SShortSingh.
TopTechnologyAI & MLStartupsProgrammingScienceBusinessFinanceCrypto & Web3World

Programming

LatestTopDiscussed
0
ProgrammingDEV Community ·

AI Coding Boom Floods App Stores, But New Apps Earn Just 3% of Revenue

The number of new subscription apps launched monthly has surged from around 2,000 in January 2022 to over 14,700 by January 2026, driven largely by AI coding tools, according to RevenueCat's State of Subscription Apps 2026 report. Despite the flood of new entrants, apps launched before 2020 still capture 69% of all subscription revenue, while apps released in 2025 or later collectively share just 3%. Global app downloads fell for the fifth consecutive year in 2025, dropping 2.7% to 106.9 billion, even as consumer spending on non-game apps rose 33.9% to $82.6 billion. The data suggests users are spending more on familiar, established apps rather than discovering new ones. The median new app earns roughly $72 per month one year after launch, and only 17.3% of new apps reach $1,000 per month within two years.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

How AI Agent Pull Requests Should Be Structured for Safe Code Review

As AI agents increasingly author code changes, reviewers need structured evidence to evaluate scope, assumptions, and risks before approving a merge. A proposed YAML-based policy framework outlines what an agent-authored pull request must include, such as the tested revision, input fixture versions, test results, and any unresolved assumptions. The policy also restricts agents from modifying sensitive files like CI workflows or infrastructure configs, and bars them from approving their own changes or bypassing merge gates. Critically, a policy file alone is insufficient — the underlying repository permissions, identity checks, and merge controls must all be aligned to enforce it. The author also highlights that reverting code does not automatically undo external side effects, making a recovery plan and a designated recovery owner essential parts of any agent-driven change.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

What Cron Jobs Are, How They Work, and Why Silent Failures Cost You

Cron is a Unix background program dating back to 1979 that runs scheduled commands, called cron jobs, based on a five-field time syntax specifying minute, hour, day, month, and weekday. Each cron job lives in a file called the crontab, and the scheduler checks every minute whether any job is due to run. A critical limitation is that cron has no memory — if a machine is off when a job is scheduled, that job is simply skipped with no automatic retry. Common pitfalls include the OR logic applied when both day-of-month and day-of-week fields are set, and hour ranges that extend further than intended, such as '9-17' running jobs until 17:45 rather than 17:00. Tools like crontab.guru, Python's croniter, and Node's cron-parser can help verify schedules before deployment, while platforms like GitHub Actions support the same cron syntax with minor differences such as defaulting to UTC time.

0 comments Read more at DEV Community
0
ProgrammingHacker News ·

Imp brings DSPy AI framework to the Erlang BEAM ecosystem

A new Elixir/Erlang package called Imp has been released as a full port of DSPy to the BEAM virtual machine. DSPy is an open-source framework designed for programming language models in a structured, programmatic way. The package is available on Hex, the package manager for the Elixir and Erlang ecosystems. This port allows BEAM-based developers to leverage DSPy's capabilities without leaving their native runtime environment. The project was shared on Hacker News, where it attracted early community attention.

0 comments Read more at Hacker News
0
ProgrammingDEV Community ·

How to migrate WooCommerce ticket data away from Event Tickets Plus safely

Developers moving away from Event Tickets Plus need to understand how ticket and attendee data is structured in the WordPress database to avoid migration errors. Each ticket is stored as a WooCommerce product, while each attendee is a separate post linked back to the event, ticket, and order via custom meta fields. Payment status is determined by WooCommerce order status rather than the attendee record itself, meaning pending or on-hold orders still generate attendee posts that should be excluded from paid counts. Sites using WooCommerce's High-Performance Order Storage keep orders in a separate database table, so migrations must query via wc_get_order() rather than standard post joins. Before switching plugins, administrators are advised to back up all data, verify paid attendee counts, and confirm that imported capacity figures prevent overselling.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

How to Authenticate Remote MCP Servers in Claude Code Using /mcp or CLI Login

Remote MCP servers such as Sentry, Linear, and Notion require browser-based authentication before they can be used in Claude Code, and appear with a 'Needs authentication' status until sign-in is completed. Users can authenticate either by running the /mcp command within an active session and selecting 'Authenticate', or by using the claude mcp login command directly from the shell. Authentication tokens are stored securely and refreshed automatically, with Claude Code triggering the sign-in flow when a server returns a 401 or 403 HTTP response. On headless or SSH environments without a display server, Claude Code prints the authorization URL instead of opening a browser, requiring the user to visit it manually and paste back the redirect URL. For servers that do not support Dynamic Client Registration, users must manually register an OAuth app and supply credentials via flags such as --client-id, --client-secret, and --callback-port.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

A2UI Architecture Lets AI Agents Describe UIs Without Writing Any Code

A2UI is a design architecture that allows language models to build user interfaces on the fly by sending structured data descriptions rather than executable code. Instead of generating JSX or HTML, the agent produces a small JSON object that names a pre-existing client-side component and specifies which values it should display. This approach keeps all executable code on the client, making the model's output validatable against a schema before anything is rendered. The architecture splits responsibilities across three distinct roles: an agent that decides what to show, a protocol that defines the wire format, and a renderer that handles visual presentation. Because components are reused and only values change, descriptions can be cached and the format can be versioned and treated as a formal contract across multiple client implementations.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

How one developer built 160 browser-only tools with no backend using Next.js

A solo developer has published the key engineering decisions behind freetoolsy.com, a collection of 160 browser-based utility tools built with Next.js 14, TypeScript, and Tailwind CSS. The site serves 399 static pre-rendered pages in English and Turkish from a single codebase, with all URL construction centralised in one file to simplify maintenance. A dynamic sitemap generated directly from route data ensures new tools appear in search indexes immediately after deployment, eliminating a recurring class of stale-sitemap bugs. The developer also built a pre-build registry consistency checker that catches mismatched tool imports in seconds, preventing mid-build failures across a growing 160-tool catalogue. Running entirely client-side with no backend, the tools function offline and require no server logs, data retention policies, or analytics infrastructure.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Lesser-Known AI Coding Tools Offer Lower Costs and Faster Onboarding in 2026

A 2026 analysis highlights that while GitHub Copilot, Tabnine, and Amazon CodeWhisperer dominate 84% of developer adoption, smaller alternatives like Cody by Sourcegraph, Autimatic, and MutableAI offer comparable or superior features at lower prices. The median monthly price for these lesser-known tools is $15, compared to $23 for Copilot, translating to an average annual saving of $96 per developer. Cody by Sourcegraph, priced at $19 per month, offers a 120,000-token context window — double that of Copilot — with private repository search included as a standard feature. Onboarding times for low-competition tools average 2.3 hours versus 3.2 hours for mainstream platforms, with MutableAI reportedly cutting onboarding time at Stripe from three days to one. Industry surveys suggest 73% of developers are overpaying for mainstream AI tools, despite the availability of focused, lower-cost alternatives with faster support cycles.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

GitHub Permanently Disables SHA-1 for HTTPS and TLS Connections from Sept 15

GitHub disabled SHA-1 support for HTTPS and TLS on github.com and its CDNs on September 15, 2026, ending years of legacy compatibility. The company had signaled the change since April 2026 and conducted an 18-hour brownout on July 14 to alert affected users before the final cutoff. Clients unable to complete a TLS handshake without SHA-1 now receive a connection error when accessing GitHub. The change affects github.com, GitHub Enterprise Cloud, and partner CDNs, but does not impact self-hosted GitHub Enterprise Server installations. Most failures have been traced to outdated infrastructure such as old curl builds, unpatched .NET Framework versions, and stale Docker base images rather than end-user browsers.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

One SQLAlchemy factory setting stops random IntegrityError failures in test suites

Developers using polyfactory with SQLAlchemy may encounter intermittent IntegrityError UNIQUE constraint failures when seeding test data, a bug that rarely appears below 50 rows but affects over 75% of runs at 100 rows. The root cause is that SQLAlchemyFactory generates integer primary keys using Faker's pyint(), which draws from only 10,000 possible values, making collisions statistically likely as row counts grow — a classic birthday problem. Testing with polyfactory 3.3.0 and SQLAlchemy 2.1.1 confirmed the failure rates match mathematical predictions across multiple table types. The fix is setting __set_primary_key__ = False in the factory class, which delegates primary key assignment to the database instead of the factory. The only trade-off is that object IDs remain None until the session is flushed, so tests that read IDs before a write must call flush explicitly.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Koobiq Design System Uses Playwright and Docker for Reliable Screenshot Testing

The Koobiq design system team has set up automated screenshot testing for its React component library using Playwright on top of Storybook. Tests capture component variants and states in a grid format across light and dark themes, reducing the number of baseline images needed while making visual regressions immediately obvious. To eliminate environment-related flakiness, all screenshots are taken and compared exclusively inside a Docker container, both locally and in CI. The configuration locks locale, time zone, and animations, and sets a zero pixel-difference threshold so any unintended style change causes an immediate test failure. Updating baselines requires no manual Docker work — a slash command on a pull request triggers CI to regenerate and commit the new reference images automatically.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Developer Earns All AWS Certifications, Shares Tips on the Toughest Ones

A software developer has achieved AWS's coveted 'Golden Jacket' by completing all available AWS certifications, a journey that began over four years ago with the Cloud Practitioner exam. The developer earned several certifications each year before deciding to complete the entire portfolio in a single year. Certifications covered a wide range, including cloud fundamentals, AI and machine learning, solutions architecture, serverless development, and data engineering. The developer noted that the Solutions Architect Associate offered the broadest technical grounding, while the Data Engineering certification was recommended as essential preparation before tackling ML and Generative AI credentials. Personal insights and difficulty ratings for each certification were shared to help others plan their own AWS learning paths.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

jev-cli Lets Developers Classify Text via Terminal Without Parsing LLM Output

A community-built open-source command-line tool called jev-cli has been released to help developers classify text without relying on parsing unpredictable LLM-generated responses. The tool interfaces with TypeSafe AI's Jev model, which returns calibrated probabilities instead of natural-language text, allowing scripts to branch on numeric outcomes. Developers can ask yes/no questions, multi-option choices, or scale-based scoring queries against any input text such as support tickets, reviews, or commit messages. Results map directly to shell exit codes, eliminating the need for fragile text parsing in automation pipelines. The project is unofficial and not affiliated with or endorsed by TypeSafe AI, and is available freely on GitHub.

0 comments Read more at DEV Community
0
ProgrammingHacker News ·

Ink & Switch Essay Argues for Malleable Software to Restore User Control

A 2025 essay published by Ink & Switch makes the case for 'malleable software,' a concept centered on giving users greater agency over the tools they use. The piece argues that modern applications have become increasingly locked down, limiting how people can adapt or customize software to their needs. The authors advocate for a design philosophy that allows end users to modify, extend, and reshape software behavior without relying on developers. The essay has surfaced on Hacker News, drawing early discussion around the future of user-driven software design.

0 comments Read more at Hacker News
0
ProgrammingDEV Community ·

Developer Builds Coding Agent That Derives Rules From Past Corrections

A developer has built a coding assistant that moves beyond simple conversation recall by deriving standing rules from past corrections — a process the author calls 'reflect'. When told once to use date-fns instead of moment.js, the agent extracted and stored that preference as a project-wide convention, applying it automatically to future unrelated requests. The system works by injecting recalled memory context directly into the LLM's system prompt, requiring no separate teaching mode from the user. The author argues that most agent memory systems stop at retrieval, while the real value lies in a reasoning layer that generalises scattered corrections into durable rules. The approach is presented as broadly applicable beyond coding assistants, addressing any agent that repeatedly makes the same mistakes due to a lack of memory generalisation.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Why CVSS Scores Alone Are Not Enough for Smart Vulnerability Management

The Common Vulnerability Scoring System (CVSS) provides a standardized measure of a vulnerability's technical severity, but it does not account for the full organizational context needed to prioritize fixes. Factors such as whether a system is internet-facing, actively exploited, or business-critical can matter more than a raw score. A vulnerability rated 9.8 on an isolated internal server may pose less immediate risk than a 7.5-rated flaw on a public-facing production system. Security teams are advised to weigh severity alongside exploitability, asset criticality, exposure, and business impact when making remediation decisions. Relying solely on CVSS rankings can cause organizations to misallocate resources and overlook higher-priority threats hiding behind lower scores.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Developer builds ad-free color-by-number game for his 8-year-old daughter

A software developer created Color Quest, a free browser-based color-by-number game, after finding that existing apps for his 8-year-old daughter were cluttered with ads, timers, and paywalls. The game lets children tap numbered squares to fill pixel art scenes featuring animals, dinosaurs, space themes, and more, with no lives, no losing progress, and gentle hints for wrong taps. Players can earn power tools like a paint brush and an explosion fill by maintaining coloring streaks. Built with TypeScript and Vite as a fully static site, the project used AI models for architecture and puzzle generation, costing the developer only a $10 domain fee. The game works offline, supports Spanish and Vietnamese, includes a photo-to-puzzle feature that keeps images on-device, and requires no account or sign-up.

0 comments Read more at DEV Community
← NewerPage 862 of 3175Older →