标签：#ev

Book: AI That Ships The series: AI in TypeScript — 5 books, from your first LLM call to agents in production — all five here My project: Hermes IDE | GitHub — an IDE for developers who ship with Claude Code and other AI coding tools Me: xgabriel.com | GitHub A new model lands. Someone on your team opens a pull request that changes one string in one config file, the model id. The diff is green in five minutes. Evals look fine, maybe a point better on the suite you happen to have. It ships. Three weeks later the invoice arrives and it is a different shape than the one before it. Nobody wrote a bad loop. Nobody shipped a prompt-injection. The system does exactly what it did last month. It just costs more to do it, because a one-line diff moved every request from $2 and $10 per million tokens to $10 and $50. OpenAI announced GPT-6 Astra on 3 September 2026. OpenAI calls it the most capable model it has shipped. That is the company's claim and I am not going to argue with it. But "most capable model available" and "the model your service should call by default" are two different questions, and the distance between them shows up on your infrastructure bill. What the launch numbers say The API list price at launch, per OpenAI: Standard tier: $10 per 1M input tokens, $50 per 1M output tokens Fast tier: $20 per 1M input tokens, $100 per 1M output tokens Astra takes text and image input and returns text only, with a 1M token context window. It went first to a limited set of organisations under OpenAI's Daybreak Access programme, with wider access to the paid ChatGPT tiers and the API announced as planned for the days after launch. It is also listed on AWS Bedrock and Microsoft Azure. Now the third-party read. Artificial Analysis runs its own evaluations independently of the vendors. On its Intelligence Index, Astra scores 60 , which puts it #14 of the 202 models the site tracks. Its cost per Intelligence Index task comes out at $0.96 . The median model in that set scores 36 a

🛡️ fetch-sentinel: El cortafuegos local (CPU-Only) para blindar la ventana de contexto de tus agentes de IA frente a inyecciones indirectas de prompts. Internet es hostil por defecto. Tu agente ya no tiene por qué estar expuesto. fetch-sentinel es un guardia estructural en el punto de entrada cuando un agente autonomo hace fetch de contenido web arbitrario. Su trabajo es decidir, antes de que el contenido externo entre al contexto del LLM, que partes son dato y que partes son instruccion. Este post NO presenta fetch-sentinel como un producto listo para produccion. Lo presenta como un repositorio alfa con cuatro capas obligatorias implementadas y verificadas localmente con 161 tests, y con dos KI conocidos abiertos (KI-10, KI-11) que una auditoria independiente identifico en la segunda ronda de revision y que requieren un refactor mayor para cerrarse. El problema: inyeccion indirecta de prompts via contenido fetched Cuando un agente LLM navega la web por su cuenta, el contenido fetched es input no confiable. Un atacante puede inyectar instrucciones en paginas que el agente va a leer como si fueran parte del prompt del sistema: Texto invisible en comentarios HTML, atributos alt, metadata. Codepoints Unicode ofuscados (TAG block, ZWSP, BIDI override) que sobreviven a la mayoria de los pipelines de sanitizacion. Manipulacion semantica sin instruccion explicita: propaganda o "hechos" seleccionados empaquetados como resumen. Exfiltracion en cadena: si el agente tiene acceso a shell, email o API keys, una inyeccion exitosa escala a accion real no autorizada. Los firewalls semanticos no resuelven esto (intentar defenderse contra manipulacion semantica convierte el componente en algo que no funciona). Lo que resuelve el problema es defender el punto de entrada. Que hace fetch-sentinel Cuatro capas obligatorias: Capa Modulo Que hace 1 - Fetch aislado core/fetcher.py Extraccion readability sobre html.parser (stdlib), descarta <script> , <style> , <iframe> , <noscript> , <objec

MLH AI Roadshow Bengaluru Hey Everyone! I am Sanidhya Goel, a new MLH Hackathon Community...

Your CircleCI pipeline is green. Every job passes. And yet your app is running against a dependency version that hasn't shipped in a month — nobody committed it, nobody bumped it, it just quietly showed up in production. If you've chased a bug like this, the culprit is almost never your code. It's your cache key. This is a five-minute read and a fifteen-minute fix. Quick Win Friday, deployed to your .circleci/config.yml . The failure mode CircleCI's dependency caching works on a simple contract: you compute a key from something that changes when your dependencies change (usually a lockfile checksum), and you save/restore a cache tied to that key. The contract breaks in three specific, extremely common ways: You checksum the wrong file. {{ checksum "package.json" }} looks reasonable until someone bumps a transitive dependency via package-lock.json without touching package.json . The checksum doesn't move. CircleCI happily hands back last week's node_modules . restore_keys does prefix matching, and people think it does exact matching. CircleCI tries your primary key first, then falls through restore_keys in order, and the first one is a prefix match against existing cache entries — not "give me the newest exact match." If your restore_keys list is too coarse (e.g. just v1-deps- ), you can restore a cache built from a completely different branch, with a completely different lockfile, and the job won't fail. It'll just quietly install nothing (cache hit, npm ci sees the modules are "there") or run against the wrong versions. There's no version escape hatch. When you inevitably need to force everyone's cache to invalidate — a corrupted cache entry, a package manager migration, a lockfile format change — there's no cheap way to do it, because the key format was never designed with a manual buster in mind. Each of these fails silently. No red X. No error in the logs. Just a build that ran with stale state, and a bug report three days later that nobody can reproduce locally

zarazhangrui/follow-builders is gaining attention for a simple reason: it focuses on the people building AI systems, not just the influencers discussing them. With 84 new stars today, the project is positioned as an AI builders digest that monitors notable creators across X and YouTube podcasts, then remixes their ideas into shorter, easier-to-scan summaries. That workflow addresses a real productivity problem. AI research and engineering conversations are scattered across long videos, fast-moving social feeds, and repeated announcements. A focused digest can reduce the time spent collecting links while preserving the practical signal: architectural decisions, implementation lessons, tools, and emerging patterns. A sensible first step is to inspect the repository locally before deciding how deeply it fits your workflow: git clone https://github.com/zarazhangrui/follow-builders.git cd follow-builders # Inspect the setup instructions and available scripts ls -la find . -maxdepth 2 -type f | sort | head -80 For an AI-assisted workflow, I would pair the project with a small review loop: Collect the generated digest. Extract claims, links, and mentioned tools. Open the original source before acting on important technical advice. Save durable findings in a project notes file or knowledge base. This keeps summaries useful without treating them as authoritative research. It also makes the tool a good companion for developers using Cursor or another AI IDE: the digest supplies discovery, while the IDE helps turn validated ideas into experiments and code. Before production use, watch for two trade-offs: Summary fidelity: compressed content can lose context, caveats, or disagreements from the original conversation. Source coverage: ranking “top builders” may introduce selection bias, so important perspectives can be missed. The strongest use case is not replacing primary sources. It is building a high-signal starting queue for developers who want to follow AI progress without

Hello, I'm Shrijith Venkatramana, and I'm building LiveReview — a blast-radius aware AI code review built for your business-critical systems. Star us to help devs discover the project, give it a try, and share your feedback to help improve the product. A modern LLM can contain hundreds of billions of parameters, run on extremely expensive accelerators, and still spend most of its inference time doing something that looks embarrassingly sequential: token 1 -> token 2 -> token 3 -> token 4 -> token 5 -> ... That is the awkward part of autoregressive generation. The model may process a whole prompt in parallel during the initial prefill, but once generation starts, the next token depends on the previous token. So generating 100 tokens looks conceptually like running the model 100 times. And for many serving workloads, that is exactly where the money goes. A family of techniques tries to break this bottleneck by asking a deceptively simple question: What if the model could predict several future tokens at once, then verify them in parallel? That idea leads to speculative decoding, Medusa-style multiple decoding heads, and the broader multi-token prediction approach used during training. The interesting part is that these are not merely "optimization tricks." They change the computational structure of decoding. This article develops that idea from first principles and then gets into the engineering details. 1. The problem: your GPU is doing an expensive sequential loop Consider ordinary autoregressive decoding. Given a prompt: The capital of France is the model predicts: Paris Then it feeds the new sequence back through the model: The capital of France is Paris and predicts the next token. Then again: The capital of France is Paris . and so on. Formally, the model factorizes the probability of a sequence as: P(x1, x2, ..., xT) = product over t of P(xt | x1, ..., x(t-1)) That conditional dependence is what makes language modeling so useful. It is also what makes decoding

I ship Codenames AI, a web game. Each game mode keeps its own save in localStorage. Reload the tab,...

Originally published at nlocoding.com 92%of IT leaders say legacy systems slow digital transformation (IBM, 2026) Every minute, a bank somewhere spends $1,200 just keeping 1970s code alive. Not replacing it, just making sure it doesn’t explode. A senior developer at Citi told McKinsey in January 2026: “We spend 53% of our engineering budget patching COBOL.” Legacy code isn’t a quirky artifact anymore. It’s a financial anchor chained to your cloud ambitions... Why AI Code Tools for Legacy System Modernization Matter in 2026 AI code tools have redefined how companies approach system upgrades. In 2026, 61% of modernization projects fail due to manual errors or missed dependencies (Gartner, 2026). You can’t afford human error when one typo in ancient assembler code can cost $500,000 in downtime. The rise of generative AI for code refactoring is the only thing standing between you and a multi-million dollar rewrite. AI Code Tools Are Slashing Modernization Timelines by 63% AI code tools for legacy system modernization have cut modernization project timelines by 63% on average (Accenture, 2026). Manual migration can take 18 months—AI-powered tools like IBM watsonx Code Assistant and Google Gemini Advanced do it in under 7 months. This isn’t a hypothetical. Banco do Brasil migrated 2.8 million lines of COBOL to Java in 2025 with Cognizant’s AI tool; downtime: 14 hours. Average cost per line dropped from $3.60 (human) to $1.15 (AI-assisted). 💡 Pro Tip: Start with small pilot modules (1000-5000 lines). Measure defect rates before scaling. AI-Assisted Code Understanding Reduces Failure Rates Code comprehension is the single biggest risk in legacy system modernization. 47% of failures in 2026 were due to “unknown dependencies” (Forrester, 2026). AI code tools now map data flows, detect dead code, and generate architectural diagrams from raw source. Microsoft’s Copilot for Azure can parse 1.5 million lines in two days and flag 96% of “code rot” blocks. One insurance company in

OpenHabitTracker is a free, open source habit tracker that also holds your notes and tasks. It runs on Windows, Linux, macOS, iOS, Android and in a browser, with no ads and no account. Filtering and sorting habits A habit here is not measured by a streak. It is measured by how much of its interval has gone by. A habit you want to do every ten days, two days late, is at 120%. A habit you want to do every four days, also two days late, is at 150%. Because that is a number, you can filter on a range of it: only habits above 50%, only habits below 150%, or only the ones in between. That last one is everything neither freshly done nor badly overdue. You can sort by it as well, and the sort takes the repeat count into account, so a habit done three times a day and one done weekly are compared against each other rather than the daily ones always sitting on top. There are other ways to sort habits by time: how long you want between repeats, how long it has actually been averaging, how long since the last one, and how much time you have spent on it in total or per completion. Notes and tasks sort by the plain things, category, priority and title, and tasks also by their planned date and duration. Each of the three keeps its own sort order. Filtering by date Tasks have a planned date. Tasks and habits have the dates they were completed on. Both are filtered separately, each before, on, after or not on a date you pick. A filter can also take a number of days from today instead of a date. Minus seven to plus seven is the week either side of now, and it still means that in a month, because the days are counted at the moment you look rather than the moment you set it. Showing what was not done The completed-date filter has a switch next to it. Turned on, the range shows what you finished in those days. Turned off, the same range shows what you did not. Searching notes, tasks and habits Searching a note searches the whole note, not just its title. Searching tasks and habits search

We're back with another DEV Weekend Challenge, a short bite-sized challenge planned to fit into your...

Refactoring Safely: A Step-by-Step Guide We all know that feeling: a function that's 200 lines long, a class that does too many things, or a variable named data2 . Refactoring is the cure, but doing it recklessly can break your app and your confidence. Here's how I approach refactoring safely, step by step. 1. Start with a Safety Net Before touching any code, make sure you have tests. If your project lacks tests, write a few key ones first. Focus on the behavior you're about to change. The goal is to have a safety net that tells you when you've broken something. # example test for a function we'll refactor import unittest from mymodule import calculate_total class TestCalculateTotal ( unittest . TestCase ): def test_with_discount ( self ): self . assertEqual ( calculate_total ( 100 , discount = 0.1 ), 90 ) If tests aren't feasible, at least have a manual checklist. But automated tests are worth the effort. 2. Make Small, Atomic Changes Don't try to refactor everything at once. Pick one logical change. For instance, extract a method or rename a variable. Each change should be small enough that if it breaks, you know exactly what caused it. // before function processOrder ( order ) { const total = order . items . reduce (( sum , item ) => sum + item . price , 0 ); const tax = total * 0.08 ; const final = total + tax ; return final ; } // after step 1: extract tax calculation function processOrder ( order ) { const total = order . items . reduce (( sum , item ) => sum + item . price , 0 ); const final = total + calculateTax ( total ); return final ; } function calculateTax ( amount ) { return amount * 0.08 ; } Run your tests after each tiny step. If they pass, move on. If they fail, you know the last change caused it. 3. Use Your IDE's Refactoring Tools Modern IDEs can rename variables, extract methods, and change signatures safely. They update all references automatically. This reduces human error. For example, in VS Code, right-click a function and choose "Extract to

Using Frore’s air-cooling technology, Lenovo’s Project AeroBlade Concept is an ultralightweight laptop with efficient heat management.

A tool claiming "up to 15x faster than cdk deploy" showed up in my feed a while back. Drop-in replacement, it said: keep your CDK app exactly as it is, just swap cdk deploy for cdkd deploy . I've learned to be skeptical of "Nx faster" claims. So I actually deployed something real to AWS with both tools and timed it. Short version: it really is that fast. What cdkd actually is cdkd deploys an existing AWS CDK app without going through CloudFormation. It calls the AWS SDK directly instead. It's built by go-to-k (Kenta Goto), an AWS DevTools Hero and CDK top contributor who also maintains cls3 (a fast S3 bucket emptier) and delstack (for cleaning up stuck CloudFormation/CDK stacks) — tools that quietly fix the annoying parts of working with AWS. cdkd feels like the biggest one yet, and I mean that as a compliment grounded in actually using it, not a throwaway one. The mechanism is straightforward. cdkd runs the exact same CDK synth step as the CDK CLI, producing the same CloudFormation template. What changes is everything after that: instead of handing the template to CloudFormation, cdkd's own engine reads the resource dependency graph ( Ref , Fn::GetAtt ), builds a DAG, and fires AWS SDK / Cloud Control API calls directly, in parallel, as soon as each resource's dependencies are satisfied. Worth saying up front: cdkd calls itself not production-ready, dev/test only. This isn't a "replace CloudFormation in prod" pitch. I actually ran both, on real AWS cdkd's own README backs up the 15x number with a VPC + Lambda + SQS + CloudFront benchmark. So I wrote that same stack as a CDK app and deployed it twice — DeployRaceCfn via cdk deploy , DeployRaceCdkd via cdkd deploy — to the same AWS account, same region (ap-northeast-1). The stack: VPC (2 AZ + NAT Gateway) with a Lambda inside it, fronted by a Function URL CloudFront, origin set to that Function URL SQS + EventSourceMapping + a consumer Lambda First attempt failed. The account had hit its VPC limit (five, the default)

A green required check is not a merge decision. It is a signal that one job graph finished without a red X. If a pre-push hook was skipped, or a snapshot fixture was regenerated without a reason, you can still ship a lie. This article walks through a merge packet: a small JSON artifact your CI publishes next to the check. The packet records hook results, fixture drift, and required-job status. A model may write the eight-line brief. It does not get a vote. Why green still lies CI dashboards collapse many facts into one glyph. You see green. You click merge. You miss three common failures. First, someone pushed with --no-verify and skipped the hook that keeps fixture hashes honest. Second, a test helper rewrote golden files because a serializer added a field. Third, a retry job went green on the second attempt and nobody recorded that the first attempt failed. You do not need a platform rewrite to catch this. You need a contract the merge button cannot ignore. Cheap code generation makes the second failure more common. When it is easy to regenerate tests, it is easy to regenerate the fixtures those tests pin. The pin becomes a moving target. Treat unexplained fixture diffs as merge blockers, the same way you treat a failed unit job. What the merge packet contains Keep the packet boring. One file. One schema. Commit it as a CI artifact, not as a comment that can be edited after the fact. { "commit" : "REPLACE_WITH_SHA" , "generated_at" : "2026-09-03T00:00:00Z" , "hooks" : { "pre_push_fixture_guard" : "passed" }, "fixtures" : { "manifest_path" : "tests/fixtures.sha256" , "changed_paths" : [], "unexplained_paths" : [] }, "required_jobs" : [ { "name" : "unit" , "conclusion" : "success" }, { "name" : "contract" , "conclusion" : "success" } ], "merge_ready" : false , "brief" : null } merge_ready is computed by a script you own. Not by a prompt. The brief is optional prose for humans who will not open the JSON. Step 1: Pin fixtures with a manifest Pick a directory you alrea

I recently built a complete IT ticket management system in Power Apps — 9 screens, role-based access, live SLA tracking, and automatic email notifications. The part I want to actually talk about here isn't the UI, it's the SLA engine, because I built it to work without Power Automate , and the trick is simpler than it looks. The problem SLA tracking normally means: a ticket is "Critical" → 60 minute target → somebody needs to know if it's about to breach or already has. The obvious way to do this is a scheduled Power Automate flow that checks every ticket on a timer and flags the ones in trouble. I wanted this app to run on Power Apps collections only — no flow, no external data source — so a scheduled flow wasn't an option. The question was: can you get "live" SLA status without a background job? The trick: recalculate on every read, not on a timer Instead of a flow updating a SLAStatus field periodically, I recalculate it every time the app or a screen is opened, using Now() against the stored due date: \ UpdateIf( colTickets, Status <> "Resolved" && Status <> "Closed", { SLAStatus: If( Now() > DueDate, "Breached", DateDiff(Now(), DueDate, TimeUnit.Minutes) <= SLAMinutes * 0.2, "At Risk", "On Track" ) } ); UpdateIf(colTickets, Status = "Resolved" || Status = "Closed", {SLAStatus: "Met"}) \ \ This runs in App.OnStart , at the top of every screen's OnVisible , and behind a manual "Refresh SLA" button. The At Risk threshold is 20% of the SLA window remaining — so a Critical ticket (60 min target) goes At Risk with 12 minutes left; a Low ticket (1440 min / 24 hrs) goes At Risk with 4.8 hours left. The honest tradeoff: this only updates when someone has the app open. A ticket breaching at 2am with nobody looking won't trigger anything until the next visit. For a real production deployment I'd pair this with a scheduled flow for after-hours detection — but for a demo, an internal tool with regular traffic, or anything where "eventually consistent within the next visit"

Design engineers increasingly work across two systems: the visual language of a product and the implementation details that make it usable. Skills for Design Engineers from ibelick focuses on that overlap, packaging practical guidance for building interfaces with stronger visual quality, clearer interaction patterns, and more consistent engineering decisions. The project is attracting attention, with +46 stars today . That momentum makes sense: design-focused AI workflows are moving quickly, but many generated interfaces still need human judgment around spacing, typography, responsive behavior, accessibility, and component reuse. The useful way to approach this project is not as a drop-in framework. Treat it as a reference layer for your development workflow. Read the relevant skill instructions, adapt them to your stack, and keep the resulting guidance close to the codebase so it can be applied consistently during implementation and review. A lightweight local setup might look like this: mkdir -p .ai/skills/design-engineering curl -L https://github.com/sponsors/ibelick \ -o .ai/skills/design-engineering/reference.html For a real team workflow, I would convert the useful parts into a checked-in Markdown file: .ai/ └── skills/ └── design-engineering/ ├── interface-quality.md ├── responsive-layouts.md └── review-checklist.md This keeps the process portable across editors and AI assistants instead of tying it to one tool. It also makes design decisions reviewable in pull requests, which is more valuable than keeping them inside an undocumented prompt. Before using the approach in production, watch for: Context drift: generic design guidance can conflict with an existing design system, so define project-specific tokens and component rules first. AI overconfidence: generated UI still requires manual checks for accessibility, keyboard navigation, mobile behavior, and performance. The strongest ROI comes from using these skills as repeatable engineering standards—not as a

At IFA 2026, Nvidia and its partners showed off the first RTX Spark-powered laptops and mini PCs, designed to run AI models right on your computer.

Most hackathons ask you to build whatever you want. Dogfood 2026 does the opposite. Everyone builds the same thing: a submission and judging platform for hackathons. The challenge is simple: Build the platform that will judge you. And there is a reason this is more interesting than it sounds. Hackathon Raptors has run 35 hackathons across 85+ countries since 2023. They have seen the same problems appear again and again: registrations, teams, submissions, judge assignments, scoring, normalization, results, certificates, and exports all becoming separate pieces of an increasingly messy workflow. Now they want to build the platform they actually wish they had. That is what Dogfood is about. About the Hackathon Dogfood 2026 is a 72-hour online hackathon organized by Hackathon Raptors . The event runs from September 25 to September 28, 2026 . At a glance 🌍 Online and global ⏳ 72 hours 💰 $2,500 prize pool 👥 Solo or teams of up to 4 💸 Free to participate 🔓 Open source 🐳 Self-hosted 🛠️ Build with the stack of your choice But this is not a normal platform-building challenge. The winning project is intended to be forked, self-hosted, and used for actual Hackathon Raptors events. So instead of building a demo that gets abandoned after the weekend, you are building something that could become real infrastructure. Why Build Another Hackathon Platform? Hackathon platforms already have most of the features organizers expect. Registration. Team formation. Project submissions. Public galleries. Judge scoring. Community voting. Organizer dashboards. CSV exports. So what is missing? The difficult part is not building another CRUD application. The difficult part is making the entire system reliable when real people start using it. Consider judging. Two judges can look at the same project and give completely different scores. One might give almost everything a 4 or 5. Another might rarely give anything above a 3. Simply averaging those scores can produce a ranking that reflects the judg

Ever wondered what's happening right now on a beach in Mexico, in Red Square, or at a harbor in Norway? I run Cam-World — a free live webcam aggregator that pulls together 7,000+ public streams from 100+ countries into one searchable place. No registration, no paywall. Here's a tour of what's inside and a few things I learned along the way. 🗺 The world map is the product The heart of the site is a dark globe where every green dot is a live camera. Click a cluster, zoom into a city, open a stream — you never leave the map. Watching it for a while teaches you something: the planet has a rhythm. Webcams go online with the morning sun, and the "online" wave rolls west around the clock. 📊 Honest uptime — you can tell a dead cam from a live one Aggregators usually show you a thumbnail and pray. We check every camera automatically and show a statistics widget: the last 24 hours and 30 days as color-coded slots (online / outage / offline / no data) plus an uptime percentage. The lesson here: webcams are ephemeral. Streams die, hotels turn off cameras, storms break them. Honest stats became our most-loved feature — users check reliability before clicking play. 🔎 Search, cities, collections Search works by name, city, country and tags. There are dedicated hubs for countries and cities, and themed collections: beaches, traffic, mountains, northern lights. 🌙 Small things that matter Dark & light themes (night couch-travel vs daytime browsing), 20 interface languages, "Near me" sorting by distance, live online/offline badges on every card. Try it 🗺 World map — pick a dot, watch live 🔎 Search — find a place you love 🏠 Home feed — a rotating mix of live cameras It's free, works on mobile, and there's always something happening somewhere. What would you check first — a beach, a mountain, or your own hometown square? 👇

Dynamic rendering in Angular sounds like a fairly narrow technical problem: “I don't know which component I need until runtime.” Angular already gives us several good tools for that. But there is a big difference between dynamically choosing a component and dynamically constructing an entire UI from a runtime specification. And that difference becomes especially important with Server-Driven UI and Generative UI. 1. ngComponentOutlet : when the problem is really just component selection For simple cases Angular already gives us: <ng-container *ngComponentOutlet= "componentType" /> This works very well when the application already knows its possible components and runtime logic only decides which one to display. componentType = condition ? UserCardComponent : AdminCardComponent ; The advantages are obvious: very little infrastructure, normal Angular lifecycle, AOT-compatible components and a relatively declarative template. But this approach starts becoming uncomfortable when the runtime input is no longer: UserCardComponent and instead becomes: { "type" : "Card" , "children" : [ { "type" : "Input" , "props" : { "label" : "Name" } } ] } Now we are no longer selecting a component. We are interpreting a UI description. 2. ViewContainerRef.createComponent() : more control, more responsibility Angular also allows components to be instantiated programmatically: const ref = viewContainerRef . createComponent ( componentType ); ref . setInput ( ' label ' , ' Name ' ); This is a powerful primitive. We control where the component is created, which component is used, how inputs are assigned and when the component is destroyed. For relatively contained dynamic behavior, this can be exactly what we need. But once a runtime specification controls many components, application code often starts evolving into something like: switch ( node . type ) { case ' input ' : ... case ' select ' : ... case ' button ' : ... case ' dialog ' : ... } Then we add input mapping. Then events. Then ne