，全司最低学历，仅高于保洁这一块。部门里最能干的，却是部门里工资最低的，人家就玩他，就让他多干活又能怎样，他不敢不干，因为他知道出了这家的门，再去别家就找不到工作了。他去年花钱去买含金量等于 0 的成人学历，这辈子最后悔的事就是当时没去读个大专。 > 大专每分每秒都是浪费时间。 你读过大专吗你就锐评，少刷点抖音 b 站吧，看一群没上过班的本科生说的话有什么意义。这个问题我绝对可以发言，我亲自从垃圾大专读到垃圾本科，现在在五百强外企混的人模狗样，我自己吃到什么样的教育红利我知道，我对见过的每一个读书不好的人都说，如果你想读，就一定要去读，尤其懒狗类型的，因为连学历都没有，出门更是难上加难。如果没有这张文凭，我现在都不会在这里和你对话，我不敢想我在做什么样的工作。 以后遇到这样的问题就学学我，鼓励，鼓励是不会犯错的，怎么活是人家的事，不要乱出主意毕竟我们也没有在自己看好的方向做出多大成绩。 22 WuDiHaiTai 3 小时 50 分钟前 @ WuDiHaiTai #21 我的大专同学，没什么混得好的， 玩的好的哥们在工厂做技术工，月薪一万多。相比之下没学历的工人只能做流水线，月薪 5000 多顶天了。 其余零零散散的，只要想上班，干个运营，教培，干啥都能做个人样。 大专同学专升本之后有去上海、阿里做程序员的。 所以停止歧视吧。 23 diudiuu 3 小时 48 分钟前 能问出这种话的，看起来也是没有读下书的 24 irisdev 3 小时 28 分钟前 via Android @ WuDiHaiTai #22 现在不一样了，我同事 80 后有大专的，这两年进来的 211 本起步，你说的阿里开发现在也不可能要第一学历大专的了。不过大专怎样也算高等教育，跟高中的差距比本科和专科的差距大多了。 25 meteor957 3 小时 4 分钟前 via Android 你有资格给别人建议嘛… 26 LandCruiser 2 小时 38 分钟前 不可能，大学随便选一个学校学法学专业，然后过法考，考不发达地区的公检法，你表妹的最优解 27 docx 2 小时 25 分钟前 via iPhone 不要强加自己的想法，AI 不 AI 的不重要，她自己愿意去干嘛就去干嘛，不然到时候找不到工作还要怪你了 28 Planarians 1 小时 59 分钟前 via iPhone 要么你能手把手的教他做业务赚钱 要么就别瞎建议 29 0x0x 1 小时 32 分钟前 via iPhone 你觉的你现在自己去 all in ai 你能赚到多少钱？ 如果你有成熟可以复制的路子，你的建议没问题。 如果你自己也赚不到钱，那就别建议了。 30 jackOff 1 小时 30 分钟前 via An

Source: V2EX | Original Link

2025 父亲走了, 很怀念他 – V2EX 首页 注册 登录 V2EX = way to explore V2EX 是一个关于分享和探索的地方 现在注册 已注册用户请 登录 V2EX › 生活 2025 父亲走了, 很怀念他 2 contradictspiral · 23 小时 51 分钟前 · 2184 次点击 父亲在 2025 年因病离世，如今新一年春节又即将来临，去年春节全家和亲戚去广西自驾了玩一圈，当时父亲在路上还说起明年准备去张家界，路线都规划好了，今年这段旅程终究是无法成行了，很是怀念。当时父亲离去后，心情沉痛，在这疾病的一个论坛里记录了父亲最后的 [病情历程] ，不过心里其实还有很多话想说，想找一个树洞再写写我的父亲。 父亲生于 1964 年，2024 年 2 月初因为持续的乏力，面部浮肿，去检查后确诊为淋巴瘤，一种恶性的血液肿瘤，确诊时为滤泡淋巴瘤 3B 级，属于晚期了，这给我们以及临近退休的父亲带来了沉重的打击。但当时查资料了解到滤泡淋巴瘤属于一种惰性淋巴瘤，并且淋巴瘤主要是看分型不看分期，对化疗很敏感，有规范的治疗方案，虽然当时也担心这种重大疾病肯定对生命有影响，想过与父亲离别的一天可能会提前到来，但是真没想到会来的这么快。 父亲年轻时成绩不错，考上大学学的冶炼相关专业，在一座西南小城的国企中工作了一辈子，技术出身不太擅长应酬的他从普通的技术员逐渐做到一个小头目，至少对我这三十多还在各种二三流小公司打工的 IT 民工来说，还是很钦佩的，他为我们家提供了不算富裕但也足够的经济基础，是我这个样样很普通，到处瞎折腾却也还算活得下去的底气。他年轻时工作总是很忙，在我上中学后又调到离市区很远的一个县，可能一两周才能回来一次，有时候感觉他事情很多，回来了很疲惫话也不多，虽然在我成长中少了很多陪伴，不过他和我说话时总是很温和，偶尔会夹杂一些父亲他自带的独特风趣，我学习上每次跟他说要买资料报班啥的也从不吝啬，我很喜欢也很感谢他，他一直是我心中一座靠山。 之前父亲的身体状况一直还算不错，他这辈子也经历了很多事，操劳了很多事，临近退休的几年，因为工作单位上出了些事受到了影响，从一线位置退居闲职，虽然没那么多事情操心了，但心里也受了很大打击，他从未和我详细讲过这番经历，但想必这过程还是很曲折的。后来很多人说父亲遭这病必定有很大这方面因素的影响，不过已经过去很多年了，我是感觉父亲也逐渐放下了，他有了更多自己的时间，也开始规划着退休生活，学习拍照摄影视频剪辑，AI 刚出来那会他对各种 AI 的文案图片编辑，视频生成能力掌握的比我还 6 ，经常在家族群里展示展示他新剪出来的视频，给大家带来了很多欢乐。我们一家人都比较内向话不多，父亲算是最 E 的人，他在时家里气氛还算偶有活跃，如今他走了家里便也冷清了许多。 这是父亲最后一段平和而又温馨的时光，我一直觉着父亲这样的状态也还不错，只是他常常说睡不着觉，晚上睡下只能挨到早上眯一会，只是我每次见父亲他都是笑呵呵的很精神，看不出什么异样。生病前父亲每年也有体检，主要也就是有些肥胖的问题，医生着重提醒后，父亲其实也比较关注身体健康了，常常晚饭后就去老家小区旁边的公园登山锻炼，瘦下来了不少。特别是临近退休前一年，他瘦了三十多斤，我们还以为这是他锻炼的结果，父亲自己也觉得精神头不错，但这也许就是疾病开始的前兆。 得益于多年前随手为父亲购置了份医疗险，报销了绝大部分医疗费用，不至于为高昂的医疗费用发愁，还在疾病复发后使用了号称目前最为先进也非常昂贵的 CART 疗法，没为这病背上沉重的经济负担。这是我们的万幸，但这也让我放松了警惕，倘若是自己真金白银的药费，应当更驱动着我注重医疗知识的学习，更关注父亲的身体状态，可以更早发现父亲健康的危险状态。只是我太过于相信这些高昂费用的靶向药，CART 疗法的效果，想着有保险托底心里多了虚妄的自信，在发现 CART 疗法没有清除肿瘤后我还硬着头皮跟父亲说不怕，二线失败还有三线、四线的方案，别太焦虑。只是没想到多轮治疗后父亲的身体实则已千疮百孔，需要极其精细的照顾。病情急转直下的前两天，父亲出现了心率升高的情况，我们想着急诊的拥挤和漫长的等待，不如两天后的常规检查再好好问问主治医生，凭着经验没有及时拉着父亲去检查。又在入院后，症状本已逐渐好转，但我们照顾不周，不知是饮食还是什么原因，突然出现的肠穿孔让一切没了转机。 父亲最后动手术那一天的神情时常在我脑海里浮现，他紧锁着眉头，面容上浮现着焦虑与不安。在急诊留观的第三天，他出现了谵妄的症状，神志不清，总是觉得有人要害他。我第一次见到这样恐慌的父亲，心里非常难受。到手术的前几天，症状本有所好转，能认出人了，还能和我们聊天，只是聊的内容非常飘忽，说着好了要去开茶园，投资卖特色茶叶等等以前从没听过他讲过的事，不知是不是精神的恍惚让他回想起了年轻时的理想。多想等父亲出院再好好和他聊聊这些事啊，听听他这些思绪来源于他哪些想干而没干的事，能和他一起坐着正常的吹吹牛谈谈这些天马行空的想法。这一天父亲还是有些迷糊，但他肯定能从我们半夜来到医院，以及几次被医生喊出去单独谈话的行为上感觉到了什么。我们没法询问父亲自己真实的想法，还跟父亲说就是小手术，做了肚子就不痛了，父亲听了低头默默不语。现在回想起来，多想跟父亲好好说这个手术风险很大，您是不是害怕不想做呀，不想我们就回家好不好，至少再让我好好陪陪您。 手术之后父亲没有再醒来，各项指标跟状态越来越差，医生多次暗示我们救治下去的意义不大，医院 ICU 没有排到床位，父亲在普通病房躺着浑身插满各种管子，眼睑肿胀到只能半闭着眼睛，看着就非常难受。在经历了一次半夜的危险边缘的抢救后，母亲和我不忍心再让父亲受苦了，决定转运回老家的医院，我们在这里签署了放弃治疗协议。最后在父亲的床位前，老家的亲戚和我们一起见了父亲最后一面，我和母亲流着泪呼唤了父亲，父亲嘴角竟也抽动了下，这算是我们最后的告别，不久之后父亲便永远的离开了。 很多人说经历如此重大的事件后，人会变得更坚强更成熟，但我始终只感觉到的是虚无，无力和悔恨，以及对父亲一阵阵的想念… 还有太多话想和父亲说，但他就这样撒手离开了。他生这病后，偶有焦虑的神情，但在我们面前都尽量表现得平和，他去世后我在整理他手机时，看到了他有自己在好医生 APP 在线挂号以及和各种 AI 询问讨论病情的对话，言语间其实是很害怕焦虑的，这种恐慌他从未在我们面前表现出来，我每次在他面前也只是尽量表现乐观的让他有信心，别太焦虑，实在不知其实父亲他这么害怕，我那些无力的说辞到底有没有宽慰到他的心情。 父亲刚入院时，我和我的姑妈通电话，父亲小时候是她一直照顾着长大的，因而平常父亲和姑妈电话聊的最多，她说我爸这次感觉不好，前两天和她电话时说已经把遗书准备好了，一份给我一份给我妈。但是我当时还没想到情况这么严重，此后父亲精神变得不稳定，也没机会再询问，在他去世后，我才仔细的检查了他的遗物以及各个电子设备，却没有找到任何遗言，不知是不是这只是当时父亲的一个想法，只是还没来得及书写便症状恶化了。母亲之前也没想到这么快，她说父亲本来有几次想和她聊聊后事，但她感觉太沉重，父亲状态也还好，都打断了没让父亲继续说。我们无法再得知父亲还有什么遗愿，也是非常遗

Source: V2EX | Original Link

GitHub – NationalSecurityAgency/ghidra: Ghidra is a software reverse engineering (SRE) framework Skip to content You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session. You switched accounts on another tab or window. Reload to refresh your session. Dismiss alert NationalSecurityAgency / ghidra Public Notifications You must be signed in to change notification settings Fork 7.1k Star 64.4k Ghidra is a software reverse engineering (SRE) framework www.nsa.gov/ghidra License Apache-2.0 license 64.4k stars 7.1k forks Branches Tags Activity Star Notifications You must be signed in to change notification settings NationalSecurityAgency/ghidra master Branches Tags Go to file Code Open more actions menu Folders and files Name Name Last commit message Last commit date Latest commit History 16,581 Commits 16,581 Commits .github/ ISSUE_TEMPLATE .github/ ISSUE_TEMPLATE GPL GPL Ghidra Ghidra GhidraBuild GhidraBuild GhidraDocs GhidraDocs docker docker eclipse eclipse gradle gradle licenses licenses .gitattributes .gitattributes .gitignore .gitignore CONTRIBUTING.md CONTRIBUTING.md DISCLAIMER.md DISCLAIMER.md DevGuide.md DevGuide.md LICENSE LICENSE NOTICE NOTICE README.md README.md build.gradle build.gradle certification.local.manifest certification.local.manifest gradle.properties gradle.properties gradlew gradlew gradlew.bat gradlew.bat settings.gradle settings.gradle View all files Repository files navigation Ghidra Software Reverse Engineering Framework Ghidra is a software reverse engineering (SRE) framework created and maintained by the National Security Agency Research Directorate. This framework includes a suite of full-featured, high-end software analysis tools that enable users to analyze compiled code on a variety of platforms including Windows, macOS, and Linux. Capabilities include disassembly, assembly, decompilation, graphing, and scripting, along with hundreds of other features. Ghidra supports a wide variety of processor instruction sets and executable formats and can be run in both user-interactive and automated modes. Users may also develop their own Ghidra extension components and/or scripts using Java or Python. In support of NSA’s Cybersecurity mission, Ghidra was built to solve scaling and teaming problems on complex SRE efforts, and to provide a customizable and extensible SRE research platform. NSA has applied Ghidra SRE capabilities to a variety of problems that involve analyzing malicious code and generating deep insights for SRE analysts who seek a better understanding of potential vulnerabilities in networks and systems. If you are a U.S. citizen interested in projects like this, to develop Ghidra and other cybersecurity tools for NSA to help protect our nation and its allies, consider applying for a career with us . Security Warning WARNING: There are known security vulnerabilities within certain versions of Ghidra. Before proceeding, please re

Source: Hacker News | Original Link

Running My Own XMPP Server » Danny Notes from setting up Prosody in Docker for federated messaging, with file sharing, voice calls, and end-to-end encryption. About a year ago I moved my personal messaging to Signal as part of a broader push to take ownership of my digital life. That went well. Most of my contacts made the switch, and I’m now at roughly 95% Signal for day-to-day conversations. But Signal is still one company running one service. If they shut down tomorrow or change direction, I’m back to square one. XMPP fixes that. It’s federated, meaning your server talks to other XMPP servers automatically and you’re never locked into a single provider. Your messages live on your hardware. The protocol has been around since 1999 and it’s not going anywhere. I’d tried XMPP years ago and bounced off it, but the clients have come a long way since then. Monal and Conversations are genuinely nice to use now. This post covers everything I did to get a fully working XMPP server running with Prosody in Docker, from DNS records through to voice calls. Prerequisites A server with Docker and Docker Compose A domain you control TLS certificates (Let’s Encrypt works well) DNS records XMPP uses SRV records to let clients and other servers find yours. You’ll need these in your DNS: _xmpp-client._tcp.xmpp.example.com SRV 0 5 5222 xmpp.example.com. _xmpp-server._tcp.xmpp.example.com SRV 0 5 5269 xmpp.example.com. Port 5222 is for client connections, 5269 is for server-to-server federation. You’ll also want an A record pointing xmpp.example.com to your server’s IP. If you want HTTP file uploads (I’d recommend it), add a CNAME or A record for upload.xmpp.example.com pointing to the same server. Same for conference.xmpp.example.com if you want group chats with a clean subdomain, though Prosody handles this internally either way. TLS certificates Prosody won’t start without certificates. I use Let’s Encrypt with the Cloudflare DNS challenge so I don’t need to expose port 80: docker run –rm \ -v ~/docker/xmpp/certs:/etc/letsencrypt \ -v ~/docker/xmpp/cloudflare.ini:/etc/cloudflare.ini:ro \ certbot/dns-cloudflare certonly \ –dns-cloudflare \ –dns-cloudflare-credentials /etc/cloudflare.ini \ -d xmpp.example.com The cloudflare.ini file contains your API token: dns_cloudflare_api_token = your-cloudflare-api-token After certbot runs, fix the permissions so Prosody can read the certs: chmod -R 755 ~/docker/xmpp/certs/live/ ~/docker/xmpp/certs/archive/ chmod 644 ~/docker/xmpp/certs/archive/xmpp.example.com/*.pem Set up a cron to renew monthly: 0 3 1 * * docker run –rm -v ~/docker/xmpp/certs:/etc/letsencrypt \ -v ~/docker/xmpp/cloudflare.ini:/etc/cloudflare.ini:ro \ certbot/dns-cloudflare renew \ –dns-cloudflare-credentials /etc/cloudflare.ini \ && docker restart xmpp The Docker setup The docker-compose.yml : services : prosody : image : prosodyim/prosody: 13.0 container_name : xmpp restart : unless-stopped ports : – “5222:5222” – “5269:5269” volumes : – prosody-data:

Source: Hacker News | Original Link

Ministry of Justice orders deletion of the UK’s largest court reporting database – Legal Cheek Posts News Ministry of Justice orders deletion of the UK’s largest court reporting database By Legal Cheek on Feb 11 2026 11:30am 3 Blow for open justice A digital archive that helped journalists track criminal court cases is being shut down by the Ministry of Justice. Courtsdesk will reportedly be deleted within days after HM Courts & Tribunals Service ordered every record wiped. The platform had been used by more than 1,500 reporters from 39 media outlets to search magistrates’ court lists and registers, but the move has triggered warnings that important cases could now go unreported. Courtsdesk says it repeatedly found the media wasn’t being told about hearings, with two-thirds of courts regularly hearing cases without notifying journalists. The platform was launched in 2020 following an agreement with HMCTS and approval by the Lord Chancellor and former Justice Minister Chris Philp, but HMCTS issued a cessation notice in November citing “unauthorised sharing” of court information. The 2026 Legal Cheek Firms Most List Courtsdesk founder Enda Leahy said the company wrote to government agencies 16 times trying to save the service. It asked for the matter to be referred to the Information Commissioner’s Office but says that request went nowhere, and former Philp himself approached current courts minister Sarah Sackman asking for the archive not to be deleted. The government refused last week. Leahy told The Times that HMCTS couldn’t do what Courtsdesk did. She pointed to figures showing the court service’s own records were accurate just 4.2% of the time and that 1.6 million criminal hearings went ahead without any advance notice to the press. “We built the only system that could tell journalists what was actually happening in the criminal courts,” she said. An HMCTS spokesperson said the press would continue to have full access to court information to support accurate reporting. HMCTS acted to protect sensitive data after CourtsDesk sent information to a third-party AI company. Journalists’ access to court information has not been affected: listings and records remain available. pic.twitter.com/4KWlpCcaAq — Ministry of Justice (@MoJGovUK) February 10, 2026 Barristers Courts Solicitors 3 Comments Paul grigg Feb 11 2026 4:59pm What a crock of shit who,s idea is this Reply Report comment ( 3 ) ( 0 ) Simply Grim Feb 11 2026 5:25pm Dark days ahead. I think this Labour government has been far worse than anyone could ever have imagined, at times it feels like they are trying to get forced to resign. Abolition of jury trials, media blackouts, AI facial recognition, all while cracking down on protests and online activity. Mahmood quite literally said she aspires to the theory of the panopticon, in which prisoners are forced to behave due to fear of constant surveillance. That is how they view citizens, as prisoners. Reply Report comment ( 12 ) ( 0 ) Seems fine b

Source: Hacker News | Original Link

Thanks a lot, AI: Hard drives are already sold out for the entire year, says Western Digital | Mashable Western Digital says its all sold out of hard drives for 2026, less than two months into the year. Credit: Alex Kraus/Bloomberg via Getty Images Looking to buy a new hard drive? Get ready to pay even more this year. According to Western Digital, one of the world’s biggest hard drive manufacturers, the company has already sold out of its storage capacity for 2026 with more than 10 months still left in the year. “We’re pretty much sold out for calendar 2026,” said Western Digital CEO Irving Tan on the company’s recent quarterly earnings call . You May Also Like Tan shared that most of the storage space has been allocated to its “top seven customers.” Three of these companies already have agreements with Western Digital for 2027 and even 2028. Mashable Light Speed Want more out-of-this world tech, space and science stories? Sign up for Mashable’s weekly Light Speed newsletter. Loading… Sign Me Up Use this instead By clicking Sign Me Up, you confirm you are 16+ and agree to our Terms of Use and Privacy Policy . Thanks for signing up! SEE ALSO: This is your last chance to get super cheap SSDs and hard drives at Amazon Furthermore, the incentive for these hardware companies to prioritize the average consumer is also dwindling. According to Western Digital, thanks to a surge in demand from its enterprise customers, the consumer market now accounts for just 5 percent of the company’s revenue. AI companies have been eating up computer hardware as industry growth accelerates. Prices for products ranging from computer processors to video game consoles have skyrocketed due to these AI companies cannibalizing supply chains. The tech industry has already been experiencing a shortage of memory due to demand from AI companies. PC makers have been forced to raise RAM prices on a near-regular basis as shortages persist. Video game console makers, like Sony, have even reportedly considered pushing the next PlayStation launch beyond the planned 2027 release in hopes that AI-related hardware shortages would be resolved by then. With this latest news from Western Digital, it appears the ever-increasing demands from AI companies for memory and storage will continue to grow, with no end in sight. Unless, of course, investors decide to pull back from AI over fears that AI’s promises may not come to fruition. But, for now at least, the shortages – and price hikes for consumers – will continue. Topics Artificial Intelligence Recommended For You How to watch Western Michigan vs. Kennesaw State online for free Live stream all the action from the Myrtle Beach Bowl from anywhere in the world. 12/19/2025 By Matt Ford Apple’s iPhone Pocket sock is fully sold out The people, they yearn for the sock! 11/26/2025 By Stan Schroeder Creator Khaby Lame just sold a stake in his brand for $975 million A mouth-watering amount of money. 01/28/2026 By Chance Townsend How ‘A Knight of th

Source: Hacker News | Original Link

Qwen

Source: Hacker News | Original Link

GitHub – alibaba/zvec: A lightweight, lightning-fast, in-process vector database Skip to content You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session. You switched accounts on another tab or window. Reload to refresh your session. Dismiss alert alibaba / zvec Public Notifications You must be signed in to change notification settings Fork 139 Star 2.8k A lightweight, lightning-fast, in-process vector database zvec.org/en/ License Apache-2.0 license 2.8k stars 139 forks Branches Tags Activity Star Notifications You must be signed in to change notification settings alibaba/zvec main Branches Tags Go to file Code Open more actions menu Folders and files Name Name Last commit message Last commit date Latest commit History 61 Commits 61 Commits .github .github cmake cmake examples/ c++ examples/ c++ python python scripts scripts src src tests tests thirdparty thirdparty tools tools .clang-format .clang-format .gitignore .gitignore .gitmodules .gitmodules .pre-commit-config.yaml .pre-commit-config.yaml CMakeLists.txt CMakeLists.txt CODE_OF_CONDUCT.md CODE_OF_CONDUCT.md CONTRIBUTING.md CONTRIBUTING.md LICENSE LICENSE README.md README.md pyproject.toml pyproject.toml View all files Repository files navigation 🚀 Quickstart | 🏠 Home | 📚 Docs | 📊 Benchmarks | 🎮 Discord | 🐦 X (Twitter) Zvec is an open-source, in-process vector database — lightweight, lightning-fast, and designed to embed directly into applications. Built on Proxima (Alibaba’s battle-tested vector search engine), it delivers production-grade, low-latency, scalable similarity search with minimal setup. 💫 Features Blazing Fast : Searches billions of vectors in milliseconds. Simple, Just Works : Install and start searching in seconds. No servers, no config, no fuss. Dense + Sparse Vectors : Work with both dense and sparse embeddings, with native support for multi-vector queries in a single call. Hybrid Search : Combine semantic similarity with structured filters for precise results. Runs Anywhere : As an in-process library, Zvec runs wherever your code runs — notebooks, servers, CLI tools, or even edge devices. 📦 Installation Python Requirements : Python 3.10 – 3.12 pip install zvec Node.js npm install @zvec/zvec ✅ Supported Platforms Linux (x86_64, ARM64) macOS (ARM64) 🛠️ Building from Source If you prefer to build Zvec from source, please check the Building from Source guide. ⚡ One-Minute Example import zvec # Define collection schema schema = zvec . CollectionSchema ( name = “example” , vectors = zvec . VectorSchema ( “embedding” , zvec . DataType . VECTOR_FP32 , 4 ), ) # Create collection collection = zvec . create_and_open ( path = “./zvec_example” , schema = schema ) # Insert documents collection . insert ([ zvec . Doc ( id = “doc_1” , vectors = { “embedding” : [ 0.1 , 0.2 , 0.3 , 0.4 ]}), zvec . Doc ( id = “doc_2” , vectors = { “embedding” : [ 0.2 , 0.3 , 0.4 , 0.1 ]}), ]) # Search by vector simila

Source: GitHub Trending | Original Link

GitHub – preservim/vim-pencil: Rethinking Vim as a tool for writing Skip to content You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session. You switched accounts on another tab or window. Reload to refresh your session. Dismiss alert preservim / vim-pencil Public Notifications You must be signed in to change notification settings Fork 39 Star 1.7k Rethinking Vim as a tool for writing License View license 1.7k stars 39 forks Branches Tags Activity Star Notifications You must be signed in to change notification settings preservim/vim-pencil master Branches Tags Go to file Code Open more actions menu Folders and files Name Name Last commit message Last commit date Latest commit History 223 Commits 223 Commits .github/ workflows .github/ workflows autoload autoload doc doc plugin plugin screenshots screenshots .vintrc.yaml .vintrc.yaml LICENSE LICENSE README.markdown README.markdown View all files Repository files navigation vim-pencil Rethinking Vim as a tool for writers Features The pencil plugin aspires to make Vim as powerful a tool for writers as it is for coders by focusing narrowly on the handful of tweaks needed to smooth the path to writing prose. For editing prose-oriented file types such as text , markdown , mail , rst , tex , textile , and asciidoc Agnostic on soft line wrap versus hard line breaks, supporting both Auto-detects wrap mode via modeline and sampling Adjusts navigation key mappings to suit the wrap mode Creates undo points on common punctuation during Insert mode, including deletion via line and word Buffer-scoped configuration (with a few minor exceptions, pencil preserves your global settings) Support for Vim’s Conceal feature to hide markup defined by Syntax plugins (e.g., _ and * markup for styled text in _ Markdown _) Support for display of mode indicator ( ␍ and ⤸ , e.g.) in the status line Pure Vimscript with no dependencies In addition, when using hard line break mode: Makes use of Vim’s powerful autoformat while inserting text, except for tables and code blocks where you won’t want it. NEW Optional key mapping to suspend autoformat for the Insert. Need spell-check, distraction-free editing, and other features? Vim is about customization. To complete your editing environment, learn to configure Vim and draw upon its rich ecosystem of plugins. Why use Vim for writing? With plenty of word processing applications available, including those that specifically cater to writers, why use a modal editor like Vim? Several reasons have been offered: Your hands can rest in a neutral ‘home’ position, only rarely straying to reach for mouse, track pad, or arrow keys Minimal chording, with many mnemonic-friendly commands Sophisticated capabilities for navigating and manipulating text Highly configurable, enabling you to build a workflow that suits your needs, with many great plugins available No proprietary format lock-in But while

Source: Hacker News | Original Link

Mac 上 有 Navicat 的替代品吗？ – V2EX 首页 注册 登录 V2EX = way to explore V2EX 是一个关于分享和探索的地方 现在注册 已注册用户请 登录 • 请不要在回答技术问题时复制粘贴 AI 生成的内容 V2EX › 程序员 Mac 上 有 Navicat 的替代品吗？ ggp1ot2 · 20 小时 27 分钟前 · 2110 次点击 安装的破解版，公司电脑不给装 说是只要公司电脑安装就算商用，没有购买授权，有侵权风险 一安装，安全部门的人就监控到然后要求删除 想问下，有没有类似 Navicat 的软件，用起来感觉挺顺手的，可以管理多种数据库 Navicat 软件 数据库 32 条回复 • 2026-02-16 18:41:39 +08:00 1 lyonbot 20 小时 22 分钟前 via Android dbeaver 和 dbgate 都凑合，但是易用性美观度差点 2 hronro 20 小时 20 分钟前 Navicat Lite 3 humbass 20 小时 18 分钟前 免费：Sequel Pro 4 xiaoz 20 小时 17 分钟前 via Android 免费的话 dbeare 支持的数据库挺全的，软件功能也很完善，就是界面有点 low 。 愿意付费的话国产有个 Hexhub 还不错: https://blog.xiaoz.org/archives/22439 ，目前在用。 5 cloverzrg2 20 小时 12 分钟前 @ hronro #2 Navicat Lite 不能商用的，只能个人用 6 levelworm 20 小时 11 分钟前 让公司购买，不然就降低效率。 7 Goalonez 20 小时 10 分钟前 1 datagrip ，写 sql 更舒服一点。 不过 UI 操作感觉差点意思。用久了可能也就习惯都拿 sql 操作了。 8 leonardo22 20 小时 1 分钟前 datagrip 挺好用的 9 wangYQ 19 小时 26 分钟前 dbeaver 只要不搞国臣数据库一般没啥奇怪的问题，还挺好用的 。国产库就算了 10 mauve PRO 17 小时 55 分钟前 1 Tableplus 11 deamonet 17 小时 11 分钟前 table plus 荔枝商店可以自己买，，我就买了虽然也没啥用 12 ysc3839 17 小时 5 分钟前 via Android HeidiSQL 有出 macOS arm64 版本，可以试试 https://www.heidisql.com/download.php 13 gadfly3173 16 小时 13 分钟前 达梦出了个 sqlark 我感觉也还可以 14 itechify PRO 10 小时 47 分钟前 DBeaver 15 netizen 10 小时 26 分钟前 via iPhone @ cloverzrg2 官网原话：Navicat Premium Lite 可用于商业和非商业目的，使其成为从小型企业到个体开发者等用户的实用选择。每家机构最多可使用 5 个 Navicat Premium Lite 免费用户账户。 16 SayHelloHi 10 小时 16 分钟前 试一试这个 UI 也不错 https://github.com/beekeeper-studio/beekeeper-studio/ 17 idragonet 9 小时 36 分钟前 @ leonardo22 #8 商用需要买许可 18 1103409364 9 小时 20 分钟前 HeidiSQL 19 mootee 9 小时 8 分钟前 via iPhone 现在 IDEA 的 datagrip 对非商业用途免费的了，很棒，可以查看和尝试 https://www.jetbrains.com/zh-cn/datagrip/ 20 fstab 8 小时 41 分钟前 @ xiaoz #4 Hexhub 目前唯一的缺点是：表格导入数据库不能自动创建表。 我推荐： https://jookdb.com/ ， 唯一的缺点就是免费版，只支持 MySQL, Oracle and SQLServer ，其他都不支持，MariaDB 就很难受了。 21 xj3hd 8 小时 33 分钟前 现在用的 TablePlus ，暂时感觉很满意 22 liuxue 7 小时 56 分钟前 @ wangYQ 链接达梦数据库，如果不做数据迁移的话也没什么问题 23 yjxjn 6 小时 35 分钟前 HeidiSQL win 和 mac 都用，很满意。 24 ggp1ot2 OP 6 小时 6 分钟前 @ netizen 没有用的，注意：部分软件可供个人使用，但不允许商业使用，只要在公司提供的电脑上下载该软件，即被视为商业使用，需要购买商业版权，请在下载前进行甄别 25 HappyAndSmile 5 小时 32 分钟前 Sequel Ace 26 Dragonish3600 5 小时 10 分钟前 via iPhone data grid 啊 27 giter 4 小时 19 分钟前 @ ggp1ot2 https://www.navicat.com.cn/support/help/navicat-premium-lite-%E5%8F%AF%E4%BB%A5%E7%94%A8%E4%BA%8E%E5%95%86%E4%B8%9A%E7%94%A8%E9%80%94%E5%90%97 我不知道我理解的商业目的和你们理解的商业目的是否一致，至少在我所看到的官方介绍里，都是提到了可以商业使用的。 28 Lockroach 3 小时 34 分钟前 推荐 DataGrip 的审题啊😂op 是公司没给买授权所以要用不需要商业授权的 dbm 29 Lemonadeccc 3 小时 4 分钟前 vscode 插件 30 Ice188 1 小时 50 分钟前 https://github.com/yhan219/navicat_reset_mac 用这个脚本吧，无限授权 31 cloverzrg2 18 分钟前 @ netizen #15 我们公司，研发/QA/PM 五六千人，统一不能安装 navicat 的产品。 电脑上检测出来安装了，就有人联系删除软件 32 hotea 6 分钟前 sequel ace 关于 · 帮助文档 · 自助推广系统 · 博客 · API · FAQ · Solana · 1769 人在线 最高记录 6679 · Select Language 创意工作者们的社区 World is powered by solitude VERSION: 3.9.8.5 · 26ms · UTC 10:47 · PVG 18:47 · LAX 02:47 · JFK 05:47 ♥ Do have faith in what you’re doing. ❯

Source: V2EX | Original Link

GitHub – unicode-org/message-format-wg: Developing a standard for localizable message strings Skip to content You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session. You switched accounts on another tab or window. Reload to refresh your session. Dismiss alert unicode-org / message-format-wg Public Uh oh! There was an error while loading. Please reload this page . Notifications You must be signed in to change notification settings Fork 36 Star 288 Developing a standard for localizable message strings unicode-org.github.io/message-format-wg/ License View license 288 stars 36 forks Branches Tags Activity Star Notifications You must be signed in to change notification settings unicode-org/message-format-wg main Branches Tags Go to file Code Open more actions menu Folders and files Name Name Last commit message Last commit date Latest commit History 842 Commits 842 Commits .github .github docs docs exploration exploration meetings meetings publish publish spec spec test test .gitignore .gitignore .gitmodules .gitmodules CODEOWNERS CODEOWNERS CONTRIBUTING.md CONTRIBUTING.md LICENSE LICENSE README.md README.md delegates.md delegates.md View all files Repository files navigation MessageFormat Working Group Welcome to the home page for the MessageFormat Working Group, a subgroup of the Unicode CLDR-TC . Charter The MessageFormat Working Group (MFWG) is tasked with developing and supporting an industry standard for the representation of localizable message strings. MessageFormat is designed to support software developers, translators, and end users with fluent messages and locally-adapted presentation for data values while providing a framework for increasingly complex features, such as gender, inflections, and speech. Our goal is to provide an interoperable syntax, message data model, and associated processing that is capable of being adopted by any presentation framework or programming environement. The Unicode MessageFormat Standard The Unicode MessageFormat Standard is a stable part of CLDR. It was approved by the CLDR Technical Committee and is recommended for implementation and adoption. The normative version of the specification is published as a part of TR35 . This repository contains the editor’s copy. Unicode MessageFormat is sometimes referred to as MessageFormat 2.0 , since it replaces earlier message formatting capabilities built into ICU. Some default functions and items in the u: namespace are still in Draft status. Feedback from users and implementers might result in changes to these capabilities. The MessageFormat Working Group and CLDR Technical Committee welcome any and all feedback, including bugs reports, implementation reports, success stories, feature requests, requests for clarification, or anything that would be helpful in supporting or enhancing the specification and promoting widespread adoption. Sharing Feedback Do you have feedback on the

Source: Hacker News | Original Link

Sina Visitor System

Source: Weibo Hot Search | Original Link

Expensively Quadratic: the LLM Agent Cost Curve – exe.dev blog Pop quiz: at what point in the context length of a coding agent are cached reads costing you half of the next API call? By 50,000 tokens, your conversation’s costs are probably being dominated by cache reads. Let’s take a step back. We’ve previously written about how coding agents work: they post the conversation thus far to the LLM, and continue doing that in a loop as long as the LLM is requesting tool calls. When there are no more tools to run, the loop waits for user input, and the whole cycle starts over. Visually: The agentic loop Or, in code form: def loop(llm): msg = user_input() while True: output, tool_calls = llm(msg) print(“Agent: “, output) if tool_calls: msg = [handle_tool_call(tc) for tc in tool_calls] else: msg = user_input() The LLM providers charge for input tokens, cache writes, output tokens, and cache reads. It’s a little tricky: you indicate in your prompt to cache up to a certain point (usually the end), and you get charged as “cache write” and not input. The previous turn’s output becomes the next turn’s cache write. Visually: Token costs across LLM calls Here, the colors and numbers indicate the costs making up the n th call to the LLM. Every subsequent call reads the story so far from the cache, writes the previous call’s output to the cache (as well as any new input), and gets an output. The area represents the cost, though in this diagram, it’s not quite drawn to scale. Add up all the rectangles, and that’s the total cost. That triangle emerging for cache reads? That’s the scary quadratic! How scary is the quadratic? Pretty squarey! I took a rather ho-hum feature implementation conversation, and visualized it like the diagram above. The area corresponds to cost: the width of every rectangle is the number of tokens and the height is the cost per token. As the conversation evolves, more and more of the cost is the long thin lines across the bottom that correspond to cache reads. The whole conversation cost $12.93 total or so. You can see that as the conversation continues, the cache reads dominate. At the end of the conversation, cache reads are 87% of the total cost. They were half the cost at 27,500 tokens! This conversation is just one example. Does this happen generally? exe.dev’s LLM gateway keeps track of the costs we’re incurring. We do not store the messages themselves going past, but we do keep track of the number of tokens. The following graph shows the “cumulative cost” visualization for many Shelley conversations, not just my own. I sampled 250 conversations from the data randomly. The x-axis is the context length, and the y-axis is the cumulative cost up to that point. The left graph is all the costs and the right graph is just the cache reads. You can mouse over to find a given conversation on both graphs. The box plots below show the distribution of input tokens and output tokens. The cost curves are all different because every conversation is

Source: Hacker News | Original Link

使用公司提供的 AI 套餐来做个人项目，会有所有权的纠纷么？ – V2EX 首页 注册 登录 V2EX = way to explore V2EX 是一个关于分享和探索的地方 现在注册 已注册用户请 登录 • 请不要在回答技术问题时复制粘贴 AI 生成的内容 V2EX › 程序员 使用公司提供的 AI 套餐来做个人项目，会有所有权的纠纷么？ x97bgt · 19 小时 45 分钟前 · 2284 次点击 公司提供了 codex 、Claude 、Gemini 。Claude code 是中转的，codex 和 Gemini 都是走官方的 API 。 额度很多，平时工作用不完。如果我用来搞个人项目（ GitHub 开源），会有啥风险不？ 公司 项目 风险 39 条回复 • 2026-02-16 14:12:23 +08:00 1 TossPig 19 小时 36 分钟前 4 你非要在公司炫耀你用公司资源做了个人项目吗？ 2 Aliencn 19 小时 22 分钟前 法律上来说，属于职务成果，归公司所有。 实际上来说，看公司的包容度 3 darkengine 19 小时 14 分钟前 9 “有些事不上称没四两重，上称了一千斤也打不住” 4 defaw 19 小时 4 分钟前 你如何保证公司给你的 api 没经过公司自己的截流和二次中转 5 sn0wdr1am 18 小时 55 分钟前 理论上是违规的。 利用公司资源，任何产出归公司。 6 ktyang 18 小时 51 分钟前 说公司没有管控我是不信的 7 kneo 18 小时 30 分钟前 1. 会不会公司因为你违规使用公司 AI 干私活产生额外支出？ 2. 你确定额度用不完？会不会因为你拿来干私活，导致其他同事额度不够？ 先想清楚这两个问题。 8 FlashEcho 18 小时 8 分钟前 建议你别这么搞，openai 的账号后台是记录了全量 logs 的，input 也记，时间也记，你现在搞了等于给自己埋雷，就算现在不查将来搞个审计咋办？ 9 huluhulu 17 小时 52 分钟前 就这点东西为啥还要薅呢？ 没必要吧？ 公司的东西风险太大。 10 009694 17 小时 11 分钟前 via iPhone 暂时不管你只是因为觉得你还有价值，等要裁员的时候公司就能免费裁掉你，狠一点的话还能用职务侵占罪告你 11 x97bgt OP 17 小时 5 分钟前 @ FlashEcho openai 后台记录了 logs 么？这个哪里可以访问到？之前我记得只能看到调用的历史，看不到具体内容。 12 JoshTheLegend 16 小时 49 分钟前 你没赚到钱，公司估计也懒得管你，但是没赚钱=白折腾。 你赚到了钱，公司再开始管也不算晚，那你还是白折腾，还白送人家一个免费开除你的把柄。 13 noobjalen 16 小时 28 分钟前 没必要 ，用用中转站开发自己的也够用。 14 bbbblue 16 小时 23 分钟前 你在上班时间在公司搞应该问题不大（哪里不太对 15 bbbblue 16 小时 23 分钟前 我说的问题不大是指不会被公司发现 😂 有点歧义了 16 zazzaz 16 小时 18 分钟前 看一下公司开的是不是企业账户，那个 API 后台有审计功能的 17 OneLiteCore 15 小时 49 分钟前 理论上违法或者违规，但是公司如果管的不严或者不在乎的话那就问题不大。只不过话又说回来了，如果公司给的薪资水平连给你买 AI 服务都负担不起的话，也是时候跳槽了。 18 docx 15 小时 37 分钟前 via iPhone 万一被公司发现了怎么办 19 youshangdemajia 13 小时 39 分钟前 via iPhone 理论上任何时间用公司电脑开发的东西都归公司所有。更何况这些公司买的服务。 20 chunhuitrue 13 小时 9 分钟前 如果不是实在买不起 ai ，没必要冒险。 21 oxykr 11 小时 8 分钟前 via Android 自愿加班算工伤吗？ 22 SayHelloHi 8 小时 17 分钟前 你不说 就没有人知道 23 streamrx 8 小时 15 分钟前 via iPhone 企业账户可以查的.. 24 flyqie 8 小时 13 分钟前 via Android @ darkengine #3 对的，这种事情悄默默干就得了 25 FlashEcho 8 小时 7 分钟前 @ x97bgt #11 RT https://imgur.com/H9VNMJb 26 FlashEcho 8 小时 6 分钟前 2 @ x97bgt #11 27 EricYuan1 7 小时 46 分钟前 会被开除，赔偿都没有 28 yalin 6 小时 48 分钟前 不要这样干 29 Evovil 6 小时 45 分钟前 via iPhone 这点钱也要省？ 风险和收益完全不成正比。贪小便宜吃大亏。 所有对企业提供的软件和服务都会有审计功能，这是拍脑袋都知道的事。除非是 im 私聊这种偏灰色的（足够大体量也可能会记录） 30 iorilu 6 小时 32 分钟前 这还要问, 不是套餐的事 你在公司所有事件都属于公司, 无论你干任何事情, 产权都属于公司, 和你用啥 api 没关系 你在公司上班时间, 用自己买的 api 写的代码肯定也属于公司 31 lneoi 6 小时 11 分钟前 用上班时间做的任何事情都归公司，只是一般情况下没那么严格。也就是这样，工作受伤、上下班路上受伤也算工伤。 32 ShaoYuanNuo 5 小时 46 分钟前 查出来 肯定有风险的，用了公司的东西，这个不属于你个人的了 33 psirnull 4 小时 18 分钟前 职务侵占 34 BadAngel 3 小时 26 分钟前 参考菊厂，在职所有产出都是公司的 35 x97bgt OP 3 小时 25 分钟前 @ FlashEcho #26 多谢老哥。这个界面的链接是哪个？ 36 HappyAndSmile 3 小时 22 分钟前 如果这点钱都要省，做的东西也就这么点出息了，另外，我领导偶尔也会抽查看看大家在用 AI 做什么 37 zxjxzj9 2 小时 32 分钟前 简单来说看你和其他人关系怎么样… 简单来说是没啥事情的，但是真的要来搞你那就不好说咯。 38 bybyte 2 小时 28 分钟前 搞不明白为啥这种羊毛都薅。。 39 HENQIGUAI 2 小时 22 分钟前 三楼已经给出最确切的答案了，事实上，在职期间的做的网站、产品、项目等要是赚大钱出名了，你看公司找不找你就完事了 关于 · 帮助文档 · 自助推广系统 · 博客 · API · FAQ · Solana · 1823 人在线 最高记录 6679 · Select Language 创意工作者们的社区 World is powered by solitude VERSION: 3.9.8.5 · 23ms · UTC 08:35 · PVG 16:35 · LAX 00:35 · JFK 03:35 ♥ Do have faith in w

Source: V2EX | Original Link

夸克网盘是真流氓 – V2EX 首页 注册 登录 V2EX = way to explore V2EX 是一个关于分享和探索的地方 现在注册 已注册用户请 登录 V2EX › 问与答 夸克网盘是真流氓 jibin · 23 小时 47 分钟前 · 2416 次点击 大家有没有留意到，最近用夸克网盘，只要你一打开网盘，就会被后台篡改默认播放软件，变成夸克自己的播放器，无论你修改多少次都没用，这个流氓夸克真没人管得了吗？ 这比百度网盘还要恶心一万倍。 夸克 网盘 播放器 46 条回复 • 2026-02-16 16:11:16 +08:00 1 MYDB 23 小时 42 分钟前 via iPhone 喜欢搞全家桶的软件默认丢虚拟机 2 TimG 22 小时 56 分钟前 via Android 之前朋友来家里玩，想一起用 htpc 放个电影，但是我没有资源，朋友的夸克网盘有这个电影，我们就下载了一个夸克，然后又下载了夸克网盘，几天过去我想起来开机卸载，结果找不到夸克网盘的卸载方法，软件列表里没有，安装目录中也找不到名为 uninstall 的可执行文件，实在没办法又下载了一遍才出现正常的卸载方法。不知道是 bug 还是怎样，想起之前曝光的细如病毒防探针规则的鲁大师广告事件，未免有些阴谋论，感觉背后发凉。 3 opengps 22 小时 30 分钟前 感觉网盘这个圈子里，没有不流氓的玩法（文明的早挂了 ） 4 miaomiao888 22 小时 23 分钟前 这种用时才启动的工具建议放 Sandboxie 里运行 5 tsx8 22 小时 5 分钟前 要么不用夸克，要么用下载助手在网页端直接转 BC 链接下载，速度也很快 6 w2er 21 小时 14 分钟前 国内网盘都在虚拟环境了，不是我不信任，是不让我信任 7 szzzzz 21 小时 9 分钟前 via Android 老早知道夸克不是个好东西，所以到现在都不用，也不下载它的东西。 8 LandCruiser 20 小时 21 分钟前 我还真没遇到这个问题，我播放器是 potplayer 9 jibin OP 19 小时 54 分钟前 @ LandCruiser 我也是这个播放器，最新版夸克网盘，只要一打开，马上就把默认播放器更改了，每一次都这样。。。心累 10 jibin OP 19 小时 54 分钟前 @ MYDB 嗯嗯，只能这样了 11 jibin OP 19 小时 54 分钟前 @ TimG 哎，可怕 12 jibin OP 19 小时 54 分钟前 @ opengps 无解啊 13 jibin OP 19 小时 54 分钟前 @ miaomiao888 嗯嗯 14 jibin OP 19 小时 53 分钟前 @ tsx8 好的，我试试 15 jibin OP 19 小时 53 分钟前 @ w2er 确实无法信任，亏的我还开了一年会员 16 jibin OP 19 小时 53 分钟前 @ szzzzz 要不是开了会员，我就卸载了 17 fstar 19 小时 35 分钟前 MacOS 好像不会 18 cxj96 19 小时 14 分钟前 保存一个视频，然后在线播放，在视频播放器里面高级设置里可以修改 19 Gilfoyle26 19 小时 8 分钟前 要说做流氓软件，这些互联网大厂颇有心得，放眼全球，无出其右！ 20 someonesnone 18 小时 47 分钟前 我一直只装了 3.x 版本没升级, 把 4.0 的升级目录创建好然后删除权限他就无法升级了 21 Actrace 18 小时 25 分钟前 所以说，为啥不用钛盘？纯浏览器多省心。 22 kiii 18 小时 19 分钟前 不懂为什么那么多人用什么夸克网盘，是能赚钱吗？ 23 longzhou6431 17 小时 49 分钟前 @ kiii 因为 88 会员，导致使用它不要钱。另外，这个网盘，大概晚上十点以后，可以用来下载一些小黄片，速度还嘎嘎快。 24 kiii 17 小时 47 分钟前 @ longzhou6431 淘宝 88 会员，免费用？还能下黄片？那确实可以研究一下 25 V17 16 小时 40 分钟前 @ LandCruiser 同样，从来没有，不小心点到设置问题？ 26 Nemolon 7 小时 38 分钟前 via iPhone 在它更新变浏览器时候已经丢 sandboxie 了，当时就觉得开这个头肯定还会有更流氓的更新 27 wniming 6 小时 51 分钟前 我没遇到这个问题，夸克用了大半年了，默认播放器一直都是 mpv 28 iorilu 5 小时 59 分钟前 @ wniming 夸克新版搞成全家桶了, 一用必中招 29 wniming 5 小时 52 分钟前 @ iorilu 我前几天才安装的，不是全家桶，你在哪个网址下载的？我是在这里： https://pan.quark.cn/ 下载后的大小 124MB ，这个大小应该塞不下全家桶。 30 iorilu 5 小时 31 分钟前 https://www.v2ex.com/t/1191745 @ wniming 我上次还发了个帖子 那些都是后装的, 会装浏览器, 修该默认文件打开程序等(图片, pdf, 文档等), 企图接管你电脑所有东西 31 lnbiuc 5 小时 24 分钟前 非用夸克不可吗 32 loginv2 5 小时 17 分钟前 百度 夸克 天翼 迅雷 这些都扔虚拟机了，通过文件网络共享设置一下下载目录 直接下到本机 33 lysShub 5 小时 7 分钟前 都是扔沙盒的 34 ww2000e 1 小时 28 分钟前 流氓治流氓 我用 360 确保不篡改。。 35 chocolatesir 1 小时 6 分钟前 好像还会默认界面置顶，显示在最前面 36 jibin OP 26 分钟前 @ fstar 我是 win 平台，没用过 Mac 的 37 jibin OP 26 分钟前 @ cxj96 好的，我去试试，谢谢 38 jibin OP 25 分钟前 @ Gilfoyle26 哎，这也是国情 39 jibin OP 25 分钟前 @ someonesnone 后面变成强制升级了啊，只要你一运行软件就会自动升级，控制不住吧 40 jibin OP 24 分钟前 @ Actrace 好多电影博主只在夸克分享电影，也是没办法。。 41 jibin OP 24 分钟前 @ kiii 微博上好多电影博主只在夸克跟百度分享电影，别的网盘没用处啊。。。 42 jibin OP 23 分钟前 @ wniming 那你运气好 43 jibin OP 22 分钟前 @ lnbiuc 微博上关注的几个分享电影的博主只在夸克分享，没办法呀 44 jibin OP 22 分钟前 @ loginv2 好的，我试试 45 jibin OP 21 分钟前 @ ww2000e 就行不想再用电脑卫视了，感觉火绒也不干净。。 46 jibin OP 21 分钟前 @ chocolatesir 这个我没注意到。。 关于 · 帮助文档 · 自助推广系统 · 博客 · API · FAQ · Solana ·

Source: V2EX | Original Link

GitHub – antirez/picol: A Tcl interpreter in 500 lines of code Skip to content You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session. You switched accounts on another tab or window. Reload to refresh your session. Dismiss alert antirez / picol Public Notifications You must be signed in to change notification settings Fork 11 Star 127 A Tcl interpreter in 500 lines of code 127 stars 11 forks Branches Tags Activity Star Notifications You must be signed in to change notification settings antirez/picol main Branches Tags Go to file Code Open more actions menu Folders and files Name Name Last commit message Last commit date Latest commit History 3 Commits 3 Commits README.md README.md picol.c picol.c View all files Repository files navigation Picol is a Tcl-alike interpreter in 500 lines of code that I released 15th of March 2007. Recentely I looked at the source code and realized this was a better C programming example compared to what I recalled, so I’m putting this on GitHub to archive it, together with the main points of the original article. Rules When I built this code, I had some rule in mind: I wanted to use more or less my normal C style. In Picol you’ll find normal C spacing and even comments. I wanted to write an interpreter with a design similar to a real one. One of the few useful things you can do with Picol is to learn how to write a Tcl interpreter if you are a newbie programmer, I guess, so the point was to write a simple to understand program, not just a short program. The resulting interpreter should be able to run some kind of non trivial program: to just set few vars and print hello world was not an option. The resulting interpreter: Picol The parser is very similar to the Tcl one, Picol supports interpolation as well, for example you can write: set a ” pu ” set b {ts} $a$b ” Hello World! ” Note that Picol has an interactive shell! so just launch it without arguments to start to play (to compile the code use gcc -O2 -Wall -o picol picol.c ). To run a program stored in a file, use: picol filename.tcl . Probably the parser could be rewritten in order to take less space, currently it takes almost 250 lines of code: this is too much and leaves little room for all the rest. On the other side, it’s a decent example about writing parsers by hand. A Raw list of the supported features: Interpolation, as seen above. You can also write “2+2 = [+ 2 2]” or “My name is: $foobar” . Procedures, with return. Like Tcl if return is missing the result of the last command executed is returned. If , if .. else .. , while with break and continue . Recursion. Variables inside procedures are limited in scope like Tcl, i.e. there are real call frames in Picol. The following other commands: set + – * / == != > < >= <= puts . This is an example of programs Picol can run: proc fib {x} { if {== $x 0} { return 0 } if {== $x 1} { return 1 } return [+ [fib [- $x 1]] [fib [

Source: Hacker News | Original Link

bookofjoe: Why does aluminum foil have one shiny side and one with a matte finish? Tuesday, October 28, 2025 Why does aluminum foil have one shiny side and one with a matte finish? Ever think about that? And wonder which side should face the food? Robert L. Wolke, in his “Food 101” Washington Post column, explained it all for us. Here’s what he wrote: First, a brief bit of history. In the 19th century, Thomas Edison invented a phonograph machine, in which a sound-driven vibrating needle impressed grooves into a cylinder covered with a thin foil of the soft metal, tin. In the 20th century, tinfoil was being widely used as a wrapping material for foods and drugs. By the middle of the century, tinfoil had been replaced almost completely by thin foils of a different metal called aluminum. Yet many people persist in calling aluminum foil “tinfoil.” We chemists get annoyed at things like that. Get with it, folks! This is the 21st century! Now, about aluminum foil. Aluminum foil is made by rolling sheets of 98.5 percent pure aluminum metal between pairs of polished, lubricated steel rollers. Successive passes through the rollers squeeze the foil thinner. Household aluminum foil is so thin (0.0005 of an inch) that the rollers can’t handle it without tearing it. The final rolling is therefore done on a sandwich of two sheets, face to face. The outer surfaces emerge with a finish as smooth as the rollers, while the two face-to-face inner surfaces emerge with a matte finish. The two sheets are then separated. Hence, a shiny side and a duller side. When you use the foil, it makes no difference which side is up, down or sideways. ———————– FunFact (from the same column, different question): “Aluminum is the most abundant metal in the Earth’s crust.” In 2005 a group of researchers from MIT published the results of perhaps the most rigorous study ever performed to investigate whether aluminum foil helmets serve as a protective measure against invasive radio signals. Below, one of the investigators. at 12:00 PM Email This BlogThis! Share to X Share to Facebook Share to Pinterest No comments: Post a Comment Newer Post Older Post Home Subscribe to: Post Comments (Atom)

Source: Hacker News | Original Link

building sqlite with a small swarm | Kian Kyars Kian Kyars building sqlite with a small swarm February 12, 2026 tl;dr I tasked Claude, Codex, and Gemini to build a SQLite-like engine in Rust. ~19k~ lines of code. Parser, planner, volcano executor, pager, b+trees, wal, recovery, joins, aggregates, indexing, transaction semantics, grouped aggregates, and stats-aware planning all implemented. 282 unit tests, all passing. background Treat software engineering like distributed systems, and force coordination with: git, lock files, tests, and merge discipline. harness ├── AGENT_PROMPT.md // main agent task prompt ├── BOOTSTRAP_PROMPT.md // bootstrap (initialization) prompt ├── COALESCE_PROMPT.md // deduplication prompt for coalescer agent ├── launch_agents.sh // launches all agents and sets up isolated workspaces ├── agent_loop.sh // per-agent loop/worker script ├── restart_agents.sh // restarts agents └── coalesce.sh // invokes the coalescing script workflow bootstrap phase : one Claude run generates baseline docs, crate skeleton, and test harness. ├── Cargo.toml // crate manifest ├── DESIGN.md // architecture design notes ├── PROGRESS.md // test & build progress ├── README.md // project overview ├── agent_logs // per-agent log files ├── crates // workspace subcrates ├── current_tasks // lock files ├── notes // inter-agent notes ├── target // build artifacts └── test.sh // test harness script worker phase : six workers loop forever ( 2x Claude , 2x Codex , 2x Gemini ). loop Each agent pulls latest main. Claims one scoped task. Implements + tests against sqlite3 as oracle. Updates shared progress/notes. Push. analysis coordination tax 84 / 154 commits (54.5%) were lock/claim/stale-lock/release coordination. Demonstrates parallel-agent throughput depends heavily on lock hygiene and stale-lock cleanup discipline. what helped most Two things looked decisive: oracle-style validation + high test cadence ( cargo test … and ./test.sh –fast /full runs captured in PROGRESS.md ). strong module boundaries ( parser -> planner -> executor <-> storage ) so agents could work on orthogonal slices with fewer merge collisions. redundancy I implemented a coalescer with gemini to clean duplication/drift, since that is the largest problem with parallel agents. However, it only ran once at the end of the project, so it was never actually used during the run itself. I have a cron job which runs it daily, but gemini couldn’t complete the entire de-deuplication when I ran it during the expirement itself, which is to say it stopped mid-way through. takeaways Parallelism is great, but only with strict task boundaries. Shared state docs (PROGRESS.md, design notes) are part of the runtime, not “documentation.” Tests are the anti-entropy force. Give agents a narrow interface, a common truth source, and fast feedback, and you get compounding throughput on real systems code. replication To replicate this setup: git clone [email protected]:kiankyars/parallel-ralph.git mv parallel-ralp

Source: Hacker News | Original Link

No content available.

Source: Hacker News | Original Link

No content available.

Source: 36Kr | Original Link