This month brings a few new features to Svelte and SvelteKit and quite a few new libraries from around the community.

Also, in case you missed it last month, the Svelte maintainers released patches for 5 vulnerabilities across the Svelte ecosystem. So be sure you’re up to date by reading the blog post, CVEs affecting the Svelte ecosystem.

Now, without further ado, let’s dive in...

What’s new in Svelte & SvelteKit

• To support newer browsers that allow it, you can now customize <select> elements using CSS and rich HTML content (svelte@5.47.0, Docs, #17429)
• Svelte’s CSS parser is now exported from svelte/compiler as parseCss. It’s a partial but lightweight CSS AST parser that is currently being used in SvelteKit (svelte@5.48.0, #17496)
• A ShadowRootInit object can now be passed to attachShadow() when shadow root is created with a customElement (svelte@5.49.0, Docs, #17088)
• A breaking change in SvelteKit’s remote functions removes buttonProps from experimental remote form functions. Use <button {...myForm.fields.action.as('submit', 'register')}>Register</button> when creating mutliple submit buttons instead (kit@2.50.0, Docs, #15144)
• Node: The server’s keepAliveTimeout and headersTimeout options can now be configured by setting the corresponding environment variable (sveltejs/adapter-node@5.5.0, Docs, #15125)
• Vercel: Remote function calls can now be found under the /_app/remote route in observability (sveltejs/adapter-vercel@6.3.1, #15098)

For a full list of changes - including all the important bugfixes that went into the releases this month - check out the Svelte compiler’s CHANGELOG and the SvelteKit / Adapter CHANGELOGs.

---

Community Showcase

Apps & Sites built with Svelte

• Frame is a high-performance media conversion utility built on the Tauri v2 framework. It provides a native interface for FFmpeg operations, allowing for granular control over video and audio transcoding parameters
• LogTide is an open-source alternative to Datadog, Splunk, and ELK. Designed for developers and European SMBs who need GDPR compliance, data ownership, and simplicity without the complexity of managing an ElasticSearch cluster
• funcalling is a P2P calling app with built in board games, word games, and motion capture games (Reddit post includes a video demo)
• book by book lets you sync book data, covers, and pricing to a Square POS in seconds. You can import distributor orders from Ingram, Edelweiss & PubEasy—no spreadsheets or manual entry
• zsweep is a minimalist, keyboard-driven Minesweeper game played with Vim motions
• PulseKit provides professional feedback management to collect and prioritize customer feedback
• Bitphase is a modern web-based chiptune tracker designed for creating music on retro sound chips
• CapCheck is a fact checker AI for images, videos, audio, text and URLs
• Distill (last mentioned in July) has been updated with a new competitor finder to make it easier to track direct and indirect competitors in a market
• Pelican lets you generate SVG vector graphics and ASCII art from text prompts. With multi-step refinement, AI sees its output and improves it iteratively

Learning Resources

This Week in Svelte

• Ep. 127 — Changelog
• Ep. 128 — Changelog

To Read

• Bundling a SvelteKit app into a single binary by Gautier Ben Aïm

Libraries, Tools & Components

UI Components and Animations

• mapcn-svelte is a Svelte port of mapcn built on MapLibre GL, styled with Tailwind, works seamlessly with shadcn-svelte
• AgnosticUI Local (v2) is a CLI-based UI component library that copies Lit web components directly into your project
• Motion Core is a collection of animated Svelte components powered by GSAP and Three.js
• Tilt Svelte is a smooth 3D tilt Svelte attachment based on vanilla-tilt.js

State Management

• Reddo.js is a tiny undo/redo utility package for JavaScript, React, Vue, and Svelte
• svstate provides deep reactive proxy with validation, snapshot/undo, and side effects — built for complex, real-world applications
• rune-sync synchronizes reactive state across various storage backends

Plugins, Compilers and Runtimes

• fastify-svelte-view is a Fastify plugin for rendering Svelte components with support for SSR (Server-Side Rendering), CSR (Client-Side Rendering), and SSR with hydration
• kit-on-lambda is an adapter for running SvelteKit on AWS Lambda. It supports deployment to Node.js and Bun runtimes bundled with esbuild/Bun
• voca is a self-hostable, stateless, and fast WebRTC signaling server written in Rust with frontend SDKs
• @svelte-safe-html/core statically analyzes .svelte files and detects unsafe {@html} insertions
• sveltekit-discriminated-fields provides type-safe discriminated union support for SvelteKit remote function form fields
• svelte-fast-check is a type and Svelte compiler warning checker for Svelte/SvelteKit projects that claims to be up to 24x faster than the built-in svelte-check

That’s it for this month! Let us know if we missed anything on Reddit or Discord.

Until next time 👋🏼!

Writers Write shares writing prompts and writing resources. Use these 28 writing prompts for February 2026 to get you writing.

February – Getting Unstuck

Month 2: Perfectly imperfect

Hello Writer

This year, I want to focus on the most common challenges we face as writers.

Every month, along with a prompt for every day, I’d like to share an exercise that helps us deal with challenges like writer’s block, perfectionism, procrastination, and imposter syndrome.

Getting started

This month, we’re kicking perfectionism off its pedestal. You know that little voice whispering, ‘It needs to be brilliant.’ Or ‘It needs to be perfect’.? We’re going to ignore it. On purpose.

Because perfectionism isn’t the pursuit of excellence. It’s fear masquerading as a well-meaning protector who only keeps us stuck. That little voice is trying to keep you safe, but when it comes to first drafts, we need to take the risk, or we won’t ever create anything. It is hard. There is a time for making it perfect, but when you are just starting on a project, you need freedom, and you need to let go a little.

Exercise:

Set a timer for five minutes and write.

Write the worst version of your story, paragraph, or poem. Make it over-the-top terrible. Give yourself permission to laugh while you do it.

Why does this work?

When you deliberately try to ‘write badly’, your inner critic goes quiet, and your creative voice finally gets to breathe. Bad writing can always be improved. Blank pages can’t. So go ahead and write badly. Write with joy and courage.

Next month, we will discuss using the middle to get unstuck.

Good luck, writer.

28 Writing Prompts For February 2026

Download your prompts here: Writing Prompts February 2026

Do you want a daily prompt?

If you would rather have a free daily writing prompt from us, sign up here: Join Our Newsletter

Make the most of your writing prompts. Read How To Use Writing Prompts

Happiness
Mia

by Mia Botha

Looking for more prompts?

1. 31 Writing Prompts For January 2026
2. 31 Writing Prompts For December 2025
3. 30 Writing Prompts For November 2025
4. 31 Writing Prompts For October 2025
5. 30 Writing Prompts For September 2025
6. 31 Writing Prompts For August 2025
7. 31 Writing Prompts For July 2025
8. 30 Writing Prompts For June 2025
9. 31 Writing Prompts For May 2025
10. 30 Writing Prompts For April 2025

Top Tip: Find out more about our workbooks and online courses in our shop.

The post 28 Writing Prompts For February 2026 appeared first on Writers Write.

Postman is used today by millions of developers, from individuals building APIs for personal projects to teams inside large, global companies scaling the APIs that power customer-facing products, partner integrations, and service-to-service systems.

Across all these use cases, we see the same shift: API workflows are becoming more collaborative, more automated, and more central to everyday development. AI is also becoming a standard part of how systems are built, tested, and operated, with APIs serving as the interface between tools, services, and agents.

That’s why we’re bringing new capabilities to Postman in 2026. These features make AI and native Git workflows core parts of the Postman platform while also expanding support for end-to-end API and service workflows across development, testing, and production.

One of the most significant additions is Postman’s new API Catalog. It serves as a live operational layer for API portfolio management, bringing together specs, collections, test execution, CI/CD activity, and production observability in a single view. For engineering leaders and their teams, the API Catalog acts as a system of record, making it easy to see not only which APIs they own, but whether those APIs are tested, automated, and performing reliably across every environment.

Key new features coming in March

Additional new features releasing March 1 include:

• Native Git workflows that let developers work in feature branches and keep Postman work aligned with their code
• AI-native capabilities that can read, write, and reason across Postman assets and code
• Terminal and Code Editors within Postman for an uninterrupted development experience
• Local and CI mock servers to support API-first development and dependency isolation throughout the dev and test lifecycle
• Support for more API and messaging protocols in the Postman app, including GraphQL, gRPC, WebSocket, Socket.IO, MQTT, MCP, and AI request types
• Deeper CLI integration for automating testing, mocking, and publishing in CI pipelines
• A unified workbench that brings collections, specs, environments, mocks, flows, and files into a single development experience
• Insights to drive production visibility across your APIs, enabling teams to spot bottlenecks, track failures, and operate APIs reliably at scale

Lastly, to help our users make the most of these new features, we’ve redesigned our UI to be more responsive, Git native, and AI-compatible. The new design simplifies navigation, prioritizes search and Agent-first workflows, and minimizes context switching in the inner development loop.

While the new features listed above will be available for customers using v11, the new UI will only be available on Postman v12, which customers can upgrade to starting on March 1.

Simplified packaging

On March 1, we’re also updating Postman’s pricing and packaging. We have learned from customers that our existing pricing plans can be challenging to navigate given our growing suite of add-ons. In response, we’re introducing simpler plans that consolidate add-on capabilities and streamline how teams adopt Postman across their development, testing, and production workflows.

Moving forward, Postman’s new plans will be Free, Solo, Team, and Enterprise.

• The Free plan and new Solo plan are designed for individual developers and support everyday development workflows, from writing and running code locally to testing before changes reach CI. Solo extends this support with additional AI and automation capabilities for developers who want more power in single-player workflows.
• The new Team plan will support collaborative API development, testing, and distribution across shared workflows for groups of developers working together across development and test environments.
• The Enterprise plan will continue to enable organizations that are standardizing APIs and services at scale, with centralized visibility, governance, security, and operational controls.

Existing customers will receive an email with details specific to their current plan, including how it maps to the new plans and what, if anything, will change. New customers will be able to select one of our updated plans starting March 1, 2026.

We’ll share more details in the coming weeks, including more details about all the features coming to Postman, along with updated documentation.

We have a lot more coming to Postman in 2026, and we’re excited to keep building the future of API development together!

The post Coming to Postman in March: AI-native capabilities, a new API Catalog, and updated plans and pricing appeared first on Postman Blog.

This week on the GeekWire Podcast: We hit the road for a driving tour of the week’s news, making stops at Starbucks, Microsoft, and an Amazon Fresh store in its final days.

First up, Starbucks reports its first U.S. transaction growth in about two years — and announces plans for an AI “ordering companion” that translates cravings and concepts into custom drinks. We taste test a TikTok trend the old-fashioned way, ordering a “banana bread latte” at the drive-through.

Then we swing by the Microsoft campus after the company beat quarterly earnings expectations but saw its stock drop 12% in a single day. One of the reasons: investor concerns about the company’s exposure to OpenAI, which now accounts for roughly 45% of Microsoft’s contracted future cloud revenue.

Finally, Amazon is closing all of its Fresh grocery stores and Go convenience stores in the U.S., exiting its homegrown retail formats entirely. We visit a Seattle location during its clearance sale, and find a long line at a store whose original promise was no lines at all.

Upcoming GeekWire Podcast Live Event: Join us from 4 p.m. to 6 p.m. Thursday, Feb 12 at Fremont Brewing in Seattle for a live recording of the GeekWire Podcast with Todd Bishop and John Cook. Register here.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

Thea Felicity reports: The anti-ICE activist app and website StopICE has reportedly suffered a major security breach, exposing the personal information of more than 100,000 users to US federal agencies, including the FBI, ICE and HSI. Hackers said they accessed names, logins, passwords, phone numbers and GPS coordinates, and sent the data directly to authorities. One...

Source

It's been quiet here for too long, and since I've started a new ambitious project, I've decided to blog the progress and steps I'm going through since a lot of if will be relevant for anyone building an embedded device.

First a little backstory: I recently got a battery backup installed on the house. If there's ever a power outage I've been able to use Home Assistant to automatically turn non-essential things off, to extend the battery lifetime. However one thing that I haven't been able to control was my pool equipment, which is one of my most power hungry devices. For me to be able to do remote-control it, I'd first have to buy a $3000 "control center" and a $500 device to get it internet connected, and likely hire an installer to install it. And what you get is something that looks like a toy made in 1995. The markup on pool equipment is absolutely insane.  So I wondered: Can't I just build my own?

Well shortly later, my desk now looks like this:

In my usual style, I went completely overboard. Instead of just making a simple remote-operated switch which would have been a few lines of code, I'm now building a full on pool controller system. Because why not?

Using MQTT I can expose the state of the pool to Home Assistant as well as expose controls to turn things off. At the same time I wanted an easy to use pool controller screen that simplifies usage for the entire family for turning on heater, switch to hot tub only etc (currently it's a mess of setting a bunch of valves in the correct position or you'll risk messing things up).

I set up a set of goals to figure out how to build a proper embedded device:

• Should be completely configurable from the screen, including setting up WiFi, MQTT etc.
• Ability to easily look for new updates online and auto-update to a new build.
• Auto-starts when device turns on, and auto-restarts in case the app crashes.
• Fully remote controllable via MQTT.

Some of these things I've figured out already, and some of them I have ideas on but haven't tried just yet. My plan is to document and share everything here, both for myself for future embedded device projects, and for everyone else to learn from. All the source code will also be shared on GitHub. At the time of writing it is pretty basic, but I expect to build a set of re-usable components into it that you can grab from to build your own embedded device: https://github.com/dotmorten/PoolController

I'll be using .NET 10 and Uno Platform to accomplish all of this.

Here's some of the blog posts I plan on posting in the coming weeks (I'll make this list clickable as I post them):

• Hardware setup.
• Developer setup.
• Uno Platform project setup basics.
• Display management (rotation, touch, auto-screen-off and so on).
• Building a WiFi configuration interface.
• Auto-updating app from the internet.
• Auto-launching app and auto-restart after failure.
• Creating a custom image/setup for easy deployment to a new device.

I don't plan on going into super detail about how the pool controller itself works, how I communicate with the pool equipment, controlling sensors etc as that is extremely project specific, and you can always look at the code and eventually look at the hardware diagrams and doc (once those things materialize) if that bit interests you. My main goal with this series is to cover several of the scenarios a developer might go through to build an embedded device.