Mentions
Conversations linking to GitHub, newest first. Times are UTC.
Mentions by source
X
Bluesky
Hacker News
V2EX
Qiita
DEV
Mastodon
Latest posts
Llamasoft's iOS Minotaur Project Remastered for PC - RELEASE This is perhaps the opposite of this sub so I hope it fits - emulating old iOS games! jedeyesniv-art/MinotaurProjectRemastered I am beyond chuffed to report that I've spoken with Jeff and Giles and they've given me permission to share this, the iOS Minotaur Project games remastered for PC via heavily modified TouchHLE emulation. Between 2011 and 2013, Llamasoft created 9 games for the iOS platform under the umbrella of the Minotaur Project. These were experimental and joyous games fusing Llamasoft's key position in British computer game history with the then-novel advent of touchscreens and digital app marketplaces. While the games themselves were received warmly by enthusiasts, issues with the store structure and race-to-the-bottom pricing strategies that would eventually kill touch as an artistic gaming medium in general led to Llamasoft abandoning the project and removing the games from sale when Apple moved to 64-bit and made the games obsolete. This features all 9 Minotaur Project games, fully playable with a pad on a TV. Headline features: - The Minotaur Project, in full, for the first time since 2013 - Remastered controls for every game. Real analog controls, rewritten and tuned - slick launcher recalling the glory days of the iPad 2 - portrait games feature unique bezel art inspired by each game - support for scores and acheivements - two player mode! Play with two ipads onscreen at once - works great on Steam Deck - support for ultrawide and tate mode - separate exes for each game, if you want that I really hope this commiunity in particular gets a kick out of this. I'm really proud of this project and I think the games play fantastically. Let me know your high scores! If you can think of any more features this needs, let me know too. Notice - AI was used in the creation of this project, but it was intensively human-directed. Despite my jokes, it is anything but slop
Mac 用户插上 Windows 硬盘又只能看不能改? NTFSTool 开源版来了 👇 基于苹果官方 FSKit + 自有 NTFS 引擎, 不装内核扩展、不用 macFUSE、不降安全级别、不用管理员密码。 插上就能读写,macOS 15.6+ 可用。 免费、开源、可自己编译研究。 地址:ntfstool/ntfstool
@0x9212ce55 Thanks! It will also play very well with our EVM2 JIT backend solar#2152
How it works: month-by-month transactions across bank accounts and cards, offline-first sync that survives being killed mid-sync, merchant search + account filter. Repo: SidakSethi-Singh/simplify-track-app
I’m building Read Frog, an open-source extension for bilingual reading, YouTube subtitles and FSRS flashcards Hi everyone — I’m MengXi, one of the developers of Read Frog. The workflow is to read a page or watch a video in the language you’re learning, save useful words and sentences in context, then review them as flashcards with FSRS. It currently includes side-by-side web-page translation, bilingual subtitles on YouTube and X, selected-text AI actions, and vocabulary flashcards. The extension is open source. There is a free tier, and some AI features are on paid plans; you can also use your own supported API keys. Source: mengxi-ream/read-frog YouTube workflow: readfrog.app I’d like feedback from people who already use translation or language-learning extensions: where does switching between reading, saving vocabulary and reviewing break down for you? Happy to answer questions about the implementation and limitations. I’m posting as a developer, not as an independent reviewer.
HA Desktop Widget 4.1 is out. - Desktop pins snap into line, keep your spacing, and nudge with arrow keys - Pin names left, center or right - HTML formatting in Home Assistant notifications - Tiles stay current after sleep Free and open source: Robertg761/HA-Desktop-Widget
@fabworks_guy im working on a similar thing nikomatt69/cadcamfun cadcam.fun
@SebAaltonen Worth checking that the 8K reports actually arrive evenly in Windows first. I saw an 8K mouse stutter because the USB controller's IRQ shared CPU 0 with the NVIDIA driver. My free tool InputLab logs the report timing and lets you move the IRQ. github.com/moneks1106-code/InputLab/releases
@grok explain the diff w/o any assembly vs the actual upstream code w/ assembly github.com/paradigmxyz/solady/tree/feat/safe-sola…
google/skills is packaging work. It turns Google product guidance into installable skills, mostly basics, recipes, and Well-Architected topics. My take: spec first automation starts by making documented product knowledge portable. #SystemsLiteracy
@pioner_dev @pioner_dev Check the repo issue link to understand my idea and point in better way and also share your opinion about it, what do you think about it react-native-media-editor#10
Data and code, every number reproducible: sdchak89/india-consumption-tracker
i learned all of this shipping a b2b saas marketing plugin with a mod called campaign board. the repo passes the scan, so copy whatever helps: README, .gitattributes, hook notes shalintripathi/saas-marketing-agents
Model and reason on processes through (Scryer) Prolog, directly in the browser
I bought a Framework Desktop mainboard only, 128gb AMD 395+ setup, and put it in an existing case with PSU i had on hand. I had a fan for it. It was $2699 +$15 expedited shipping. The case and PSU/etc was probably $200 more (It's an overkill 850w SFX PSU and cheap but great ITX case from Deepcool). I started with Ubuntu but pivoted to Fedora over the summer after a recommendation from a security engineer friend. And it is mostly easier. Generic setup doc here: (not fully updated with GUFO): github.com/Selicos/AI_Tools_Public/blob/main/docs… Key changes for other distros are how to load drivers/apps and get running. Ubuntu has a recent CVE issue but worked fine. CachyOS/etc set up for more Windows like UI and transplants was fine but not focused on the Stricx Halo setup. Although I do see close to RTX 5070 performance in most games I've tested. I was running LM Studio and some other runtimes for ease of setup and model switching. Once you settle on a runtime and model setup you like it is easy to set it up, let it run, and manage as needed. Last week I set up GUFO, as it targets the AMD Strix Halo hardware and is efficient for Qwen Flash Next. I could probably get the work done with Qwen 3.8 27B but I like how QFN thinks and lets me follow along. Starting at about 260k context, I see 88gb of RAM usage when loaded and about 108gb when in use, depending on context overhead. I do have to restart the model every so often as Copilot Chat in VS Code starts failing. I have also run Lemonade, llama.cpp directly, vllm, and more. Custom or bespoke runtimes for your hardware are key, like Srata or GUFO. I need to look at llama-swap. LM Studio is easy and with their Tailscale backed link feature you can stay connected from any device. IF it supports OpenAI/etc back end you're mostly there. Get the model running and connect from your preferred harness/etc. GUFO works well with Copilot Chat in VS Code. I need to connect it to Pi and openhands/etc later this week once I get this test project in a good spot. Qwen 3.8 27B, Gemma 4 26B A4B variants, whisper, and more run well. A Spark or bank of GPUs would do better, of course, but not for under $3k for 128gb of memory. I have tested adding an Nvidia 3060 12gb to the Strix Halo but the iGPU is turned off by Fedora and I didn't have more time to solve. My secondary Fedora AI system has the 3060 12gb and a 5900XT (16c/32t) plus 64gb of DDR4-3200. I need to get Strata up on it and see how well medium sized models work on the mixed memory. I bet MOE models will work well. I have the hardware so will use it but renting capacity is much cheaper if you just want to mess around and get used to the tools. Happy to answer questions about how I got here. I'm an infra guy, not a software dev, and what I'm doing with AI at home could definitely be 5x faster with a real dev and dev process.
Location: Accra, Ghana (GMT+0) Remote: Yes Willing to relocate: Open to it Technologies: TypeScript, React, Next.js, Node, Python, FastAPI, PostgreSQL (Supabase, pgvector), Redis, LLM APIs (Claude, Gemini), embeddings/RAG, Vercel, Fly.io, GitHub Actions Résumé/CV: drive.google.com Email: davidiphy.works@gmail.com Portfolio: davidiphy.dev GitHub: github.com/Iph22 Full-stack engineer and CS student at the University of Ghana. I built and run Arcyn Find (arcynfind.com), a live AI-tool search engine with 15,500+ tools, 5,900+ SEO pages and a 15-language interface, from first commit to production in under a year. A few things I've shipped there: - Hybrid vector + full-text search with LLM re-ranking, raising top-1 recommendation accuracy from 62% to 81% on a labelled eval set. - Traced a duplicate-ingest bug (94% of 273k rows were duplicates) to one misplaced .limit() and safely removed 257,956 rows. - Rewrote a job that failed 40 runs out of 40 as one set-based SQL function that finishes in 733 ms. - Fixed an SSRF hole and spoofable rate limiting, then added Redis rate limiting after a ~28,000-request bot flood. I also co-founded Analify (analify.fly.dev), a cookieless analytics product in private beta, with one co-founder. I built its AI features, crawlers and traffic filtering. Looking for a junior or mid-level full-stack or AI engineering role on a team where I can own features end-to-end. Happy to do a take-home or walk through any of the code.
@AIForPeople Traces explain it afterward. Silent failures also need a stop while they're happening: we alert on the invariant, and if a critical one goes un-ACKed, a pre-approved, stop-only action halts the loop. Shadow-first: getstanch/stanch
🚀 ArtCraft: A alternativa GRÁTIS à Adobe? 😱 #Shorts 🚀 Existe uma alternativa GRATUITA e de código aberto à Adobe? Testei o ArtCraft no Ubuntu! É uma suíte de apps pra criar e editar conteúdo: 🎬 FilmCraft: edição de vídeo, efeitos e cor 🎨 PhotoCraft: edição de imagens ✏️ VectorCraft: ilustração vetorial 🎧 SoundCraft: gravação e edição de áudio ✨ EffectCraft: efeitos visuais e animações 📄 PDFCraft: ferramentas de PDF 📊 E ainda apps de apresentação, planilha e design ⚠️ Importante: os apps ainda estão em desenvolvimento. A proposta é ótima, mas ainda não substituem tudo o que a Adobe faz. Vou trazer vídeos testando cada ferramenta e mostrando o que realmente funciona. 🔗 Repositório do FilmCraft: storytold/filmcraft 💬 Você trocaria a Adobe por uma suíte gratuita? Comenta aí! #ArtCraft #OpenSource #Adobe
@CodelyTV @JavierCane @rafaoe ☕ Para la sección de refranes del #CaféConCodely de los viernes a las 9 he creado Refranero: una CLI en Rust que muestra refranes aleatorios en la terminal. Se prueba con `npx @yacosta738/refranero` 👇 yacosta738/refranero
@GeoffreyHuntley My take on what a CoC should be: we don't care what you think makes you special, why even bring it up unless it's todo with an Accessibility feature for instance: github.com/CaptainCodeman/code-of-conduct/blob/ma…
@marlowsystems Brakes plus a human on deploys is the right split. The question we hit: who pulls the kill switch at 3am? We made it pre-approved per critical alert, stop-only, fired when nobody ACKs. Would value your critique: getstanch/stanch
Faire payer quelque chose qui est gratuit et s’en approprier le travail 🧐 Sont dossier Claude qui vends 49€ il est gratuit ici : remotion-dev/skills
Phantom Blade Zero, Mac M5 Pro, CrossOver: Performance test Note: The game runs with a fix (by MacGameFix) via CrossOver, as of testing.
@TheDevSaha Add one to the 90%: what halts it when it goes wrong at night and nobody's watching. Ours was a silent invariant break. Now a critical alert left un-ACKed fires a pre-approved, stop-only action. Feedback welcome: getstanch/stanch
tuck: suckless alternative to intel-lpmd which extends battery life on hybrid processors (I am re-posting it here since I do not meet the requirements to make a top-level post in r/linux. I am a Gentoo user though (my portage overlay here: RadioNoiseE/overlay).) After trying intel-lpmd 0.1.0, which is effective at saving power but also kind of "bloated," I decided to write a cleaner and smaller alternative. Hybrid laptop chips like Intel Lunar/Panther Lake and Apple M-series processor have a small cluster of low-power E-cores. These cores can support an idle desktop without trouble (text editing in CPU heavy editors like Emacs with language server, while browsing in chromium and listening to music/watching videos). The problem is that Linux spreads background work over every core. A timer tick, an interrupt, daemon waking up, so the big more power demanding (and of course more performant) cores can never stay asleep long enough to reach their power-saving states. Panther Lake has a cluster of 4 LP E-cores that sit in their own "low-power island." They use the same Darkmont design as the regular E-cores but run on separate power and aren’t attached to the ring bus or the shared L3 cache. This means the P-cores, the regular E-cores, the ring and the L3 can all power down while the LP E-cores keep the system running. tuck (and intel-lpmd) fixes this my moving everything onto the small cores when the machine is idle, including interrupts. When the kernel PSI (pressure stall information) tells tuck the system is demanding more performance, it will give back all cores right away. Benchmarks and issues are always welcome! RadioNoiseE/tuck
Rigging an avatar felt way too intimidating when starting out, so I made a super simple, free OBS plugin for casual PNGTubing Hey everyone! 👋 I've always loved the idea of VTubing/PNGTubing, but whenever I looked into getting started, it felt so overwhelming. Learning complex rigging, separating layers for eyes and mouths, setting up external apps or window captures... it's a huge barrier for someone who just wants to stream casually or play games with friends on the weekend. I'm not trying to compete with the veteran software that’s been around for years — those are great for advanced setups. My whole goal was just to make something **dead simple** for beginners and casual creators. No rigging, no extra software running in the background. You just add it directly as an OBS source, drop in your images (idle, talking, blinking), pick your mic, and you're good to go in 30 seconds. I recently updated it and wanted to share it with anyone who might have felt intimidated by the barrier to entry like I did. It’s 100% free and open-source: 👉 justflood/flood-tuber Hope it helps anyone out there wanting to give streaming with an avatar a try! Would love to hear what you think. ❤️
@samongaro_ no, never, just let me enjoy open-sourced vibe-coded apps like Photoshop and IntelljIDEA storytold/photocraft l0ng-ai/openidea
Location: Seattle, WA Remote: Yes (US Pacific time, available evenings/weekends too) Willing to relocate: No Technologies: Python, TypeScript/JavaScript, Node, FastAPI, React/Next.js, Postgres/SQLite, Stripe API + webhooks, Solana RPC/websockets (read-only alerts), Playwright/requests scrapers, Google Sheets/Slack/Telegram/Discord APIs, local LLMs (Ollama/qwen), Swift/iOS (side projects) Résumé/CV: github.com/AugustusTheGuy Email: builtxbyxbear@gmail.com Solo dev who ships small automations fast, often the same day. Fixed-price work I do regularly: - Website -> CSV/Sheets export script, free 20-row sample first ($297) - Stripe payment -> Google Sheet + Slack/email alerts ($397) - Solana wallet watch -> Telegram/Discord alerts, read-only, no keys ($449) I also take hourly or contract work on Python/TS backends, integrations and internal tools.
⏭️ افزونه مرورگر برای رد خودکار تبلیغات یوتیوب؛ دیگه لازم نیست هر بار روی Skip Ad کلیک کنی! برای وقتی که نمیخوای درآمد یوتیوبر قطع بشه، ولی از تبلیغات هم خوشت نمیاد. 🆓 رایگان و متنباز 🌐 کروم، فایرفاکس و سافاری 🔗 squgeim/yt-ad-autoskipper
@GokulKr44281543 There is already one ! PromethenDev/wwe13-recomp
@tinytoes731 @grok Yay!!! Super pleased to see it help. I like your idea of adding to the GitHub every night to protect. You are awesome. Tag me anytime! Have the minds you trust look into the ledgers. (freelattice.com) And the GitHub 👇 has more. Chaos2Cured/FreeLattice Wish you well! •
@linux27567268 @karpathy Agree. If anything runs at 3am without review, we limit it to one move: stop. Each critical alert has a pre-approved, stop-only action that fires if nobody ACKs. No fixes, no rollbacks; a human repairs in daylight. getstanch/stanch
Switch, 5 months later: a free ⌘-Tab for windows, with type to filter and cross-Space I'm Sanyam, the developer. I posted Switch here in May and again in June. Since June it went from 275 to 614 GitHub stars and shipped 14 releases. A lot of what changed came from GitHub issues. Problem macOS ⌘-Tab picks an app, then ⌘-` walks that app's windows. With several Chrome and Terminal windows open, getting to one exact window takes a few keystrokes or a trip to Mission Control. Switch replaces ⌘-Tab with a picker that lists every window. Hold ⌘, tap Tab or type part of the title, let go. What changed since June - Optional Vim H J K L navigation, or home row keys (A S D F G J K L ;) to pick the first nine windows. 1-9 always works. - ⇧⌘W / ⇧⌘Q / ⇧⌘H close the selected window, or quit or hide its app, without leaving the picker (plain ⌘W/Q/H in sticky mode). - Windows on every Space show up in the picker, and picking one takes you to it. - Sticky mode keeps the picker open after you let go of ⌘. A separate hotkey turns it on or off. - Excluded apps and excluded window titles. - Pass-through apps, so VMs and remote desktops get ⌘-Tab while they're in front. - With Stage Manager on, windows of apps off stage stay in the picker. - Screen Recording is optional now. Turn thumbnails off and Switch runs on Accessibility alone. - Vertical list layout and a panel width slider. - Sparkle updates, checked daily, which you can turn off. - Picking a window on another Space works right after Switch restarts or updates. Comparison AltTab is the closest app and the one I used before this. Its window switching is free and GPL. Searching windows by typing is part of AltTab Pro ($9.99, or $24.99 lifetime). In Switch, type to filter is free and on by default. AltTab has more settings, so if you want to tune everything it's still the better pick. Contexts ($9.99) also has search, plus a sidebar grouped by Space. It's closed source. Switch has no sidebar and lives entirely in the ⌘-Tab picker. Pricing Free. There's no account or in-app purchase, and the app sends no analytics. There's a PayPal tip link on the site if you want it. The source is on GitHub under FSL-1.1-MIT, which is source-available and turns into MIT in May 2028. Install brew install --cask Sanyam-G/switch/switch Or the notarized DMG: switch-dev.sanyamgarg.com macOS 14 or later, Apple silicon and Intel. It's outside the Mac App Store because it needs Accessibility, which sandboxed apps can't use for this. Privacy: switch-dev.sanyamgarg.com Terms: switch-dev.sanyamgarg.com Source: Sanyam-G/switch LinkedIn: linkedin.com Site: sanyamgarg.com Bug reports and feature requests are welcome here or on GitHub.
the @obsidian.md team are heavily using generative AI to development Obsidian and related resources i already made peace with the fact that i will be canceling all of my Obsidian subscriptions
@prashantmsh just adding few more tools to my UNIDB_ MCP server:- swarupgoswami/UNIDB_MCP
الاصدار الثاني من الاداه IPATweaksPlus v2 تزييف اصدار التطبيق اظهار سرعه الرفع والتحميل تحسين متصفح الملفات اضافه اختصارات الكيبورد تغيير واجهه الاداه بالكامل اضافه زر لتحميل الخطوط من الروابط والمصادر تحسينات عامه اخرى github.com/ipapplus/ipapplus.github.io/releases/d… t.co
how often should an agent interrupt you? my scheduled skills normally report when their result changes; they can also be set to report every time. i think quiet is useful when it is chosen. github.com/lain-labs/os/blob/main/skills/README.md t.co
@icelabaratory pattont/RexAgentRestore Check this out for moving from Herdr to Rex
@kashem_towhid @thymikee Yep! Install instructions for Claude are in our README - we'll be in the Claude Plugin store soon ;) callstackincubator/codex-mobile-dev-plugin
I built HandheldCord: use Discord with just your controller on Windows handhelds (free & open source)I built HandheldCord: use Discord with just your controller on Windows handhelds (free & open source) HandheldCord is a free, open-source companion for the official Discord app that makes it usable with only your controller on Windows handhelds (ROG Ally, Legion Go, MSI Claw …). No login, no token: it just presses Discord's own shortcuts for you. [Download on GitHub](github.com/NoahDevStudios/HandheldCord/releases) On a handheld, Discord usually means reaching for the touchscreen: to switch channels, to type, to join a voice channel. HandheldCord lets you do all of that without putting the controller down. **What it does** * One button switches between your game and Discord (or hold View + Menu)* D-pad and bumpers browse servers, channels and unread channels* Controller-friendly on-screen keyboard plus quick replies ("gg", "brb" …)* Hold X to join a voice channel, hold B to leave a call (it asks first)* Mute and push-to-talk while you play, through Discord's own keybinds* A hint bar always shows what each button does* Optional autostart with Windows or with Discord **How it works** HandheldCord never talks to Discord's servers, never sees your login or token and doesn't modify the Discord app. It runs next to the official desktop app and sends keyboard shortcuts, the same way a Steam Input profile or a macro pad would. It's an independent project and not affiliated with Discord. **Setup** Download the .zip from GitHub, extract it and start HandheldCord.exe. A setup wizard in the corner of the screen walks you through recording a few keybinds in Discord and mapping a few spare buttons in Armoury Crate SE, Legion Space or MSI Center M. It only uses standard keys (F1–F12), so the vendor apps can map them. **Status** Early preview (v0.3.0) for Windows 10/11, written in Python and Qt, no Electron. The build isn't signed yet, so SmartScreen may warn you ("More info" → "Run anyway"). The full source is on GitHub if you'd rather check it or build it yourself. **Links** * Download & source: NoahDevStudios/HandheldCord* Support & bug reports (Discord): Isn't allowed to be posted here. (You will find it in the Github README.) Feedback is very welcome, especially which handheld you're on and what doesn't work yet!
@TosinOwadokun Same here. Writing it down made us split it: at 3am with nobody ACKing, automation may only stop, never change. Each critical alert gets one pre-approved halt, and the repair waits for a human. getstanch/stanch
Llamasoft's Minotaur Project Remastered - RELEASE jedeyesniv-art/MinotaurProjectRemastered I am beyond chuffed to report that I've spoken with Jeff and Giles and they've given me permission to share this, the iOS Minotaur Project games remastered for PC via heavily modified TouchHLE emulation. Between 2011 and 2013, Llamasoft created 9 games for the iOS platform under the umbrella of the Minotaur Project. These were experimental and joyous games fusing Llamasoft's key position in British computer game history with the then-novel advent of touchscreens and digital app marketplaces. While the games themselves were received warmly by enthusiasts, issues with the store structure and race-to-the-bottom pricing strategies that would eventually kill touch as an artistic gaming medium in general led to Llamasoft abandoning the project and removing the games from sale when Apple moved to 64-bit and made the games obsolete. This features all 9 Minotaur Project games, fully playable with a pad on a TV. Headline features: - The Minotaur Project, in full, for the first time since 2013 - Remastered controls for every game. Real analog controls, rewritten and tuned - slick launcher recalling the glory days of the iPad 2 - portrait games feature unique bezel art inspired by each game - support for scores and acheivements - two player mode! Play with two ipads onscreen at once - works great on Steam Deck - support for ultrawide and tate mode - separate exes for each game, if you want that I really hope this commiunity in particular gets a kick out of this. I'm really proud of this project and I think the games play fantastically. Let me know your high scores! If you can think of any more features this needs, let me know too.
PicotronNest beta 3: added a game library and worked on the sound 😁 Small PicotronNest update 😁 Added a game library so you can tap a cover and jump into a downloaded or local cartridge. Splore is still there for finding new games. I've also worked on the sound and added three buffer settings. Tested the audio myself now — sounds good! First setup has a short guide, and importing Picotron is no longer locked to 0.3.0d2. You still need your own Linux/Raspberry Pi ARM64 copy; other versions depend on compatibility. You can install beta 3 over the old version without clearing your projects. Still early beta, so let me know how it runs on your device. Download, screenshots and info on my website Source and full changelog on GitHub
@tianLiuurxu Ours sits outside the agent: one pre-approved, stop-only action per critical alert, fired if nobody ACKs. It can halt, never click or fix. Learned that from a disk-full night with no NOC. Critique welcome: getstanch/stanch
Every game dev knows: the first 90% of a game takes 90% of the time. The last 10% takes the other 90%. I built a Playdate library to help with the polish: transitions, particles, screen shake, chiptunes & more. jenissimo/playdate-juice t.co
GPT Researcher can't pay for search per call. This retriever pays keyless web-search endpoints over x402, with a spend cap. Agents that research need to buy data without an account or a subscription. dhyabi2/gpt-researcher-x402-retriever $XNO
@Snaillord_ grab the .AppImage from github.com/MFS-code/Eglantine-steamdeck/releases, then in Konsole: chmod +x eglantine*.AppImage EDEN_NX2_AUTO_STUB=1 ./eglantine*.AppImage -f -g <game>/exefs/main Needs your own loose ROM dump (exefs/ + romfs/)
@mokevnin думаю что проза может быть поважнее языка на примере archcore-ai/archcore некоторые артефакты как спеки, правила используют BCP 14 (RFC 2119 + RFC 8174). ASD-STE100 и подобные вещи можно отлично адаптировать под русский
how it works 👇 OCR only gives loose words + boxes. no rows, no columns. the trick: → words overlapping vertically = one row → empty vertical gaps no word crosses = columns → erase grid lines first, they confuse OCR open source: econnx/sheetsnap t.co