#10

Top Tech Picks of October

2025-11-04

Reading time: 6:34 min

Even in the October edition of Coders Corner, we bring you a roundup of the most interesting updates from the world of development. Java JDK 26 will soon introduce native support for HTTP/3, Elementor is taking website creation to the next level with its AI agent and new editor, and React 19.2 makes working with effects easier while bringing Partial Pre-rendering. The Zapier platform has entered the AWS Marketplace and now allows integrations via Action Runs. Finally, OpenAI has surprised everyone with its own browser, Atlas, which brings together the web and artificial intelligence in a single interface.

1

JDK 26 Gains HTTP/3 Support

Lukáš Huňár

Developer

The release is scheduled for March 2026 and finally brings HTTP/3 support to the standard Java HTTP Client API. This isn’t just another “oh wow, something new” update, but rather a quiet yet significant upgrade in how Java communicates with the modern web.

From an HTTP protocol perspective, HTTP/3 isn’t drastically different from HTTP/2 in terms of functionality. The main difference lies in the underlying transport protocol: while HTTP/2 runs over TCP, HTTP/3 uses UDP and is built on top of the QUIC protocol.

What does this mean for us?

  • Faster connections – say goodbye to slow TCP handshakes.
  • No blocking – streams no longer block each other like in HTTP/2.
  • Smarter mobility – switch from Wi-Fi to 5G without dropping the connection.
  • Better built-in security – TLS 1.3 is integrated by default.

In other words, it’s like moving from a bicycle (HTTP/2) to a motorcycle (HTTP/3).

If this news caught your attention and you’d like to learn more, be sure to take a moment – you can find more details here.

2

Elementor Fall 2025: New Features, AI Tools, and a Look into the Future

Mikuláš Žačok

Visual Content Creator

The Elementor platform continues to build on its vision of “intelligent and connected website creation.” The fall 2025 update recap brings a host of improvements—from an all-new generation of the editor and groundbreaking AI tools to accessibility enhancements and marketing automation upgrades.

The goal is for website creators to spend less time on technical details and more on design, content, and strategy.

Angie

One of the biggest highlights is the introduction of Angie, the first “agent-based” AI system for WordPress—an intelligent assistant that can act autonomously directly within your website.

Angie is more than just a chatbot—it’s an autonomous AI agent that understands the structure of WordPress, Elementor, and plugins, and can directly execute tasks.

What Angie can already do:

  • Create entire pages or sections based on natural language input.
  • Adjust design to match the style of an existing website.
  • Optimize content for SEO.
  • Generate texts, images, and contextual CTA blocks.

What’s coming next:

  • Integration with Ally – Angie will be able to automatically fix accessibility issues.
  • Integration with Send and marketing flows – generation of emails, campaigns, and autoresponders.
  • Extended memory and learning from website style – the AI will remember the creator’s preferences.
  • Access to the WordPress API – it will be able to manage plugins, menus, users, and backups.

Ally

The new Elementor Ally module helps creators build websites accessible to everyone.

Already available:

  • Ally Assistant can automatically scan a website and identify over 180 accessibility issues—from missing alt texts to color contrast problems.
  • Users can choose whether to fix issues manually or let the AI handle it.
  • The new accessibility widget allows visitors to adjust contrast, font size, or spacing directly on the site.

Coming soon (Q4 2025):

  • Heading structure (H1–H6) audit with SEO recommendations.
  • Automatic alt text generation for images.
  • Ability to run an audit without a plugin—directly from the browser.
  • Bulk fixes across the entire website.

Editor V4

The new Editor V4 is the biggest change since the inception of Elementor.

Core philosophy: more control, less code, and greater consistency.

Current features:

  • Single div wrappers – fewer HTML layers = faster websites.
  • Variables and style classes – a design system like in Figma.
  • States & interactions – new options for hover, click, and other states.
  • Responsive controls and a unified style panel.
  • Micro-animations directly in the editor.

Upcoming features:

  • Components – reusable elements (e.g., buttons, pricing tables) that can be synchronized or customized locally.
  • Variables Manager – visual management of variables.
  • Custom CSS directly on elements.
  • Nested tabs and advanced animations.
  • Coexistence of V3 and V4 – creators can transition gradually without losing compatibility.

The Send and Hosting modules will also receive updates, bringing faster websites, smarter campaign management, and advanced marketing automation directly within Elementor. Read more here.

3

React 19.2 Release

Ján Hažinčák

Developer

React 19.2 is the third release of the framework this year. The latest update brings several interesting features: the new React hook useEffectEvent, more efficient rendering with the Activity component, enhancements for DevTools, the new cacheSignal for React Server Components, and the concept of Partial Pre-rendering.

useEffectEvent

useEffectEvent is a new React hook that solves a common problem when using useEffect—specifically, when an effect uses an outdated variable value (a so-called stale closure). Example without useEffectEvent:
				
					import { useEffect, useState } from "react";

function ChatRoom({ roomId }) {
  const [user, setUser] = useState("Jano");

  useEffect(() => {
    const connection = createConnection(roomId);
    connection.onMessage((msg) => {
      console.log(user, "dostal správu:", msg);
    });
    connection.connect();

    return () => connection.disconnect();
  }, [roomId]); // user tu nie je v závislostiach
}
				
			
In the application, the effect (useEffect) runs, for example, when connecting to a WebSocket, but it needs to use the latest values from state or props—in this case, the user. Problem: When the user changes, the effect does not re-run, because user is not in the dependency array, so the callback keeps using the old value of the user variable. If you add user to [roomId, user], the effect would re-run unnecessarily every time the user changes. Solution with useEffectEvent: The effect only runs when roomId changes, but onMessage always “sees” the latest user without needing to re-run the effect:
				
					import { useEffect, useState, useEffectEvent } from "react";

function ChatRoom({ roomId }) {
  const [user, setUser] = useState("Janko");

  const onMessage = useEffectEvent((msg) => {
    console.log(user, "dostal správu:", msg);
  });

  useEffect(() => {
    const connection = createConnection(roomId);
    connection.onMessage(onMessage);
    connection.connect();

    return () => connection.disconnect();
  }, [roomId]);
}
				
			

Activity Component

React 19.2 introduces a new component, which allows you to efficiently manage parts of the UI that are either visible or hidden—impacting overall performance.
The Activity component includes a mode prop, which can take values such as "visible" or "hidden".
  • If mode="hidden": all child components are hidden and their effects (event listeners, subscriptions) are paused.
  • If mode="visible": everything works as normal.
Use case:
It’s very common in React to conditionally render a component like this:
				
					{isVisible && <Page />}
				
			
The problem is that when a component is hidden, it often loses its state (such as scroll position or form data), effects are unmounted, and remounting can be slow. With , you can avoid this issue.
				
					import { Activity } from "react";

function Sidebar({ isOpen }) {
  return (
    <Activity mode={isOpen ? "visible" : "hidden"}>
      <div className="sidebar">
        {/* nejaký obsah */}
      </div>
    </Activity>
  );
}
				
			
When isOpen is false, the sidebar is in hidden mode—its state remains preserved, but it doesn’t impact performance like a fully active part of the UI.

Performance Tracks in DevTools

React 19.2 brings improvements to Chrome DevTools—introducing the new Scheduler and Components Tracks.
  • The Scheduler track shows what work React is performing according to different priorities (e.g., “blocking,” “transition”).
It displays:
    • which events triggered a render,
    • the start and end time of the render,
    • whether React was waiting for something (e.g., “blocked waiting for paint”).
This helps to understand how React schedules and distributes work based on priorities.
  • The Components track shows the component tree React is currently working with—whether it’s rendering, running effects (useEffect, useLayoutEffect), or being blocked.
This makes it easy to detect which components spend the most time rendering and where potential bottlenecks are.

cacheSignal for Server Components

cacheSignal is available only for React Server Components. It helps interrupt long-running actions (like fetches) when a user navigates to another page and the result is no longer needed. This saves both performance and network resources.

Partial Pre-rendering

The new API for Partial Pre-rendering allows you to combine static and dynamic parts of your application. Static parts are pre-rendered (on the server or at build time) and dynamic parts are computed during client or server rendering. In other words:
The server (or build process) renders a “static shell” (app shell) that can be quickly delivered, for example, from a CDN, and then the dynamic content “fills in” once loaded on the client.
Find more details about React 19.2 here.

4

Zapier introduces AWS Marketplace integration and a new feature: Action Runs

Matej Matfiak

SAP CPI Specialist

The Zapier platform has made two strategic moves that are reshaping its position in the automation and AI ecosystem.

  • The first is availability on AWS Marketplace, allowing businesses to easily procure Zapier via Amazon Web Services.
  • The second is a new feature called Action Runs, which enables other software platforms to integrate Zapier’s automation logic directly into their own applications.

What do these updates bring?

Entering the AWS Marketplace streamlines the purchasing and approval process for larger organizations and opens the door to the enterprise space. Companies can leverage existing contracts, cloud credits, and AWS security standards, lowering the barrier to adopting automation solutions.

The Action Runs feature changes how SaaS platforms approach integrations. Instead of developing dozens of connectors, they can embed Zapier’s engine directly into their products, giving users access to thousands of actions across 8,000+ apps. This means faster implementation, less maintenance, and richer user options.

Why does it matter?

These steps confirm that Zapier is evolving from a no-code tool for individuals into infrastructure for AI and enterprise workflows. The partnership with AWS and the platform openness through Action Runs means greater scalability, easier integration, and new opportunities for both developers and end users. Read more here.

5

OpenAI launches its own browser, Atlas

Štelmák Michal

Developer

In October 2025, OpenAI introduced a new browser, ChatGPT Atlas, which combines the well-known Chromium engine with the intelligence of ChatGPT. Atlas is more than just another browser—its assistant is available to users throughout their browsing, helping search for information, analyze content, compare products, and even automatically fill out forms or plan tasks.

One highlight is the so-called Agent Mode, which allows the assistant to perform tasks independently on the user’s behalf. You can work with tables or PDF documents without switching between apps—everything is handled directly in the browser.

However, Atlas also raises questions around privacy and data security. OpenAI addresses this by letting users choose which sites the assistant can “see,” or operate in incognito mode.

Is ChatGPT Atlas the future of browsing? Only time will tell, but for everyone working daily with the web, development, or data, this is definitely a browser worth trying.

Missed the latest edition of Coder’s Corner? Read it here.