https://federicocalo.dev/en/blog Articles about software development, AI, machine learning, architecture and technology en Wed, 05 Aug 2026 12:00:00 GMT info@federicocalo.dev (Federico Calo) info@federicocalo.dev (Federico Calo) https://federicocalo.dev/assets/images/logo.png https://federicocalo.dev/en/blog https://federicocalo.dev/en/blog/after-gemini-cli-antigravity-gemini-api https://federicocalo.dev/en/blog/after-gemini-cli-antigravity-gemini-api Fri, 30 Oct 2026 00:00:00 +0000 Federico Calo Google Gemini for Developers Since 18 June 2026 Gemini CLI no longer serves individual accounts. What Antigravity CLI is, what gemini-api is for, and how to migrate your configuration. https://federicocalo.dev/en/blog/codex-vs-claude-code-copilot-cursor-comparison https://federicocalo.dev/en/blog/codex-vs-claude-code-copilot-cursor-comparison Thu, 29 Oct 2026 00:00:00 +0000 Federico Calo OpenAI Codex Codex against Claude Code, GitHub Copilot and Cursor: prices as listed in October 2026, real differences and a way to choose without fandom. https://federicocalo.dev/en/blog/getting-cited-by-perplexity-what-is-proven https://federicocalo.dev/en/blog/getting-cited-by-perplexity-what-is-proven Thu, 29 Oct 2026 00:00:00 +0000 Federico Calo Perplexity How an answer engine picks sources is unknown from outside: what is documented, what is conjecture and how to measure what can be measured. https://federicocalo.dev/en/blog/codex-cloud-subagents-skills-mcp https://federicocalo.dev/en/blog/codex-cloud-subagents-skills-mcp Wed, 28 Oct 2026 00:00:00 +0000 Federico Calo OpenAI Codex Delegate to Codex cloud, run subagents in parallel, write skills and connect MCP servers: what pays off, what it costs and when it does not fit. https://federicocalo.dev/en/blog/gemini-api-tiers-limits-privacy https://federicocalo.dev/en/blog/gemini-api-tiers-limits-privacy Wed, 28 Oct 2026 00:00:00 +0000 Federico Calo Google Gemini for Developers Free tier, Tiers 1-3, spend thresholds and rate limits of the Gemini API: what is published, what is not, and why the free tier suits no sensitive data. https://federicocalo.dev/en/blog/perplexitybot-and-perplexity-user-robots-txt https://federicocalo.dev/en/blog/perplexitybot-and-perplexity-user-robots-txt Wed, 28 Oct 2026 00:00:00 +0000 Federico Calo Perplexity Two agents, different rules: what robots.txt controls, what it does not, how to verify from your logs who really visits and what Cloudflare reported in 2025. https://federicocalo.dev/en/blog/codex-exec-ci-sdk-automation https://federicocalo.dev/en/blog/codex-exec-ci-sdk-automation Tue, 27 Oct 2026 00:00:00 +0000 Federico Calo OpenAI Codex How to use codex exec, JSON output, response schemas and the SDK to run Codex in CI while keeping the API key away from the code in your repo. https://federicocalo.dev/en/blog/gemini-context-thinking-caching https://federicocalo.dev/en/blog/gemini-context-thinking-caching Tue, 27 Oct 2026 00:00:00 +0000 Federico Calo Google Gemini for Developers One million tokens shared across input, thinking and output: what changes with Gemini 4 Argon, thinking_level, and when implicit caching saves money. https://federicocalo.dev/en/blog/costs-qps-and-429s-on-the-perplexity-api https://federicocalo.dev/en/blog/costs-qps-and-429s-on-the-perplexity-api Tue, 27 Oct 2026 00:00:00 +0000 Federico Calo Perplexity Usage tiers, QPS, the kinds of 429 and a spending ceiling of your own: how to estimate and control Perplexity API costs without knowing everything. https://federicocalo.dev/en/blog/codex-models-reasoning-effort-costs https://federicocalo.dev/en/blog/codex-models-reasoning-effort-costs Mon, 26 Oct 2026 00:00:00 +0000 Federico Calo OpenAI Codex Astra, Sol, Terra or Luna? The Codex models as of October 2026, plan limits, credits and how to pick a reasoning level without wasting usage. https://federicocalo.dev/en/blog/perplexity-mcp-in-claude-code-and-cursor https://federicocalo.dev/en/blog/perplexity-mcp-in-claude-code-and-cursor Mon, 26 Oct 2026 00:00:00 +0000 Federico Calo Perplexity Perplexity's official MCP server in Claude Code, Cursor and VS Code: installation, the four tools, the costs and the risk of web text in the context. https://federicocalo.dev/en/blog/gemini-function-calling-loop-modes https://federicocalo.dev/en/blog/gemini-function-calling-loop-modes Sun, 25 Oct 2026 00:00:00 +0000 Federico Calo Google Gemini for Developers How to let a Gemini model call your functions: the declaration, the function_call and function_result loop, the auto, any, none and validated modes. https://federicocalo.dev/en/blog/from-sonar-to-the-agent-api-migrating-safely https://federicocalo.dev/en/blog/from-sonar-to-the-agent-api-migrating-safely Sun, 25 Oct 2026 00:00:00 +0000 Federico Calo Perplexity Sonar has been absorbed by Perplexity's Agent API: what changes in request, response and citations, and how to migrate with a test on your own traffic. https://federicocalo.dev/en/blog/gemini-api-key-sdk-interactions https://federicocalo.dev/en/blog/gemini-api-key-sdk-interactions Sat, 24 Oct 2026 00:00:00 +0000 Federico Calo Google Gemini for Developers From API key to first response: the google-genai SDK, the new Interactions endpoint, server-side state, and what changes compared with generateContent. https://federicocalo.dev/en/blog/perplexity-search-api-for-your-rag https://federicocalo.dev/en/blog/perplexity-search-api-for-your-rag Sat, 24 Oct 2026 00:00:00 +0000 Federico Calo Perplexity Perplexity's Search API returns sources, not answers: parameters, costs, limits and a RAG pipeline that checks the citations in code. https://federicocalo.dev/en/blog/codex-agents-md-config-toml-setup https://federicocalo.dev/en/blog/codex-agents-md-config-toml-setup Fri, 23 Oct 2026 00:00:00 +0000 Federico Calo OpenAI Codex How Codex reads AGENTS.md, where config.toml lives, which file wins when they conflict, and how to keep your rules under the 32 KiB limit. https://federicocalo.dev/en/blog/gemini-for-developers-models-pricing https://federicocalo.dev/en/blog/gemini-for-developers-models-pricing Fri, 23 Oct 2026 00:00:00 +0000 Federico Calo Google Gemini for Developers Which Gemini models exist on 30 October 2026, what they cost and which to pick: Gemini 4 Argon, 3.8 Flash and Flash-Lite, with trade-offs and dated sources. https://federicocalo.dev/en/blog/perplexity-agent-api-effort-and-citations https://federicocalo.dev/en/blog/perplexity-agent-api-effort-and-citations Fri, 23 Oct 2026 00:00:00 +0000 Federico Calo Perplexity How to use Perplexity's Agent API: effort levels from fast to xhigh, how to read the sources in the response and where the real cost hides. https://federicocalo.dev/en/blog/codex-sandbox-approvals-permissions https://federicocalo.dev/en/blog/codex-sandbox-approvals-permissions Thu, 22 Oct 2026 00:00:00 +0000 Federico Calo OpenAI Codex Read-only, workspace-write or full access? How Codex combines sandbox and approvals, what each choice costs and which one to use locally and in CI. https://federicocalo.dev/en/blog/perplexity-for-developers-the-map https://federicocalo.dev/en/blog/perplexity-for-developers-the-map Thu, 22 Oct 2026 00:00:00 +0000 Federico Calo Perplexity Perplexity's five developer APIs under one key: when to pick the Agent API, the Search API or the others, what it costs, and what became of Sonar. https://federicocalo.dev/en/blog/codex-cli-install-first-task https://federicocalo.dev/en/blog/codex-cli-install-first-task Wed, 21 Oct 2026 00:00:00 +0000 Federico Calo OpenAI Codex How to install Codex CLI, sign in with ChatGPT or an API key and run your first task: /init, /permissions, /review and Git checkpoints to roll back. https://federicocalo.dev/en/blog/openai-codex-what-it-is-today-product-map https://federicocalo.dev/en/blog/openai-codex-what-it-is-today-product-map Tue, 20 Oct 2026 00:00:00 +0000 Federico Calo OpenAI Codex Codex is now a family of surfaces: CLI, IDE extension, app, web and cloud. What it is, which plans include it and when it pays off, with dated sources. https://federicocalo.dev/en/blog/formal-methods-06-06-petri-nets-structure-reachability-liveness https://federicocalo.dev/en/blog/formal-methods-06-06-petri-nets-structure-reachability-liveness Thu, 08 Oct 2026 10:20:45 +0000 Federico Calo Formal Methods Graphical-mathematical model for concurrent systems; places, transitions, markings; liveness, safety and boundedness properties; deadlock and starvation. https://federicocalo.dev/en/blog/formal-methods-03-03-finite-state-machines https://federicocalo.dev/en/blog/formal-methods-03-03-finite-state-machines Thu, 08 Oct 2026 00:00:00 +0000 Federico Calo Formal Methods Deterministic and non-deterministic automata, modeling behavior, verifying reachability and liveness properties. https://federicocalo.dev/en/blog/information-theory-03-03-source-coding-huffman-kraft https://federicocalo.dev/en/blog/information-theory-03-03-source-coding-huffman-kraft Tue, 06 Oct 2026 11:00:00 +0000 Federico Calo Information Theory Prefix codes and Kraft inequality. Huffman code construction. Shannon's first theorem: average length of any code >= H(X). Optimal compression. https://federicocalo.dev/en/blog/information-theory-06-lossy-compression-rate-distortion https://federicocalo.dev/en/blog/information-theory-06-lossy-compression-rate-distortion Tue, 06 Oct 2026 09:45:00 +0000 Federico Calo Information Theory Shannon's rate-distortion theory: compression with distortion constraint, the third theorem, and the bridge to machine learning model quantization. https://federicocalo.dev/en/blog/formal-methods-05-05-model-checking-how-works-state-explosion https://federicocalo.dev/en/blog/formal-methods-05-05-model-checking-how-works-state-explosion Tue, 06 Oct 2026 09:42:15 +0000 Federico Calo Formal Methods Exhaustive state space search; scalability challenges; state compression and abstraction techniques; counterexamples as debugging tools. https://federicocalo.dev/en/blog/formal-methods-04-04-temporal-logic-ltl-ctl-eventually-always https://federicocalo.dev/en/blog/formal-methods-04-04-temporal-logic-ltl-ctl-eventually-always Sun, 04 Oct 2026 08:15:22 +0000 Federico Calo Formal Methods Linear Temporal Logic (LTL) and branching-time temporal logic (CTL): formal semantics, temporal operators, liveness and safety properties. https://federicocalo.dev/en/blog/information-theory-02-02-mutual-information-kl-divergence https://federicocalo.dev/en/blog/information-theory-02-02-mutual-information-kl-divergence Sat, 03 Oct 2026 09:30:00 +0000 Federico Calo Information Theory Mutual information I(X;Y): how much Y reduces uncertainty about X. Kullback-Leibler divergence: measure of distance between distributions. Jensen's inequality, relationship between both concepts. https://federicocalo.dev/en/blog/formal-methods-02-02-logic-for-specification https://federicocalo.dev/en/blog/formal-methods-02-02-logic-for-specification Sat, 03 Oct 2026 00:00:00 +0000 Federico Calo Formal Methods Propositional logic, first-order logic, preconditions, postconditions, loop invariants. How to formalize behavior. https://federicocalo.dev/en/blog/information-theory-05-error-correcting-codes-hamming https://federicocalo.dev/en/blog/information-theory-05-error-correcting-codes-hamming Fri, 02 Oct 2026 10:30:00 +0000 Federico Calo Information Theory Hamming distance, linear block codes, Hamming and Singleton bounds for error correction: from theory to practice on disks and networks. https://federicocalo.dev/en/blog/information-theory-01-01-shannon-entropy https://federicocalo.dev/en/blog/information-theory-01-01-shannon-entropy Wed, 30 Sep 2026 10:15:00 +0000 Federico Calo Information Theory Definition and properties of Shannon entropy: quantitative measure of a source's uncertainty. Formula, joint and conditional entropy, chain rule. https://federicocalo.dev/en/blog/information-theory-04-noisy-channels-capacity-shannon https://federicocalo.dev/en/blog/information-theory-04-noisy-channels-capacity-shannon Tue, 29 Sep 2026 08:15:00 +0000 Federico Calo Information Theory The binary symmetric channel, channel capacity, and Shannon's second theorem: the foundation of reliable communication over real channels. https://federicocalo.dev/en/blog/formal-methods-01-01-when-to-invest https://federicocalo.dev/en/blog/formal-methods-01-01-when-to-invest Mon, 28 Sep 2026 00:00:00 +0000 Federico Calo Formal Methods Formal specification, mathematical verification, when the cost of a bug exceeds the cost of a proof. ROI in critical systems. https://federicocalo.dev/en/blog/walled-garden-legaltech-portabilita-dato https://federicocalo.dev/en/blog/walled-garden-legaltech-portabilita-dato Wed, 05 Aug 2026 12:00:00 GMT Federico Calo Data Sovereignty A law firm's documents living inside a proprietary platform raise an architectural question before a legal one: in what format do they come out, and at what cost? This article describes an on-prem architecture for legal documents — what's a verifiable fact, what isn't — and why... https://federicocalo.dev/en/blog/architettura-on-prem-dati-sanitari-baa https://federicocalo.dev/en/blog/architettura-on-prem-dati-sanitari-baa Wed, 05 Aug 2026 09:00:00 GMT Federico Calo Data Sovereignty A workload that handles healthcare data can't send that data outside the client's perimeter, and there's no third-party processor to hand it to. This article describes what an on-prem architecture does — what's a verifiable fact, what isn't — and why the question of a BAA... https://federicocalo.dev/en/blog/framework-tco-infrastruttura-on-prem https://federicocalo.dev/en/blog/framework-tco-infrastruttura-on-prem Mon, 03 Aug 2026 09:00:00 GMT Federico Calo Data Sovereignty The total cost of ownership of an infrastructure that keeps data in-house isn't a sales figure: it's a sum of line items — hardware, maintenance, in-house skills, lock-in risk, billing model. A calculation framework with the real figures measured on the production machine... https://federicocalo.dev/en/blog/ottimizzare-costi-llm https://federicocalo.dev/en/blog/ottimizzare-costi-llm Sun, 19 Jul 2026 09:00:00 GMT Federico Calo AI Cost Optimization How to reduce LLM inference costs from cloud API consumption to zero marginal cost with self-hosted Ollama on VPS. Strategies: GGUF quantization, semantic caching, batching, model selection... https://federicocalo.dev/en/blog/java-21-spring-boot https://federicocalo.dev/en/blog/java-21-spring-boot Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Clean Architecture Java 21 LTS transforms Spring Boot development with virtual threads, advanced pattern matching, immutable records and string templates. Practical guide with real examples... https://federicocalo.dev/en/blog/jpa-flyway-clean-architecture https://federicocalo.dev/en/blog/jpa-flyway-clean-architecture Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Clean Architecture How to integrate JPA/Hibernate and Flyway database migrations in the infrastructure layer of Clean Architecture Spring Boot, keeping the domain layer pure, separating JPA entities from domain models, and managing schema versioning with full traceability. https://federicocalo.dev/en/blog/testing-spring-testcontainers https://federicocalo.dev/en/blog/testing-spring-testcontainers Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Clean Architecture Complete guide to Testcontainers for Spring Boot integration testing. Real PostgreSQL and Redis containers, @DynamicPropertySource, container reuse, testing infrastructure layer in Clean Architecture, CI/CD best practices. https://federicocalo.dev/en/blog/solid-java-typescript https://federicocalo.dev/en/blog/solid-java-typescript Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Clean Architecture The 5 SOLID principles (SRP, OCP, LSP, ISP, DIP) explained with real-world examples in Java 21 and TypeScript. For each: an anti-pattern (wrong) and the correct version (right). Link SOLID to Clean Architecture. Examples from Spring Boot and Angular. https://federicocalo.dev/en/blog/domain-layer-clean-architecture https://federicocalo.dev/en/blog/domain-layer-clean-architecture Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Clean Architecture The domain layer is the heart of Clean Architecture: pure business logic, zero Spring, zero JPA. Learn how to model immutable entities with Java records, implement value objects with constructor-based validation, and protect business rules. https://federicocalo.dev/en/blog/use-case-spring-clean https://federicocalo.dev/en/blog/use-case-spring-clean Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Clean Architecture Application layer with use case (interactor), ports and DTO as Java records. Orchestration domain-infrastructure without presentation coupling. Transactions, dependency injection, anti-pattern business logic in controller. https://federicocalo.dev/en/blog/infrastructure-jpa-redis https://federicocalo.dev/en/blog/infrastructure-jpa-redis Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Clean Architecture The infrastructure layer implements the ports of domain/application through adapters: JPA for persistence, Redis for caching, WebClient for external services. How to maintain the dependency arrow toward the domain, avoid fatal couplings, and scale with concrete patterns on Spring Boot 3.4, PostgreSQL 16 and Redis Stack 7. https://federicocalo.dev/en/blog/testing-clean-architecture https://federicocalo.dev/en/blog/testing-clean-architecture Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Clean Architecture The test pyramid in Clean Architecture: pure domain unit tests (no mocks, JUnit 5 only), use case tests with mocked ports (Mockito), infrastructure integration tests with Testcontainers. Coverage per layer, what NOT to test, Java 21 + JUnit 5 examples. https://federicocalo.dev/en/blog/lightgbm-sport-prediction https://federicocalo.dev/en/blog/lightgbm-sport-prediction Wed, 15 Jul 2026 10:00:00 GMT Federico Calo AI Engineering Sports prediction pipeline with LightGBM in Python: feature engineering, gradient boosting, probabilistic calibration, evaluation with Brier score on a dataset of 2000+ matches. https://federicocalo.dev/en/blog/saas-pricing-founder-eu https://federicocalo.dev/en/blog/saas-pricing-founder-eu Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Founder Tools Guide to SaaS pricing models (flat, tiered, per-seat, usage-based, freemium), value-based pricing, price anchoring, VAT/MOSS EU compliance, reverse charge B2B, currencies. How to test prices and maximize revenue for European startups. https://federicocalo.dev/en/blog/hiring-startup-eu https://federicocalo.dev/en/blog/hiring-startup-eu Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Founder Tools How to hire in a European startup. Employment contract types, equity with vesting and cliff, cross-border employer of record, tax compliance and GDPR, technical hiring process, remote-first team building. https://federicocalo.dev/en/blog/privacy-by-design-ai https://federicocalo.dev/en/blog/privacy-by-design-ai Wed, 15 Jul 2026 10:00:00 GMT Federico Calo EU AI Act Privacy by Design & by Default (GDPR Article 25) applied to AI systems — data minimization, pseudonymization, DPIA for high-risk AI, legal basis for training, retention, data subject rights in ML pipeline, intersection with EU AI Act. With Ollama on-prem for guaranteed privacy. https://federicocalo.dev/en/blog/prompt-engineering-code https://federicocalo.dev/en/blog/prompt-engineering-code Wed, 15 Jul 2026 10:00:00 GMT Federico Calo AI Engineering Practical guide to prompt engineering techniques for reliable code generation with LLMs. Few-shot learning, chain-of-thought, role prompting, structured output, iteration and evaluation. Real examples with Ollama on-prem. https://federicocalo.dev/en/blog/angular-i18n-18-lingue https://federicocalo.dev/en/blog/angular-i18n-18-lingue Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Angular Moderno Practical guide to i18n consolidation in Angular 21 SSR: migrating from 18 runtime languages (XLF translation) to Pattern A (IT+EN compiled AOT). i18n markers, build gates, configuration, trade-off reasoning, and lessons learned. https://federicocalo.dev/en/blog/reranking-rag-cross-encoder https://federicocalo.dev/en/blog/reranking-rag-cross-encoder Wed, 15 Jul 2026 10:00:00 GMT Federico Calo RAG Engineering Two-stage retrieval for perfect relevance: bi-encoder for fast retrieval, cross-encoder for intelligent re-ranking of top-k. Quality metrics (nDCG, MRR, recall@k), latency-quality trade-offs, and real-world implementation with sentence-transformers + RediSearch on local Ollama. https://federicocalo.dev/en/blog/fastapi-ml-pydantic https://federicocalo.dev/en/blog/fastapi-ml-pydantic Wed, 15 Jul 2026 10:00:00 GMT Federico Calo AI Engineering Build ML microservices with FastAPI (Python 3.12) and Pydantic v2: request/response validation, async/await, lifespan for model loading at boot, dependency injection, health checks, graceful degradation with on-prem Ollama. https://federicocalo.dev/en/blog/provider-deployer-eu-ai-act https://federicocalo.dev/en/blog/provider-deployer-eu-ai-act Wed, 15 Jul 2026 10:00:00 GMT Federico Calo EU AI Act The distinction between provider and deployer in EU Regulation 2024/1689 is critical: each role has precise responsibilities for risk management, technical documentation, logging, and transparency. Guide to 4 risk levels, high-risk system obligations, timeline, and operational checklist. https://federicocalo.dev/en/blog/conformity-assessment-eu-ai https://federicocalo.dev/en/blog/conformity-assessment-eu-ai Wed, 15 Jul 2026 10:00:00 GMT Federico Calo EU AI Act Operational guide to conformity assessment for high-risk AI systems. Internal evaluation, Notified Body involvement, EU Declaration of Conformity, CE marking, NEBA registration, post-market monitoring. Checklists and templates. https://federicocalo.dev/en/blog/llm-italiani-2024 https://federicocalo.dev/en/blog/llm-italiani-2024 Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Italian NLP In 2024, the first Large Language Models fully trained or significantly adapted for Italian emerge: Minerva (Sapienza NLP), LLaMAntino, Camoscio, and Fauno represent a breakthrough in Italian NLP. Features, datasets, benchmarks... https://federicocalo.dev/en/blog/landscape-llm-italiani https://federicocalo.dev/en/blog/landscape-llm-italiani Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Italian NLP 2025 overview of the Italian language model ecosystem — new models, Italian benchmarks, how to select the right model for your use case, and local deployment via Ollama on-prem. https://federicocalo.dev/en/blog/nlp-italiano-spacy https://federicocalo.dev/en/blog/nlp-italiano-spacy Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Italian NLP Complete NLP pipeline for Italian: tokenization, POS tagging, lemmatization, NER and dependency parsing with spaCy and Hugging Face transformers. Italian models (BERT, UmBERTo), language-specific challenges, sentiment analysis and text classification with Python 3.12 examples. https://federicocalo.dev/en/blog/fine-tuning-llm-italiano https://federicocalo.dev/en/blog/fine-tuning-llm-italiano Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Italian NLP How to fine-tune an open-source model for Italian using LoRA on consumer GPUs, prepare quality datasets, evaluate performance with Italian benchmarks, and deploy locally via Ollama... https://federicocalo.dev/en/blog/rag-italiano-embedding https://federicocalo.dev/en/blog/rag-italiano-embedding Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Italian NLP Building an effective RAG pipeline for Italian content: choosing multilingual embeddings, semantic chunking, RediSearch HNSW, retrieval + re-ranking, generation with Ollama on-prem, quality evaluation. https://federicocalo.dev/en/blog/crewai-produzione-multi-agent https://federicocalo.dev/en/blog/crewai-produzione-multi-agent Wed, 15 Jul 2026 10:00:00 GMT Federico Calo Multi-Agent AI Systems CrewAI orchestrates specialized AI agents with clear tasks, integrated tools, and persistent memory. Complete guide on production architecture, Ollama on-prem, PostgreSQL, observability and scaling. https://federicocalo.dev/en/blog/inp-cwv-angular-ssr-59-to-88-four-fixes https://federicocalo.dev/en/blog/inp-cwv-angular-ssr-59-to-88-four-fixes Tue, 14 Jul 2026 11:05:00 GMT Federico Calo Original Research Four real interventions that took an Angular SSR home page from a mobile PageSpeed score of 59 to 88: precompute instead of fan-out, PlatformLocation instead of REQUEST at prerender, read/write split with afterNextRender, and a prerender-compatible host-based CSP. With measured numbers and a public repository. https://federicocalo.dev/en/blog/projecthub-poker-retro-be-fe-slice-schema-real-time https://federicocalo.dev/en/blog/projecthub-poker-retro-be-fe-slice-schema-real-time Tue, 14 Jul 2026 11:00:00 GMT Federico Calo Architecture Real-time collaboration is a schema problem, not a WebSocket problem. I document the ProjectHub Poker/Retro vertical (V184/V185) as a real end-to-end slice: eleven Postgres tables, a one-directional phase FSM, STOMP broadcast, server-side anonymity, and a dot-voting budget, with the trade-offs declared at every fork. No code changes. https://federicocalo.dev/en/blog/production-postmortems-flyway-jar-deploy-duplicate-beans https://federicocalo.dev/en/blog/production-postmortems-flyway-jar-deploy-duplicate-beans Tue, 14 Jul 2026 10:50:00 GMT Federico Calo Platform Engineering A comment in an applied Flyway migration breaks the checksum. A non-atomic cp during deploy causes an intermittent NoClassDefFoundError. Two beans with the same simple name crash the boot. Three real incidents, one root cause: violated immutability. https://federicocalo.dev/en/blog/multi-bucket-rate-limiting-bucket4j-postmortem https://federicocalo.dev/en/blog/multi-bucket-rate-limiting-bucket4j-postmortem Tue, 14 Jul 2026 10:45:00 GMT Federico Calo DevSecOps Four separate Bucket4j buckets (API, AI, report download, token) versus one shared number. Routing to the right bucket is where real bugs are born: the postmortem of the CSP-report flood that blocked email verification with 429s. https://federicocalo.dev/en/blog/archunit-clean-architecture-4-layer-diagram-lie https://federicocalo.dev/en/blog/archunit-clean-architecture-4-layer-diagram-lie Tue, 14 Jul 2026 10:40:00 GMT Federico Calo Architettura Software The backend's ArchUnit gate was running zero executed tests, a total false green. How I found it, what I froze with FreezingArchRule (30+4 historical violations), and why a 4-layer diagram without automated enforcement is not an architecture. https://federicocalo.dev/en/blog/production-postmortems-flyway-archunit-jar-race-duplicate-beans https://federicocalo.dev/en/blog/production-postmortems-flyway-archunit-jar-race-duplicate-beans Tue, 14 Jul 2026 10:10:58 GMT Federico Calo Original Research Four real production incidents that all started from a fully green build: a Flyway crash from a removed comment, an ArchUnit gate dead for months, a JAR deploy race condition, a duplicate Spring bean. Root cause, fix, and a reproducible public repository for each. https://federicocalo.dev/en/blog/how-i-build-in-4-hours-gsd-agentic-workflow https://federicocalo.dev/en/blog/how-i-build-in-4-hours-gsd-agentic-workflow Tue, 14 Jul 2026 10:00:00 GMT Federico Calo Claude Code Workflow Claude Code Workflow series. How I built a custom booking module in a little over four hours with GSD and specialized agents: the real per-phase time datum, the trade-offs, and why AI is accompaniment and never the director. Portfolio AI provider: Ollama on-prem. https://federicocalo.dev/en/blog/onpush-signals-default-not-exception-angular-21 https://federicocalo.dev/en/blog/onpush-signals-default-not-exception-angular-21 Tue, 14 Jul 2026 09:52:54 GMT Federico Calo Angular Moderno 1215 out of 1219 components in this portfolio use ChangeDetectionStrategy.OnPush. Why "turn it on when I need it" is a comfortable lie, what immutability discipline actually costs, and when Default is the better call instead. https://federicocalo.dev/en/blog/strict-mode-not-enough-any-typescript-monorepo https://federicocalo.dev/en/blog/strict-mode-not-enough-any-typescript-monorepo Tue, 14 Jul 2026 09:52:54 GMT Federico Calo TypeScript Avanzato strict, strictTemplates, and strictInjectionParameters all on, yet 54 ': any' and 46 'as any' survive across 3884 files. Why strict mode only eliminates accidental errors, not deliberate escape hatches, and what to actually do about the remaining anys. https://federicocalo.dev/en/blog/prerender-vs-server-on-demand-angular-blog-articles https://federicocalo.dev/en/blog/prerender-vs-server-on-demand-angular-blog-articles Tue, 14 Jul 2026 09:52:54 GMT Federico Calo Angular Moderno app.routes.server.ts assigns Prerender to 50 route configurations and Server to 134, deliberately including every article page. Why "prerender everything" stops being safe advice past a certain scale, and the real trade-off between TTFB and build stability. https://federicocalo.dev/en/blog/clean-architecture-spring-boot-template-archunit-gate-that-runs https://federicocalo.dev/en/blog/clean-architecture-spring-boot-template-archunit-gate-that-runs Tue, 14 Jul 2026 09:30:00 GMT Federico Calo Original Research Most Clean Architecture templates ship a green-for-nothing ArchUnit gate that runs zero tests. I extracted a Spring Boot 3.4 template from my production backend, sanitized by hand (no git filter-repo), with JWT, Flyway, and an ArchUnit gate that genuinely executes four rules. Public repo, green build on JDK 21. https://federicocalo.dev/en/blog/side-project-case-study-italiani https://federicocalo.dev/en/blog/side-project-case-study-italiani Tue, 14 Jul 2026 09:30:00 GMT Federico Calo Indie Maker Italia Analysis of 5 Italian side projects launched in 2024-2025: strategies, mistakes, real costs, and ROI. How indie makers from Turin, Milan, and Rome transformed an idea into 5-6 figure revenue. https://federicocalo.dev/en/blog/ai-on-prem-vs-cloud-privacy-cost-data-sovereignty https://federicocalo.dev/en/blog/ai-on-prem-vs-cloud-privacy-cost-data-sovereignty Tue, 14 Jul 2026 09:24:30 GMT Federico Calo Platform Engineering A B2B comparison with real TCO: on-prem AI with Ollama on a VPS costs zero extra per month, metered cloud bills per token. An honest trade-off on cost, latency, model ceiling, and data sovereignty for decision-makers. https://federicocalo.dev/en/blog/geo-retrofit-benchmark-741-to-0-structural-errors https://federicocalo.dev/en/blog/geo-retrofit-benchmark-741-to-0-structural-errors Tue, 14 Jul 2026 09:14:16 GMT Federico Calo Original Research Most GEO advice is cosmetic. An additive engine (injects, does not rewrite) took 741 structural errors to zero across 194 bilingual articles: 386 missing-FAQ errors closed automatically, 310 missing-citation errors closed only after manual review. https://federicocalo.dev/en/blog/sport-intelligence-benchmark-97k-matches https://federicocalo.dev/en/blog/sport-intelligence-benchmark-97k-matches Tue, 14 Jul 2026 08:51:44 GMT Federico Calo Original Research Why accuracy lies about football predictions and the Brier score does not. The portfolio's Sport Intelligence engine publishes a Brier score of 0.5783 across 97,000 historical matches, with a V2 LightGBM ensemble, isotonic calibration, and a reproducible public repo. https://federicocalo.dev/en/blog/funding-oss-eu-nlnet-sovereign https://federicocalo.dev/en/blog/funding-oss-eu-nlnet-sovereign Sat, 20 Jun 2026 08:00:00 GMT Federico Calo OSS Sustainability How to access European funding for open source: NLnet Foundation, Sovereign Tech Fund, and NGI programs. Complete guide to applications, budgeting, and winning case studies from the OSS community. https://federicocalo.dev/en/blog/open-core-business-case-studies https://federicocalo.dev/en/blog/open-core-business-case-studies Sat, 20 Jun 2026 04:41:27 GMT Federico Calo OSS Sustainability The Open Core model through three cases: ESLint, Supabase and HashiCorp. Technical anatomy, what stays open, what turns paid, and pricing architecture. https://federicocalo.dev/en/blog/maintainer-burnout-prevenzione https://federicocalo.dev/en/blog/maintainer-burnout-prevenzione Sat, 20 Jun 2026 01:22:55 GMT Federico Calo OSS Sustainability Article from the Open Source Sustainability 2027 series. Maintainer burnout is a silent crisis that strikes open source project stewards. Concrete strategies through automation, community management, and economic sustainability to prevent exhaustion. https://federicocalo.dev/en/blog/oss-commercial-agpl-bsl-elastic https://federicocalo.dev/en/blog/oss-commercial-agpl-bsl-elastic Fri, 19 Jun 2026 22:04:22 GMT Federico Calo OSS Sustainability When you publish an open source project, you must choose how to protect yourself from those who profit without contributing. AGPL, BSL, and Elastic License offer three different strategies for commercial copyleft. https://federicocalo.dev/en/blog/sponsor-models-github-open-collective https://federicocalo.dev/en/blog/sponsor-models-github-open-collective Fri, 19 Jun 2026 18:45:50 GMT Federico Calo OSS Sustainability Discover the two leading platforms for monetizing open source projects: GitHub Sponsors vs Open Collective. Analysis of pros/cons, revenue models, case studies and the right strategy for your project. https://federicocalo.dev/en/blog/playwright-a11y-axe-core https://federicocalo.dev/en/blog/playwright-a11y-axe-core Fri, 19 Jun 2026 15:27:17 GMT Federico Calo Playwright Components Article from the Playwright Component Testing Avanzato series. Published November 26, 2027. https://federicocalo.dev/en/blog/playwright-mock-network-interception https://federicocalo.dev/en/blog/playwright-mock-network-interception Fri, 19 Jun 2026 12:08:45 GMT Federico Calo Playwright Components Advanced mocking and network interception in Playwright: routes and network events, mock patterns, dynamic interception, context versus page… https://federicocalo.dev/en/blog/test-parallelism-playwright-shard https://federicocalo.dev/en/blog/test-parallelism-playwright-shard Fri, 19 Jun 2026 08:50:13 GMT Federico Calo Playwright Components Test parallelism in Playwright: how workers and shards work, handling shared state, hybrid strategies, configuration, and troubleshooting. https://federicocalo.dev/en/blog/visual-regression-playwright-percy https://federicocalo.dev/en/blog/visual-regression-playwright-percy Fri, 19 Jun 2026 05:31:40 GMT Federico Calo Playwright Components Automate visual testing of your components with Percy and Playwright. Learn to capture snapshots, manage baselines, and integrate visual regression into CI/CD. https://federicocalo.dev/en/blog/playwright-component-testing-setup https://federicocalo.dev/en/blog/playwright-component-testing-setup Fri, 19 Jun 2026 02:13:08 GMT Federico Calo Playwright Components Article in the Playwright Component Testing Advanced series. Complete Playwright setup for component testing in Angular with fixtures, configuration and best practices. https://federicocalo.dev/en/blog/bun-production-docker-kubernetes https://federicocalo.dev/en/blog/bun-production-docker-kubernetes Thu, 18 Jun 2026 22:54:35 GMT Federico Calo Bun Runtime 2027 Bun in production with Docker and Kubernetes: a multi-stage Dockerfile, building and pushing the image, YAML manifests, health checks, and rolling updates. https://federicocalo.dev/en/blog/bun-hono-stack-edge https://federicocalo.dev/en/blog/bun-hono-stack-edge Thu, 18 Jun 2026 19:36:03 GMT Federico Calo Bun Runtime 2027 A minimal edge REST API with Bun and Hono: setup, JWT authentication, database and cache integration, and why this stack stays lightweight. https://federicocalo.dev/en/blog/bun-nativo-sqlite-redis https://federicocalo.dev/en/blog/bun-nativo-sqlite-redis Thu, 18 Jun 2026 16:17:31 GMT Federico Calo Bun Runtime 2027 Bun's built-in SQLite and Redis clients: native usage, a cache pattern with SQLite fallback, a benchmark against Node.js, and production best practices. https://federicocalo.dev/en/blog/migrazione-node-bun-gotchas https://federicocalo.dev/en/blog/migrazione-node-bun-gotchas Thu, 18 Jun 2026 12:58:58 GMT Federico Calo Bun Runtime 2027 Migrating from Node.js to Bun without surprises: package manager and lockfiles, CommonJS versus ES Modules, divergent APIs, async behavior, and native addons. https://federicocalo.dev/en/blog/bun-vs-node-deno-2027 https://federicocalo.dev/en/blog/bun-vs-node-deno-2027 Thu, 18 Jun 2026 09:40:26 GMT Federico Calo Bun Runtime 2027 Node.js, Bun and Deno compared: startup time, HTTP throughput, memory use, and TypeScript compilation, plus how to choose a runtime for your project. https://federicocalo.dev/en/blog/tauri-mobile-android-ios https://federicocalo.dev/en/blog/tauri-mobile-android-ios Thu, 18 Jun 2026 06:21:53 GMT Federico Calo Tauri Rust Desktop Tauri on Android and iOS: the state of mobile support, reusing desktop Rust code, architecture, roadmap, dev environment setup, and native API access. https://federicocalo.dev/en/blog/cross-platform-deploy-tauri https://federicocalo.dev/en/blog/cross-platform-deploy-tauri Thu, 18 Jun 2026 03:03:21 GMT Federico Calo Tauri Rust Desktop Shipping a Tauri app on macOS, Windows and Linux: tauri.conf.json, code signing, macOS notarization, and installers and bundles for each system. https://federicocalo.dev/en/blog/tauri-plugins-ecosystem-2027 https://federicocalo.dev/en/blog/tauri-plugins-ecosystem-2027 Wed, 17 Jun 2026 23:44:49 GMT Federico Calo Tauri Rust Desktop The Tauri 2 plugin ecosystem: the IPC model, official and community plugins, how to write a custom plugin, and notes on performance and security. https://federicocalo.dev/en/blog/rust-frontend-integration-tipi https://federicocalo.dev/en/blog/rust-frontend-integration-tipi Wed, 17 Jun 2026 20:26:16 GMT Federico Calo Tauri Rust Desktop Article from the Tauri 2 + Rust Desktop Apps series. Published 17 June 2026. https://federicocalo.dev/en/blog/tauri-2-vs-electron-bundle https://federicocalo.dev/en/blog/tauri-2-vs-electron-bundle Wed, 17 Jun 2026 17:07:44 GMT Federico Calo Tauri Rust Desktop Tauri 2 versus Electron: why bundle size matters, the Rust backend plus web frontend architecture, how to measure it, and a performance comparison. https://federicocalo.dev/en/blog/cost-optimization-vector-db-tier https://federicocalo.dev/en/blog/cost-optimization-vector-db-tier Wed, 17 Jun 2026 13:49:11 GMT Federico Calo Vector DB Decision Cutting vector database costs with tiering: cold storage on PostgreSQL, a warm index on Redis Stack, a hot cloud index, and an in-memory cache layer. https://federicocalo.dev/en/blog/vector-db-scaling-100m https://federicocalo.dev/en/blog/vector-db-scaling-100m Wed, 17 Jun 2026 10:30:39 GMT Federico Calo Vector DB Decision Modern Vector Databases can scale beyond 100 million vectors while maintaining sub-100ms query latency using HNSW hierarchical indexing, geographic sharding, quantized compression, and batch processing. https://federicocalo.dev/en/blog/hybrid-search-bm25-dense https://federicocalo.dev/en/blog/hybrid-search-bm25-dense Wed, 17 Jun 2026 07:12:07 GMT Federico Calo Vector DB Decision Hybrid search combines BM25 (lexical search) and dense embeddings (semantic search) for the best of both worlds. Discover how to implement it with RediSearch and Ollama, practical trade-offs, and real-world use cases.