SShortSingh.
0
ProgrammingDEV Community ·

Testing Optimistic UI Rollback by Letting Demo Users See a Failed Change

A public SaaS demo can show whether an interface handles a rejected change by displaying an immediate update, deliberately rejecting the underlying operation, and checking what users understand afterward. The approach separates a pending change from a confirmed one, since a fast animation signals responsiveness but does not prove the change was saved. The exercise uses a fictional task board with a disposable item, a known saved value, and a reversible action so the difference between displayed and stored state is easy to follow. Testers are asked what they believe has happened, what they expect after a refresh, and whether the interface makes the failure clear. The author notes that the example is a proposed review exercise, not a report on any specific product, and that each product needs its own recovery policy.

0
ProgrammingDEV Community ·

Developer Controls Local AI Coding Agent via Telegram Voice Notes

A developer built agy-pilot, a tool that links Telegram with Google's Antigravity, a local AI coding environment, so tasks can be assigned remotely. Voice notes sent from a phone are downloaded to the Mac running the agent, which can read the audio directly without a separate speech-to-text service. The setup relies on outbound long-polling over HTTPS, so no ports need to be opened on the home router, and only an allowlisted Telegram user ID is accepted. To avoid flooding the chat and hitting Telegram's rate limits, progress is shown in a single status message that is updated at most once every 1.5 seconds. Risky actions can be approved with a tap from the phone.

0
ProgrammingDEV Community ·

Tabs in One gathers open Chrome tabs into a single-column page to free up RAM

A developer has released Tabs in One, a free Chrome extension that moves all open tabs onto one clean, single-column page and closes the originals. The developer says the tool was built because Chrome's tab bar becomes hard to read with many tabs and heavy tab use drives up memory consumption. Users can restore tabs individually or all at once, and the extension offers daily backups, JSON or HTML export and import, and usage statistics. It requires tabs, storage, contextMenus and alarms permissions, does not sync between devices, and is open source, according to the developer. It is available free on the Chrome Web Store with no account required.

0
ProgrammingDEV Community ·

DuckHarness Lets Coding Agents Adapt to How You Prefer to Work

DuckHarness is a tool that gives coding agents shared guidance so users do not have to repeat the same instructions each session. It is installed from a checkout using an install script, which sets up DuckHarness and the Harness² component and registers it with uv. Commands such as dharness skills create or refresh the agent guidance files, while dharness status, gate, and decline help users review and manage proposed changes to that guidance. The project is licensed under Apache-2.0, with updates pulled from its SourceForge repository and Harness² refreshed through the dharness update command.

0
ProgrammingDEV Community ·

Codex CLI dapat membaca Note dan knowledge base osFoundry lewat MCP

Panduan ini menjelaskan cara menghubungkan Codex CLI ke proyek osFoundry melalui server MCP, sehingga asisten dapat membaca Note spesifikasi, knowledge base support, dan Record bug sambil bekerja di repositori pengguna. Setiap tindakan yang menulis meminta persetujuan pengguna terlebih dahulu, dan semua panggilan tool tercatat di aktivitas terbaru koneksi tersebut. Pengaturannya mencakup satu tabel konfigurasi di ~/.codex/config.toml, satu kali login di browser, dan penambahan koneksi ke proyek. Konfigurasi juga membatasi scope dan daftar tool yang tersedia, sehingga hanya sejumlah tool baca dan edit Record yang diaktifkan. Data tetap berada di osFoundry, sementara item yang hanya bersifat lokal tidak pernah meninggalkan perangkat.

0
ProgrammingDEV Community ·

Tabs in One: a Chrome extension that gathers open tabs into one list to free RAM

A developer has released Tabs in One, a free Chrome extension designed to tackle the slowdown caused by dozens of open tabs. It collects all open tabs onto a single clean, single-column page and closes the originals, which immediately frees memory. Users can restore any tab or all of them with one click, and the tool skips duplicate links and offers daily automatic backups with JSON or HTML export. The extension requires tabs, storage, contextMenus and alarms permissions, needs no account, and its code is publicly reviewable, though it does not sync sessions across devices.

0
ProgrammingHacker News ·

Claude Code's "Don't Ask Again" setting linked to secrets exposed in 1,122 public GitHub files

A security analysis reported that Claude Code's "Don't Ask Again" permission option may be associated with passwords appearing in 1,122 public GitHub files. The findings were published by Apiiro in a blog post examining how coding agents may be leaking secrets to GitHub. The source material is limited to a link and brief discussion metadata, so the specific mechanism and scope of the exposure cannot be fully confirmed from this item. The story gained attention on Hacker News, where it received 30 points and 11 comments.

0
IndiaTimes of India ·

Virginia restaurant hit with $51,594 water bill after years of unrecorded usage

A restaurant in Virginia received a water bill totaling $51,594 following a three-year period in which its meter recorded no readings. The large charge was reportedly tied to the unmonitored consumption that accumulated over that span. The figure drew attention because the establishment had not been receiving regular billing updates during that time. The case highlights how gaps in meter reading can lead to substantial, unexpected costs for businesses.

0
IndiaTimes of India ·

15-Year-Old Rodion Zaichikov Wins Science Prize for Building Arc-Jet Wind Tunnel

Fifteen-year-old Rodion Zaichikov has drawn attention in the scientific community for building an arc-jet wind tunnel designed to test scramjet engines. The device is meant to recreate extreme flight conditions, which are usually expensive and difficult for researchers to access. His work earned him the Ellison Scholars Prize at Germany's European Union Contest for Young Scientists. The project reflects his commitment to developing new aerospace technology.

0
IndiaTimes of India ·

Squids Found to Have Sensory Hair Cells Across Their Entire Bodies

Researchers at Case Western Reserve University have discovered that squids carry sensory hair cells throughout their bodies rather than only on their heads. These cells let squids sense nearby movements of water. The process works somewhat like how humans use their ears to detect sound. Differences in the lengths of the hair bundles help squids tell apart water motions of different frequencies.

0
IndiaTimes of India ·

Arkansas Appeals Court Upholds Sale of 111 Acres to Couple Who Paid With Truck and Cows

Destin and Jennifer Clark were involved in a legal dispute over a 111-acre property in Magazine, Arkansas. The couple initially had difficulty keeping up with payments and reached several arrangements with the sellers, George and Sandra Thomas. The conflict escalated when the sellers declined to supply a signed payoff figure that the Clarks needed to close the transaction. The Arkansas Court of Appeals upheld the earlier lower-court decision, clearing the way for the Clarks to finish buying the land.

0
ProgrammingDEV Community ·

Claude Code Action Can Automatically Review GitHub Pull Requests

The anthropics/claude-code-action GitHub Action lets Claude review pull requests and answer @claude mentions in issues without a dedicated server, running on the user's own GitHub runner. It selects its behavior from the workflow context, handling code questions, PR reviews, and simple fixes, and sends API calls to a chosen provider such as Anthropic, Amazon Bedrock, Google Vertex AI, or Microsoft Foundry. Setup is done by running claude /install-github-app from a repo terminal, which requires repository admin rights and is documented for direct Anthropic API users, with other providers covered in the repo's cloud-providers guide. The project's Solutions Guide lists review patterns such as automatic PR review, path-specific checks, and OWASP-aligned security reviews, and the article recommends starting small and limiting tool access. The author stresses that the AI review should be treated as a first pass rather than a replacement for human approval, and notes they did not run the action on a live repository.

0
ProgrammingDEV Community ·

A 65B Local Model Outscored a 501B Model on Identical Eval Questions

A developer reported that a 65-billion-parameter model run locally scored 80.52 on their evaluation suite, higher than a 501-billion-parameter frontier model on the same test. Both models answered the same 200 multiple-choice questions at temperature 0 with no system-prompt changes, using the same scoring script. The local model was quantized to Q4_K_M and ran on a single consumer GPU, while the larger model was accessed through its hosted API with default settings. The smaller model reportedly won 14 of 20 categories. The author noted that verifying the larger model's result costs about twelve dollars per full run, compared with roughly forty minutes of GPU time and electricity for the local run, and shared a short video walkthrough of the setup.

0
ProgrammingDEV Community ·

Developers Explore Running Local LLMs in Browsers Using WebGPU and Transformers.js

A technical article from DEV Community describes how WebGPU gives browsers direct access to local GPU compute, enabling quantized open-weight models to run client-side. It names Qwen 2.5, Phi-3.5, and Llama 3.2 as examples of models that can execute within the browser sandbox. The author highlights a code example using the Transformers.js pipeline API and notes challenges such as shader compilation latency during initial load. Caching weights through the browser's Cache Storage API is suggested so models are downloaded only once. The piece argues that keeping prompts on the device avoids server round trips and improves data privacy for sensitive information.

0
ProgrammingDEV Community ·

Deduplizieren einer Kundenliste in Python, wenn keine Zeilen exakt identisch sind

Doppelte Einträge in Kundenlisten sind selten identische Zeilen, weil sich etwa das Bestelldatum, die Stadt oder die Schreibweise der E-Mail unterscheiden. Welche Zeilen als Duplikate gelten, hängt daher vom gewählten Vergleichsschlüssel ab, der bestimmt, wie viele Zeilen entfallen und welche erhalten bleiben. Der Artikel zeigt dies anhand einer Beispielliste mit 412 Zeilen in reinem Python und vergleicht die Ergebnisse mit dem Befehl Remove Duplicates im Spreadsheet Editor von osFoundry. Ein Schlüssel über die ganze Zeile entfernt keine Einträge, die E-Mail allein entfernt alle 38 Wiederholungen, und E-Mail plus Stadt entfernt 25. Zudem behält die übliche Konvention die erste Zeile, was bei den meisten Dubletten nicht die neueste Bestellung ist, weshalb die Sortierung vor dem Deduplizieren wichtig sein kann.

0
ProgrammingDEV Community ·

Addy Osmani's agent-skills brings senior-engineer workflow habits to Claude Code

The addyosmani/agent-skills repository packages engineering practices such as writing specs first, breaking work into small tasks, and reviewing before merge into skills that AI coding agents follow. Its README organizes the skills around a lifecycle of define, plan, build, verify, review, and ship, with nine slash commands mapped to those phases. In Claude Code, the plugin installs through the marketplace, with HTTPS workarounds offered for SSH permission errors, and the review command must be called in namespaced form because of a built-in conflict. The same skills can be installed across more than 70 agents via the open skills CLI, though per-skill installs omit shared reference files. The publication notes it followed the README without running the workflow end to end, and warns that commands and skill counts may change.

0
ProgrammingDEV Community ·

AI가 항균 펩타이드 탐색을 앞당기며, 구조 이해 단계의 병목이 주목받다

DeepMind가 Google Antigravity를 활용해 항균 펩타이드(AMP) 탐색 속도를 높이는 방법을 블로그에 공개했다. 이 파이프라인은 후보 서열과 구조를 빠르게 스크리닝하는 데 초점을 둔다. 그러나 후보가 막과 어떤 기하학으로 상호작용하는지, 어떤 잔기가 결합 포켓을 이루는지는 서열 점수만으로 알기 어렵고 3D 구조를 직접 살펴봐야 한다. 기사는 2D 화면에서의 반복적 조작이 인지 부담을 키워 탐색 이후 단계의 처리량을 제한한다고 지적한다. 이를 보완하기 위해 Seven Sportz가 개발한 Meta Quest 앱 넥소피크는 구조 데이터를 공간 컴퓨팅 환경에서 보여주는 방식을 제시한다. 해당 앱은 교육 및 연구 시각화 도구이며 의학적 진단이나 치료를 제공하지 않는다고 명시한다.

← NewerPage 473 of 7351Older →