标签：#tor

If you're shipping a commercial WordPress plugin, sooner or later you'll need a licensing system. Something that lets paying customers activate the plugin on their site, locks it to that domain, and stops people from sharing the same key across fifty sites. The default answer in the WordPress world is Freemius or EDD Software Licensing. They're great. They're also a revenue share, a third-party dependency, and a black box you don't control. When we built RideCab WP , a commercial WooCommerce taxi booking plugin, we decided to build our own. Here's the architecture, the code, and the gotchas we hit along the way. What a Licensing System Actually Needs to Do Before writing any code, get clear on the requirements. A real plugin licensing system needs to: Generate unique license keys when someone buys Let the customer activate the key on their site Bind that key to one (or N) domains Validate the key periodically so revoked or expired keys stop working Handle deactivation when a customer moves to a new domain Fail gracefully — never lock a paying customer out because your license server hiccuped Optionally: deliver plugin updates only to valid license holders We'll cover 1 through 6 in this post. Update delivery is a separate beast and I'll write it up next. The Architecture The system has two halves that live in two different places. The license server runs on your own infrastructure — for us, it's a WordPress must-use plugin (mu-plugin) on the same WordPress install that powers our marketing site. It: Stores license keys in a custom database table Exposes a small REST API for activate, deactivate, and validate calls Provides an admin dashboard to view, create, and revoke keys The client is a PHP class shipped inside the commercial plugin (RideCab WP, in our case). It: Adds a license settings page to the plugin Calls home on activation Caches the validation result Re-validates quietly in the background Two pieces, talking over HTTPS, with the customer's domain as the b

json.loads(response) fails at a certain point. You told the model "return JSON only," but it added a ```json markdown code fence around everything. A quick regex strips it — until that regex hits an edge case, and that edge case blows up in production. Since Ollama 0.3.0, passing a JSON schema to the format parameter eliminates this problem at the root. The model's inference itself is constrained by the schema, so no code fences, no explanatory text, no mid-thought artifacts. Just parseable JSON. I ran these tests locally with Gemma4 and Ollama 0.30.7 to see how well it holds up in practice. Why LLM Response Parsing Is Tricky The most common problem when running Ollama locally — without a cloud LLM API — is JSON parsing. Two reasons. First, text generation models are trained toward "natural text." Even if you ask for JSON only, they'll often wrap it in json ... blocks or prepend "Of course! Here is the JSON you requested:" style text. Here's what I reproduced directly: json Input: 'Give me 3 Python tips as JSON with keys: tips (array), difficulty (1-5)' Model output (no format parameter): ```json { "tips": [ "Master the fundamentals first...", ... ] } JSON parse: FAILED Python ' s `json.loads()` can ' t handle the markdown wrapper . The " JSON only " instruction is unreliable in production . Second , speed . I measured the same query both ways : 32 seconds without structured output , 5 seconds with it . More on why below . ## How the Ollama format Parameter Works Ollama ' s `/api/generate` endpoint has a `format` field. Pass a JSON schema object and Ollama applies **constrained decoding** during inference. python import json import urllib.request def ollama_structured(prompt, schema, model="gemma4:e4b"): payload = { "model": model, "prompt": prompt, "format": schema, # ← pass JSON schema object directly "stream": False, "options": {"temperature": 0} } data = json.dumps(payload).encode() req = urllib.request.Request( " http://localhost:11434/api/generate ", data=data

🚀 Technical Briefing: This tutorial is part of our deep-dive series on Agentic Workflows at Gate of AI . For the full technical breakdown, interactive code sandbox, and the native Arabic translation, visit the original article here . <span>Tutorial</span> <span>Advanced</span> <span>⏱ 45 min read</span> <span>© Gate of AI 2026-06-16</span> Learn how to fine-tune large language models (LLMs) to enhance communication capabilities in specialized domains, such as homeless shelters, using modern AI tools and techniques like LoRA. Prerequisites Python 3.10+ OpenAI API key (latest version) Familiarity with machine learning concepts What We're Building In this tutorial, we will embark on a journey to fine-tune a large language model (LLM) to cater to the specific communication needs of homeless shelters. By leveraging a bespoke dataset compiled from the Youth Spirit Artworks (YSA) Tiny House Empowerment Village website, we aim to create a model that can effectively assist in the nuances of communication required in such environments. The finished project will result in a model capable of generating contextually relevant and empathetic responses to inquiries typical within the homeless shelter community. This involves structuring data into a standardized question-and-answer format to enhance the training process, ensuring the model's outputs are aligned with the communication style and needs of the target audience. Setup and Installation To begin, we need to set up our development environment with the necessary tools and libraries for model fine-tuning. We'll be using Python along with the OpenAI library to interact with the LLMs. pip install openai pandas numpy Additionally, you'll need to configure environment variables to securely store your API keys. This ensures that sensitive information is not hardcoded into your scripts. .env file OPENAI_API_KEY=your_openai_api_key Step 1: Data Collection and Preparation The first step in fine-tuning our model involves collecting and

QTRACTOR BIBLE Volume 1 — Foundations PART I — QTRACTOR CONCEPTS & WORKFLOW MODEL Chapter 1 — What Qtractor Is Quick Start Qtractor is a non-destructive, multi-track audio and MIDI sequencer designed primarily for Linux-based production environments. Unlike applications that attempt to integrate every aspect of the audio ecosystem into a single package, Qtractor focuses on recording, sequencing, editing, routing, mixing, and rendering while cooperating with external audio infrastructure and specialized tools. Qtractor is best understood as a timeline-centered production environment where audio clips, MIDI clips, plugins, automation, and routing configurations are organized into sessions. Common uses include: Music production MIDI composition Podcast production Voice recording Sound design Film scoring Hybrid hardware/software studios Live backing-track preparation Design Philosophy Qtractor follows several fundamental principles: Non-Destructive Editing Source media files remain unchanged. When a clip is trimmed, split, faded, moved, stretched, or processed, Qtractor modifies references and parameters rather than rewriting the original recording. Benefits include: Unlimited experimentation Reversible editing Reduced storage requirements Safer project management Session-Based Workflow Every operation belongs to a session. A session contains: Track definitions Clip placements Bus configurations Plugin assignments Automation data Routing information Tempo maps Markers Audio and MIDI source files remain separate from session instructions. Timeline-Centered Production The timeline is the primary workspace. Nearly every task ultimately relates to a position on the timeline: Recording Editing Automation Arrangement Export Qtractor is optimized for linear productions rather than clip-launching performance systems. Open Ecosystem Integration Qtractor assumes cooperation with: Audio servers MIDI systems External synthesizers External samplers Video playback tools Modular audi

A collection of stories about how militaries are using AI models to make decisions. This subscriber-only eBook is a package of six stories that were originally published in MIT Technology Review between April 11, 2025, and April 21, 2026, and have been updated to reflect recent developments. by James O’Donnell Choose which file format to…

Quick Tip: How to Choose the Right Model for Slack AI Workflows in 2026 I've been running Slack-integrated AI workflows in production for about three years now, and the question I get asked most often is deceptively simple: "Which model should I actually use?" Back in 2024, the answer was easy — you picked GPT-4o and moved on. But in 2026, with 184 models accessible through Global API and price points ranging from $0.01 to $3.50 per million tokens, that decision has become a genuine engineering problem. Pick wrong and you're either burning budget or shipping a sluggish experience. Pick right and your CFO actually smiles at you. Let me walk you through how I think about this, what the numbers actually look like, and where I've landed after months of benchmarking across multi-region deployments. Why Slack Workloads Are Weird Most people underestimate what a Slack AI assistant needs to do well. It's not a chatbot. It's a latency-sensitive, always-on, context-heavy workload that has to feel native inside a chat client where users expect responses faster than they can refresh the channel. In my experience, the three constraints that matter most are: p99 latency under 1.5 seconds for the first token — anything slower and users start double-messaging 99.9% uptime across at least two regions — Slack itself is up, so your AI better be too Cost per active user per month under $0.40 — this is the line where finance stops asking questions If a model can't hit those numbers consistently, it's not viable, no matter how clever the benchmark scores look. The Pricing Landscape I Actually Use Here's the table I keep pinned in my team's documentation. These are the models we rotate between depending on the workload. I haven't changed a single number — these are the exact rates as of writing this: Model Input ($/M) Output ($/M) Context DeepSeek V4 Flash 0.27 1.10 128K DeepSeek V4 Pro 0.55 2.20 200K Qwen3-32B 0.30 1.20 32K GLM-4 Plus 0.20 0.80 128K GPT-4o 2.50 10.00 128K The spread is w

How we live now is defined by unprecedented forces. In this special issue, WIRED and Architectural Digest help you understand what home will look like tomorrow—and beyond.

To build one family’s dream house on a flood-prone Mississippi bayou, AD100 architect Tom Kundig decided the sky’s the limit.

Here is a problem I hit building a long-running agent: I needed to inject a new instruction partway through a session ("the project is Go, write Go") but editing the top-level system prompt to add it invalidated my entire prompt cache. Every cached turn got reprocessed at full price. The fix is a feature that landed in the current Claude models: mid-conversation system messages. Here is what it is and when to use it. The setup that breaks A long agent session has a large, stable system prompt and a growing message history, and you cache the prefix so each turn reuses the prior work cheaply. That works until you learn something mid-session that the agent needs to know: a mode toggled, the user delivered async context, files changed on disk, the token budget dropped. The naive move is to edit the system prompt to include the new fact. But the system prompt sits at the front of the cached prefix. Change one byte there and you invalidate everything after it. Your whole conversation history reprocesses at full input price on the next request. For a long session, that is expensive and slow. The fix: a system message in the messages array The current models let you put a system -role message directly in the messages array, after the history, instead of editing the top-level system : const response = await client . messages . create ( { model : " claude-opus-4-8 " , max_tokens : 16000 , system : [ { type : " text " , text : STABLE_SYSTEM , cache_control : { type : " ephemeral " } }, ], messages : [ ... history , // cached prefix, untouched { role : " user " , content : latestUserMessage }, // @ts-expect-error: role:"system" SDK types may still be landing { role : " system " , content : " This project is Go. Write all code in Go. " }, ], }, { headers : { " anthropic-beta " : " mid-conversation-system-2026-04-07 " } }, ); Because the new instruction sits after the cached history, it invalidates nothing before it. The cached prefix stays intact, you pay full price only for the

Designers are finding sustainable building solves close to home—in ancient practices and cutting-edge innovations alike.

An AI agent has two kinds of memory: conversation (semantic) and context (exact reference). Keep them separate with Strands and AgentCore.

Tech has been encroaching on the family domicile for years—but actor, writer, and satirist Jill Kargman is all in on analog.

From friend compounds and meditation spaces to shaded outdoor areas and rooms just to make coffee, homes are getting even more multipurpose.

Buying a first house used to mark entry into adulthood—and the beginning of wealth-building. But a shifting economic landscape is threatening to close the door on this American milestone.

The state of Florida is suing TikTok over claims the company isn't complying with the state's child safety law, which bans kids under 14 from creating social media accounts, as reported earlier by Reuters. The lawsuit, filed on Monday, alleges that TikTok still allows 13-year-olds in Florida to use the platform and is "actively deceiving" […]

WIRED surveyed readers on their housing costs. The answers paint a stark portrait of unaffordability, climate adaptation, and the death of the homeowner dream.

A new, AI-assisted model of insurance is quietly exploding in disaster-prone areas—and may be coming for FEMA too. Is it the answer to climate change, or a trap?

An icon of Silicon Valley’s counterculture, Stewart Brand is confronting his final years in a home that embodies the self-sufficient, DIY ethos of his famous Whole Earth Catalog.

The global editorial directors of WIRED and Architectural Digest on teaming up to help you understand how we live today, and what comes next.

Devices that monitor seniors for safety are appealing to worried loved ones and underresourced home care agencies.