SShortSingh.
TopTechnologyAI & MLStartupsProgrammingScienceBusinessFinanceCrypto & Web3World

Programming

LatestTopDiscussed
0
ProgrammingDEV Community ·

Security review flags concentrated voting power and short timelock risks in OKX governance

A DeFi security researcher has published a governance attack surface review of OKX, which operates a hybrid on-chain and off-chain governance model with roughly $30.6 billion in TVL. The review examines core components including the OKB token, governor and timelock contracts, a multisig treasury, cross-chain bridges, and an oracle registry, most of which use upgradeable proxy patterns. Identified risks include concentrated voting power held by a small number of entities, flash-loan-based voting manipulation, a 24-hour timelock considered too short for community response, and the possibility of malicious upgrades if the governor is compromised. The author rates the overall risk at 7 out of 10 (high-medium), noting the system is well engineered but dependent on several critical design choices. The review is dated 20 October 2026.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

A Practical Afternoon Plan for Automated QA on a Small Web App

Small web apps tend to fail in the same few places, including signup, login, the main form, checkout, and the most-visited page. The article recommends listing the five to ten flows whose failure would cost money or users, and naming for each a start page, steps, and a clear on-screen success signal. It advises running tests against a stable, safe environment such as a staging deployment, with a dedicated test account and test payment cards, and planning how one-time login codes will be handled. A free, no-account audit tool can show what an automated service could cover from the outside, which helps identify flows that must be tested manually. The author outlines three approaches: writing tests by hand with Playwright or Cypress, using recorder tools that tend to be brittle, or using AI agents to write and review a suite for a fee starting at $29 per month.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

ghostsubs uses local Gemma 4 to find recurring charges in bank CSV exports

ghostsubs is a command-line tool that scans CSV exports from bank or card accounts and lists every recurring payment, such as subscriptions. Its main difficulty is identifying the business behind each charge, since bank descriptors are often cryptic, so it uses the Gemma 4 12B model running locally through Ollama to name each one. The rest of the processing, including dates, amounts, and totals, is handled by ordinary code. The report flags price increases, trials that converted to paid plans, upcoming annual renewals, and services that have stopped charging. Because the model runs on the user's own machine, bank data stays private unless the user opts into the ntfy notification or a remote model.

0 comments Read more at DEV Community
0
ProgrammingHacker News ·

NixOS Is Making Me Less Afraid, Blogger Writes

A personal blog post titled "NixOS is making me less afraid" was shared on Hacker News, linking to an essay on entropicthoughts.com. The post reflects the author's personal experience with the NixOS Linux distribution and how it has reduced their anxiety about system changes. The submission drew only 3 points and no comments on Hacker News at the time of the listing. The available material consists solely of the title and link, so the specific reasoning and details behind the author's experience cannot be confirmed from this item alone.

0 comments Read more at Hacker News
0
ProgrammingHacker News ·

Demographic 'Doomsday' Predictions Miss the Mark, Science Article Argues

A Science article titled 'Demographic Doomsday Misses the Mark' challenges predictions of catastrophic population decline. The piece is available via a DOI link on science.org and was shared on Hacker News, where it gathered modest attention. The provided source material includes only the headline, link, and basic engagement metrics, so the specific arguments and data cited in the article cannot be detailed here. Readers seeking the full analysis are directed to the original Science publication.

0 comments Read more at Hacker News
0
ProgrammingHacker News ·

Developer Shares Four-Year Experience Using the Decker Tool

A blog post reflects on four years of working with Decker, a tool for building interactive creative projects. The author describes their experience and what they have learned over that period. The post was shared on Hacker News, where it gathered 46 points and 9 comments. The article is available on the author's personal blog at beyondloom.com. The source text provided is limited, so the specific details of the author's conclusions cannot be fully confirmed here.

0 comments Read more at Hacker News
0
ProgrammingHacker News ·

Apple Accused of Breaking a Key iPad Presentation Feature

A blog post by Teddy Svoronos claims that Apple broke an essential presentation feature on iPads. The article was shared on Hacker News, where it received 6 points and no comments at the time of capture. The source text provided is limited to the link and metadata, so the specific feature and the nature of the breakage are not described here. Readers would need to consult the original article for details on the issue and Apple's response, if any.

0 comments Read more at Hacker News
0
ProgrammingHacker News ·

Hazel, a YC W24 startup, is hiring a full-stack TypeScript engineer

Hazel, a company from Y Combinator's W24 batch, has posted a job listing for a full-stack engineer with TypeScript experience. The listing appears on Y Combinator's company job board. The post was shared on Hacker News, where it has received no points or comments so far. The source material provides no further details about the role's responsibilities, compensation, or location.

0 comments Read more at Hacker News
0
ProgrammingDEV Community ·

Why .NET chat loops lose OpenAI encrypted reasoning between turns

A reasoning model can appear to forget the first turn of a conversation, and the cause is often in the application's own chat loop rather than the prompt or model. When stateless requests ask for encrypted reasoning, the response includes an encrypted reasoning item that the service does not store, so it must be sent back on the next turn. In Microsoft.Extensions.AI, this data sits in TextReasoningContent parts, which are separate from the plain text the response exposes. Developers who save only the response text in their history discard this encrypted reasoning, so the follow-up request carries no prior reasoning items. The author demonstrates the problem with an offline .NET sample and tests, and recommends appending the entire response to the history instead.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Crafting and Parsing L2-L7 Packets with Scapy: A Hands-On Lab

The author walks through a networking lab aimed at debugging layer 2, 3, 4, and 7 behavior using Scapy. At layer 2, an ARP request sent to a broadcast address is shown, and a layer 3 test illustrates how a target ignores a broadcast-addressed packet while tcpdump captures traffic. For layers 4 through 7, the author builds a packet with Ether, IP, TCP, and Raw layers, carrying an SSH banner payload to port 22. The packet is verified on a VyOS system using tcpdump, which captured it even though no TCP handshake was performed. The author concludes that Scapy provides full control over payload content within the TCP/IP envelope.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Five Unhappy-Path Tests to Write Before Trusting a .NET 10 Multi-Agent Workflow

A developer building a support-ticket triage workflow on Microsoft Agent Framework for .NET 10 argues that demos only show the happy path, while real failures include malformed model output, unapproved promises and provider outages. They built the workflow with a classifier agent, two specialist agents and plain C# steps, and wrote 58 tests that run without an API key or network access. The core technique is a deterministic fake chat client that can be told to produce specific failures, such as cut-off JSON or an HTTP exception, via trigger strings. The author also describes a reply-guard step that checks specialist drafts for oversized refunds, guarantee language and sensitive data, routing risky replies to human review. The post presents five such tests as a starting point for any agent workflow, along with the bug each one catches.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Kraken CSV Format Change Exposes Risks in Crypto Tax Importers

A CoinTaxman bug report from July 2026 flagged that a newer Kraken Ledgers export adds subclass and wallet columns, which older parsers cannot handle. The same report noted unhandled ledger types such as earn and hybridearnwithdrawal, along with asset naming mismatches like BTC/XBT and DOGE/XDG. The author argues that importers should reject unrecognized layouts and unclassified events rather than guess their tax treatment. The proposed fix recognizes the new twelve-column header as a distinct format and flags unsupported wallet types for manual review. The author stresses that the patch is not a full tax determination or complete Kraken importer.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

GSTR-3B Filing History Offers Lenders a Credit Signal for MSMEs

GSTR-3B is a monthly or quarterly GST summary return in which businesses self-declare outward supplies, input tax credit and tax paid, and it cannot be revised once filed. Because each filing is a fixed, dated record, its history can show how regularly a business trades, reports and pays tax. This matters for MSME borrowers in India, who often lack audited financials but do have GSTINs and years of returns. Lenders are urged to read the full period-wise record rather than simply confirming that a GSTIN is active. The article outlines how 2025 changes, including locking of auto-populated liability and a three-year time bar on filing, have strengthened the return as a data source. It also lists signals such as filing regularity, delay trends, nil returns, turnover consistency and cash-versus-credit tax payments, none of which is conclusive alone.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Samsung y Reality Defender llevan la detección de deepfakes al teléfono Galaxy

Samsung Electronics Singapur y la empresa neoyorquina Reality Defender firmaron un memorando de entendimiento durante GovWare 2026 para integrar la detección de contenido falso en dispositivos Galaxy. La herramienta analizará voz, video e imágenes generados con inteligencia artificial directamente en el teléfono, sin enviarlos a la nube. El despliegue arranca con clientes empresariales en Singapur y se apoya en la plataforma de seguridad Knox. Las cifras citadas muestran la magnitud del problema, con víctimas de estafas de suplantación de funcionarios que perdieron al menos 242,9 millones de dólares singapurenses en 2025. Los responsables sostienen que el análisis local protege la privacidad y evita demoras antes de que un empleado actúe sobre un mensaje falso.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Why CSS Variables Can Slow Rendering: Runtime Resolution Costs Explained

A developer found that CSS custom properties, often assumed to be free abstractions, can cause noticeable slowdowns in complex web apps. Unlike fixed values, variables are resolved at runtime because they can be inherited and overridden anywhere in the DOM tree. The browser must therefore climb ancestor elements to find each variable's value during every style recalculation. Changes to variables on parent elements can trigger repeated invalidations and re-resolutions across many descendants, which is costly during animations or theme switches. The article's author traced the problem to frequent style recalculations using Chrome DevTools' Performance tab.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Abbottabad Founder Launches Independent Digital News Platform Buzz Capitol

Sami Ullah, a 23-year-old from Abbottabad, Pakistan, has launched Buzz Capitol in 2026 as a self-funded, independent digital news platform. He serves as its founder, CEO and editor-in-chief, operating without investors or a traditional newsroom. Ullah previously ran more than 15 YouTube channels and built a combined following of over 10 million across social platforms, which he describes as his practical training in online distribution. He says the project aims to cover stories about places outside major cities and to offer young readers independent reporting. He emphasizes that trust and consistent editorial standards, rather than the technology stack, are the platform's core assets.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

AWS CDK Adds Deterministic Guardrails to Block Destructive Changes from AI-Written Code

Developers increasingly rely on AI to write AWS CDK code and sometimes to deploy it, raising concerns about accidental deletion of critical resources. Relying on AI to judge diffs is non-deterministic, so the article proposes a mechanical alternative. Deleting or replacing resources is considered destructive, since it can cause downtime or data loss, and some resource properties trigger replacement even when no deletion appears in the template diff. AWS CDK CLI v2.1145.0 introduced the --fail-on destructive option for cdk diff and the --require-approval destructive option for cdk deploy to detect and block such changes. This setup lets AI drive development while requiring human approval only for destructive changes before the CI pipeline proceeds.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

LangGraph, CrewAI and AutoGen Compared on Architecture for Production Agents

A comparison of three popular agent frameworks, LangGraph, CrewAI, and AutoGen, focuses on how each executes workflows, manages state, and recovers from failures rather than on feature lists. LangGraph models workflows as cyclical graphs with typed state checkpointed after every step, enabling pause, resume, and replay. CrewAI organizes agents by role and task within crews, with simpler mental models but less explicit state handling, and offers Flows for more control. AutoGen relies on conversational message exchange among agents, which suits discussion-style tasks but can be harder to predict. The article notes that AutoGen has been in maintenance mode since October 2025, with Microsoft Agent Framework as its successor. No single framework is presented as best; the choice depends on whether deterministic, resumable execution or faster prototyping is the priority.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

playwright-webext Simplifies Testing Browser Extensions in Playwright

A DEV Community post introduces playwright-webext, a project designed to make testing browser extensions easier than with Playwright alone. Playwright's official guide covers launching a browser with an extension and locating its service worker, but extension-specific setup such as storage, popups, and version upgrades still requires custom fixtures and helpers. The author demonstrates the library using a simple todo-list extension that stores tasks in chrome.storage.local. The tutorial walks through installation and configuration, then shows three example tests covering installation behavior, popup interaction, and migration between extension versions.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Building MOSAIC: Why Reproducible Replay Matters for Multi-Agent AI Experiments

The author describes MOSAIC, a research project for studying populations of language-model agents, and the difficulty of knowing whether observed behaviour changes stem from a model swap or ordinary generation variation. MOSAIC uses an append-only event log and a deterministic reducer so that a run's state can be reconstructed and checked against its recorded history. The system also records model identity via digests, generation parameters, and protocol settings, so differences between experiments can be explained. A record-and-replay mechanism saves model generations so that runs can be re-executed without fresh model calls, supporting debugging and validation. The author stresses that a system can execute exactly as designed yet still fail to yield the observations needed to answer the research question.

0 comments Read more at DEV Community
← NewerPage 280 of 3054Older →