It's Foss News

There used to be a time when Linux gaming was a tricky affair, filled with trial and error, obscure fixes, and things randomly breaking. Many gamers used to avoid gaming on the platform due to those issues.

Now? Things have changed dramatically. Tools like Wine, Proton, DXVK, etc. have taken Linux gaming to another level. Bottles is one of those handy tools helping make the experience that much easier for gamers.

Sadly, the project has hit a funding roadblock.

Hard Work Deserves Appreciation

The lead developer behind Bottles, Mirko Brombin, recently shared an update on the project’s current state. He points out that, while Bottles has sponsorships from companies like Linode, JetBrains, and Hyperbit, they are still facing funding shortages that make sustained development difficult.

Despite having over 3 million downloads on Flathub, the project receives only about €100 per month in donations, an amount easily overshadowed by the server costs alone.

That sounds concerning. 🫤

Mirko also brought attention to Bottles Next, a complete rewrite of the app designed to modernize the codebase and improve performance. He said that they are still working on it, and while it’s due sometime in the future, continued support from Bottles users will help the team focus on development and deliver a better product faster.

He further added:

I am actively working to find sponsorships, I am in contact with a possible funding that could allow us to accelerate development, to pay a small bonus to those working on Next, to give some breathing room to those who are contributing. But here too, it takes time. And that’s precisely why today I feel the need to speak openly.

We don’t want to make Wikipedia-style appeals, with the usual “just one euro each.” But it’s right that those who love Bottles know how things really are. If you want to see Next grow, if you want to see Bottles finally become what it’s meant to be, we invite you to consider supporting us. Even just a symbolic donation, even just a monthly subscription, if done by many, can become what we need to take the next step.

If you use Bottles and want to see it grow, even a small donation helps more than you might think. Supporting the project now means faster updates and a better experience down the line.

Donate to Bottles

Suggested Read 📖

‘Don’t be Afraid to Contribute’: Mirko Brombin Talks about Vanilla OS and Other Future ProjectsA conversation with Mirko Brombin, founder of Vanilla OS and Bottles creator.It's FOSS NewsAnkush Das

From It's FOSS News via this RSS feed

Virtual Private Networks (VPNs) are a must for privacy-minded Linux users. They encrypt your internet traffic, mask your IP address, and help bypass censorship or surveillance.

Mozilla VPN is one such option from the makers of Firefox that has been available to Linux users for some years now, though installing it was limited to Debian-based systems via the .deb package.

This is no longer the case. 😉

Mozilla VPN Comes to Flathub

The Mozilla VPN client for Linux is now available as a Flatpak on Flathub, making it accessible to a wider range of Linux distro users. With this, anyone can install the VPN easily, regardless of their distribution, and without needing to worry about compatibility or manual setup.

/The app itself has yet to receive the verified badge on Flathub, but it should get approved soon, as Mozilla officially maintains it.

If you haven’t used it before, Mozilla VPN offers the basics, like encrypted connections, no logging, and the ability to block ads and trackers. You can use it on up to five devices, and it provides access to servers in over 30 countries. Plans start at $9.99 per month, with about 50% savings on the annual plan, which costs $59.88 before taxes.

Proton VPN - Protect yourself online | Proton VPNProton VPN is an open-source and audited Swiss VPN that cares about your privacy. Secure your internet traffic and keep your browsing history private with Proton VPN.Proton VPN

Want to Install It?

If you have Flatpak set up on your system, then you can install Mozilla VPN directly from your distribution’s Software Center (e.g., GNOME Software, KDE Discover, etc.).

You can also install it via the terminal by running this command:

flatpak install flathub org.mozilla.vpn

Launch it using the command below:

flatpak run org.mozilla.vpn

If you have a graphical interface, you can also start it from your app menu or launcher.

Via: OMG! Ubuntu

Suggested Read 📖

10 Best VPN for Linux Users To Secure Their Privacy [2025]Here are our recommendations for the best VPN services for Linux users to secure their privacy and enhance their online security.It's FOSSAnkush Das

From It's FOSS News via this RSS feed

Europe's push towards digital sovereignty is gaining serious momentum in 2025. Amid growing concerns over data privacy, access, and surveillance, many regions are distancing themselves from U.S.-based tech giants like Microsoft in favor of open source alternatives.

A growing number of European initiatives are working to keep infrastructure, data, and innovation within the continent. Their goal is to reduce dependence on foreign tech platforms that lock in customers and to build long-term digital independence through local open source development.

In a notable development, two European open source organizations have teamed up to offer an alternative to two popular workflow tools.

What's Happening: XWiki and OpenProject have announced that they are working on creating a fully open source alternative to Atlassian's Confluence and Jira, two of the most widely used workflow and collaboration tools out there.

Confluence is a popular platform for team documentation and knowledge sharing, while Jira is known for issue and project tracking across teams. Together, they dominate enterprise collaboration but remain proprietary and primarily U.S.-based.

As for the two organizations, XWiki is a well-known provider of open source enterprise wiki and knowledge management solutions, offering advanced collaboration features for businesses. It also develops CryptPad, a privacy-focused, end-to-end encrypted suite that many of us know and love.

OpenProject complements this with its open source project management platform, designed for agile workflows, task tracking, and team collaboration.

Speaking on the matter, Ludovic Dubost, CEO and Founder of XWiki, added that:

By combining our expertise, we’re creating a compelling alternative to Atlassian’s Jira and Confluence — one that puts control and transparency back into the hands of the user. It’s a step forward for open-source collaboration in Europe, and most importantly, it responds directly to what our customers have been asking for.

What to Expect: The goal of this initiative is to create an open, modular stack that allows teams to collaborate efficiently without the worry of vendor lock-in, forced upgrades, or loss of control over their data.

By combining their strengths in documentation handling and knowledge sharing, XWiki and OpenProject are working to deliver a unified, open source platform for managing both content and projects.

With backing from public sector efforts like ZenDiS and its openDesk platform, the project lines up with Europe's broader goals of digital sovereignty and accessible, self-hosted infrastructure.

Suggested Read 📖

French City of Lyon Kicks Out MicrosoftMicrosoft faces growing rejection in Europe whereas open source software sees growing adaption.It's FOSS NewsSourav Rudra

From It's FOSS News via this RSS feed

If you are a user of Ubuntu 24.10 'Oracular Oriole', then take note, as the release has officially reached the end-of-life (EOL) status. This means that, going forward, there will no longer be any new updates or important security patches for it.

This aligns with how Canonical manages Ubuntu's interim releases, which are supported for only nine months. Ubuntu 24.10 was never intended for long-term use. Instead, it was aimed at users who wanted easy access to the latest features and refinements Ubuntu had to offer.

With its support window now closed, it's time to consider your next steps. 👇

Ubuntu 24.10 is EOL, Act Now!

Ubuntu 24.10 is not covered under the Ubuntu Pro program.

Running Ubuntu 24.10 after support ends means leaving your system vulnerable to security threats and broken software. This is particularly risky in enterprise setups where stability and security are essential.

Unless you plan to do a fresh install of Ubuntu 24.04 LTS or wait for the next interim release (25.10) in October 2025 or the next LTS (26.04) in April 2026, upgrading to Ubuntu 25.04 is the only way forward.

Check Ubuntu Version

The lsb release -a command and the About page are handy to check which Ubuntu version you have.

Before you do anything, you should confirm that you are running Ubuntu 24.10. To do that, either run the command below or head into the Settings app and click on "System" to check.

lsb_release -a

Upgrading to Ubuntu 25.04

Then, make sure you've backed up any important data, and open the Software & Updates app. Under the "Updates" tab, check that you are set to receive notifications for all new Ubuntu versions by selecting "For any new version" from the dropdown menu in the last item listed.

Software & Updates app on the left and the terminal on the right, running the apt update and upgrade commands.

Now, run the following command to update your system:

sudo apt update && sudo apt dist-upgrade

After that, run the Update Manager using the command below. Install any pending updates, and restart your computer if prompted to do so:

sudo update-manager

Now run the same command again. Here, you should see an "Upgrade…" button. Click it and follow the on-screen instructions to begin the upgrade process.

You will see a welcome page with a link to the release notes for Ubuntu 25.04 'Plucky Puffin'. When you click "Upgrade," the necessary files will start downloading.

The release notes for Ubuntu 25.04 and the new packages list in the upgrader.

After the downloads finish, a new page will appear showing the packages to be installed, removed, or upgraded. Click "Start Upgrade" to begin the upgrade process.

The upgrader may ask you to choose some preferences. Once you’ve made your selections, the upgrade will continue, and, when finished, it will prompt you to reboot your computer.

📋If the upgrader crashes for any reason, don’t worry. The terminal will keep running the upgrade in the background. Let it finish, then simply restart the update-manager.

Suggested Read 📖

16 Killer Tips To Speed Up Ubuntu LinuxSome practical tips to speed up Ubuntu Linux. The tips here are valid for most versions of Ubuntu and can also be applied to Linux Mint and other Ubuntu-based distributions.It's FOSSAbhishek Prakash

From It's FOSS News via this RSS feed

Having a secure way to communicate, especially in conflict zones, is more important than ever. When authorities or the occupying power feel threatened, they shut down networks, block apps, and surveil everything, going to great lengths to suppress dissent.

Against this bleak backdrop, Jack Dorsey, co-founder of Twitter, has launched an open source project. It's a decentralized messaging app that works offline without needing the internet or cellular networks.

bitchat: What to Expect?

Built with offline and physical proximity use in mind, bitchat enables secure, peer-to-peer messaging by connecting nearby devices via Bluetooth LE Mesh Networking, allowing messages to hop between phones without involving conventional network connectivity.

bitchat’s architecture depends on a decentralized mesh in which each device simultaneously acts as both a node (client) and relay (server), where each device broadcasts and receives encrypted message packets using Bluetooth LE, enabling multi-hop communication without central servers.

As for its features, bitchat requires no accounts or phone numbers, supports IRC-style commands for easy navigation, enables group chats with channel management, and offers an emergency wipe option to instantly clear all data. It also includes performance optimizations like message compression and adaptive battery modes.

If you ask me, the features it provides can actually be useful for people who are in places where their government can’t be trusted and are facing constant surveillance, censorship, and sudden internet shutdowns that block normal communication channels, e.g., during protests, unlawful crackdowns, armed conflicts, or other emergencies.

Get bitchat

The app is currently in the testing phase via TestFlight (the beta spots are currently full) and is only available for iOS and macOS right now. There’s no word on an Android version yet.

The source code is hosted on GitHub as an Unlicense project. You can also check out the detailed whitepaper for more information on the underlying bits.

Explore bitchat

From It's FOSS News via this RSS feed

LibreOffice is a powerful, open source office suite that offers word processing, spreadsheets, and more. It’s an excellent alternative to Microsoft’s user-data-hungry Office 365, which was recently dropped by the French city of Lyon.

According to a recent development, it turns out LibreOffice has finally begun work on a long-requested feature, something that many have been waiting on for years now.

What's Happening: Work is underway to add native Markdown import support to LibreOffice Writer. This development, led in collaboration with a Google Summer of Code (GSoC) developer, aims to enable opening and editing of Markdown files (.md) using the MD4C Markdown parser.

The initial proposal to develop Markdown import support for LibreOffice Writer came from Ujjawal Kumar Chouhan, a student at IIT (BHU), India.

What to Expect: So far, the development has made it through four key milestones. The first one involved adding a basic stub for Markdown export, laying the foundational code that prepares LibreOffice to handle .md files.

Next, the integration of the MD4C parser was completed. This C-based library provides a reliable and CommonMark-compliant way to parse Markdown content, enabling LibreOffice to understand the structure and syntax of Markdown documents accurately.

Following that, a dummy Markdown Filter was introduced. It lets Writer open .md files, but for now, it only creates empty documents without showing any content.

And, the latest milestone, brings actual functionality, allowing Writer to import and render basic Markdown elements such as paragraphs, headings, and lists.

While promising, these improvements and Markdown support as a whole are still under active development. The work is progressing steadily, and the feature is expected to be fully completed and included in the 26.2 release scheduled for next year.

Once live, this will add to LibreOffice’s existing support for file types, making it even more useful and flexible for users.

Via: Phoronix

Suggested Read 📖

15 LibreOffice Tips to Get More Out of itLibreOffice is an excellent office suite. These LibreOffice tips will enable you to use it more effectively.It's FOSSAbhishek Prakash

From It's FOSS News via this RSS feed

The upcoming Blender 5.0 release is set to bring a notable improvement for Linux: experimental support for High Dynamic Range (HDR) rendering when using the Wayland display server with the Vulkan graphics backend.

This is a big step towards expanding Blender’s rendering capabilities on Linux, offering a more accurate visual preview for artists working in areas like animation, visual effects, and game development.

HDR Support on Blender: Why Does it Matter?

Source: Blender

If you're working in 3D art, animation, or visual effects, HDR can make a noticeable difference. It lets you use displays that support higher brightness, deeper contrast, and a wider range of colors, so what you see in your Blender viewport will be closer to what you’ll see on other HDR screens or in final renders.

This comes at a time when many Linux distributions are transitioning away from the older Xorg display server in favor of Wayland, which is being adopted as the new default display protocol. Wayland offers a more modern architecture designed to improve security, performance, and support for advanced graphical features.

Want to Try it Out?

🚧HDR support in Blender 5.0 is still experimental and under active development. It isn't stable, and is not recommended for production use at this time.

If you’d like to test the feature and explore its capabilities, follow these steps:

Download the Blender 5.0 alpha build Get the latest alpha version from the Blender Builds page.Ensure HDR is enabled on your system Use an HDR-capable monitorRun a Wayland session (e.g., GNOME or KDE Plasma)Enable HDR output in your system’s display settingsConfigure Blender for Vulkan and HDR Launch Blender and go to Preferences > System > Display GraphicsSelect Vulkan as the graphics backendGo to Preferences > Interface and enable Developer ExtrasNavigate to Preferences > Experimental and enable Vulkan HDR support Linux/Wayland.Restart Blender to apply changes

You can learn more about this addition in the original post by Jeroen Bakker from the Blender team.

Blender 5.0 Alpha

Via: Phoronix

From It's FOSS News via this RSS feed

Fedora got into a sticky situation recently, when a proposal to drop 32-bit support sparked significant backlash from the community. The change, aimed at reducing maintenance overhead for maintainers, was met with opposition from users and developers who rely on 32-bit compatibility for certain use cases.

The response was strong enough that the proposal had to be withdrawn, but the change itself will still need to be tackled sometime in the near future to avoid ongoing maintenance challenges.

Now, a new change proposal has popped up, looking to drop UEFI boot support on MBR-partitioned disks.

End of UEFI Boot Support on MBR?

Posted by Katerina Koukiou of the Anaconda installer team, the change proposal aims to drop support for installing Fedora in UEFI mode on MBR-partitioned disks, specifically on x86-64 systems starting with the upcoming Fedora 43 release.

The goal with this is to simplify the installer code and stop supporting a boot configuration that is rarely used, inconsistently implemented across firmware configs, and not officially tested by Fedora.

Plus, this setup can cause the bootloader to crash, adding to its list of problems. In the merge request for this change, Katerina puts it like this:

This patch introduces a disklabel_types constraint in platform definitions and enforces it during bootloader device validation. For the x86 EFI platform, only GPT is permitted. Other EFI-capable platforms (e.g., Aarch64, Arm, RISCV64) are unaffected and may continue using MBR where appropriate - such as in cloud or embedded environments where MBR UEFI booting is still used. This change will help us avoid bootloader crashes caused by firmware failures when MBR is used.

If you are an existing user, then you don’t need to worry, as this change won’t affect your installations. It only applies to new Fedora installations done via Anaconda, where GPT would be required instead of MBR for UEFI boot.

Also, while this change targets x86_64 systems, systems running Arm and RISC-V architectures will not be impacted by it.

This proposal is currently in the community feedback phase and needs to be voted on by the Fedora Engineering Steering Committee (FESCo) before it can be implemented. You can follow the conversation and contribute to it on Fedora Discussion.

Via: Phoronix

Suggested Read 📖

Linux Community Wins as Fedora Cancels 32-Bit Removal PlanFedora’s 32-bit removal plan has been dropped due to intense backlash from the community.It's FOSS NewsSourav Rudra

From It's FOSS News via this RSS feed

European countries have been growing increasingly wary of relying on Microsoft for critical government and public sector services. Concerns about data privacy, digital sovereignty, and potential governmental surveillance have led many to question the viability of depending on an American tech giant for sensitive infrastructure.

Many worry that dependence on Microsoft could leave them vulnerable to sudden service interruptions or the risk of sensitive data being accessed without consent.

This growing unease has already led Denmark to begin transitioning from Office 365 to LibreOffice, while the German state of Schleswig-Holstein recently abandoned Microsoft Teams and started the shift from Windows to Linux.

And now, another European region is jumping on the "Ditch Microsoft" bandwagon.

Microsoft Software is a No-Go

A translated version of the announcement by the city of Lyon.

The French city of Lyon has taken a major step toward digital sovereignty by officially starting a move away from Microsoft software. . The city is gradually set to replace Microsoft Office with open source alternatives like ONLYOFFICE and switch from Windows to Linux-based operating systems to reduce dependency on proprietary offerings.

Coinciding with this, they have launched the Territoire Numérique Ouvert (TNO), an open source collaborative platform developed with the Métropole de Lyon and the SITIV.

TNO integrates tools such as Jitsi for video conferencing, Nextcloud paired with ONLYOFFICE for document sharing and co-editing, Zimbra for email, Chamilo for online training, and Matrix for instant messaging.

According to the official statement, TNO is already being used by thousands of employees across several local governments in France. All its services are hosted in regional data centers located in the Auvergne-Rhône-Alpes region to ensure full control over public data and meet national digital sovereignty standards.

The project has received €2 million in funding from the Agence Nationale de la Cohésion des Territoires (ANCT).

💬 What do you think of this move? Do you expect more European countries to follow suit?

From It's FOSS News via this RSS feed

Born out of governance concerns in Organic Maps, CoMaps has been created with a goal of restoring transparency, community control, and privacy-focused navigation. It’s a fully open source navigation app that works offline and doesn’t track users.

A few weeks after the project began, CoMaps has officially launched with apps for both Android and iOS. Join me as I take you through its key features and what makes it stand out.

CoMaps: What to Expect?

A screenshot of CoMaps showing the Wikipedia entry of a location for additional information.

We begin with its most handy feature, offline navigation and search. With CoMaps, you can download entire regions before you begin a trip for easy navigation in places without an internet connection or mobile network connectivity.

The user interface is clean and intuitive. You won’t need a manual to figure it out. Whether you're a daily commuter or a weekend hiker, everything is simple, fast, and clear.

Moreover, privacy is baked into the app. CoMaps doesn’t track you, collect data, or run background analytics. You are responsible for what you share, and your location stays on your device.

Before you ask, what makes this different from Organic Maps?

CoMaps can remember your navigation routes as tracks, making it easy to revisit past trips or explore alternate routes. It also includes a re-mappable action button (just below the search bar), letting you customize it for quick access to your favorite function.

The map style itself is another highlight; it feels fresh, modern, and thoughtfully designed without being cluttered. Roads, paths, and terrain are clearly differentiated, making it easier to plan routes at a glance.

Bonus Tip: You can pair the default map view with the "Outdoors" map style to highlight more navigation routes. It’s a small change that makes a big difference when exploring.

Get CoMaps

The latest builds of CoMaps are available on F-Droid and the Play Store for Android users, and on the App Store for iOS.

If you’re looking for APK files or want to explore the source code, then you should visit the project’s Codeberg repository.

CoMaps

Suggested Read 📖

Organic Maps Forked Over Governance Concerns: CoMaps is BornAn Organic Maps fork has emerged due to broken trust.It's FOSS NewsSourav Rudra

From It's FOSS News via this RSS feed

It's effortless to manage photos on Linux thanks to some really great software that offer powerful organization, editing, and automation tools, all without needing pricey licenses from proprietary providers.

Among those, digiKam stands out, providing a ton of features while being a fully open source and community-driven solution. With the release of digiKam 8.7, it has received some useful upgrades that further enhance its capabilities.

digiKam 8.7: What's New?

One of the most notable highlights in this release is the introduction of an AI-powered auto-rotation tool that automatically detects the correct orientation of images using deep learning. Earlier, people had to manually rotate images or apply a fixed rotation value to an entire batch, which was inefficient when dealing with photos taken at different angles.

This has been integrated into the Batch Queue Manager for quick batch processing of images that need rotating.

Similarly, digiKam 8.7 adds new options to check if your system can use the GPU for AI tasks, leveraging OpenCL via OpenCV for general GPU support, and CuDNN SDK through the CUDA framework for NVIDIA GPUs.

Plus, you can also turn off OpenCL just for the reference engine if required, and there's a new test button in the settings to make sure GPU support is working right.

As for the internal tooling, this release comes equipped with Qt 6.8.3 for Linux and Windows and Qt 6.9 for macOS, with updates to the RAW decoder Libraw supporting over 1,260 camera models and a recent ExifTool version for more efficient metadata management.

The AI-powered face recognition system also gets upgrades, with digiKam now being able to remember when you reject a suggested face match and automatically offer the next best option, making tagging more accurate and efficient. Additionally, you can enable automatic face scans whenever new faces are confirmed or tagged by going into:

Settings > Miscellaneous > Behavior > Enable background face recognition scan

Beyond these new additions, digiKam 8.7 includes the following improvements:

Several fixes for the face management workflow.digiKam no longer remains active in memory after quitting.PKG installation issues on a recent macOS version have been resolved.Translations have been updated, with digiKam now supporting 61 interface languages and offering documentation in 16 languages (use the switcher on the bottom-left).

Get digiKam 8.7

The official website has links to the various packages for Linux, Windows, and macOS. By default, it should pick the download mirror closest to you, but if you face slow downloads, then you can always choose a different download mirror.

The source code is hosted over on KDE's GitLab instance.

digiKam 8.7

Suggested Read 📖

9 Best Linux Photo Management SoftwareLooking for a replacement for the good-old Picasa on Linux? Take a look at best photo management applications available for Linux.It's FOSSAnkush Das

From It's FOSS News via this RSS feed

The past few days have seen a strong response from the Linux community, who were not happy about Fedora's plan to drop 32-bit support with the upcoming Fedora 44 release.

The proposal sparked heated debates across forums and social media, with many users voicing concerns over how it would affect gaming, legacy hardware, and Fedora-based projects.

Things got so tense that the founder of Bazzite spoke out publicly, highlighting how the change could negatively impact their user base and future development plans.

Luckily, after receiving feedback from the community, the proposal has been dropped.

32-bit Support Stays, For Now

Fabio Valentini, a Fedora Engineering Steering Committee (FESCo) member, has formally withdrawn the proposal to discontinue 32-bit support in Fedora. The proposal, originally targeted for Fedora 44, aimed to address the increasing technical challenges in maintaining 32-bit compatibility across the distribution.

Acknowledging the backlash, Fabio admitted that the target for this change was too early and pointed out that the technical problems mentioned in the proposal are not going away.

He added that:

I don’t think the problem that was attempted to be addressed with this proposal will go away. With more and more projects dropping official support for building / running their software on 32-bit architectures, it’s just going to get worse over the next few years.

Fabio also said that he was disappointed by some of the reactions. He felt that people misunderstood the proposal and saw it in a more negative way than intended, and that it was meant to solve technical issues, not some conspiracy to break gaming use cases.

In the end, I figured this would be the end result of the proposal, and I get that 32-bit support in Fedora won’t last forever, and that’s okay. The good thing here is that Fedora-based distros like Bazzite and people who rely on 32-bit now have plenty of time to prepare for when that day eventually comes.

Suggested Read 📖

Fedora Looks to Completely Ditch 32-bit SupportFedora plans to drop 32-bit packages completely.It's FOSS NewsSourav Rudra

From It's FOSS News via this RSS feed

Secure video conferencing software is vital for organizations handling sensitive data, more so when they have to comply with standards like HIPAA, GDPR, and PCI-DSS to protect private information and ensure regulatory compliance.

Having said that, Cloudflare recently released Orange Me2eets, a newish video calling app with end-to-end encryption, that's designed to provide privacy and security for group calls.

Orange Me2eets: What is It?

Just a placeholder image of Orange Meets. (Source: Cloudflare)

You see, Cloudflare has this open source project called Orange Meets. It’s more of a demo for developers and researchers to explore than a polished app for everyday users.

That said, Orange Me2eets is an upgrade over the original offering, now strengthened with end-to-end encryption (E2EE) using the Messaging Layer Security (MLS) protocol, keeping video calls private, even from the infrastructure provider.

Cloudflare built this as an open source project to explore secure, E2EE group video calls at scale. By default, it uses the VP8 video codec for good quality and compatibility, and the app itself runs on a Selective Forwarding Unit (SFU) architecture, where the SFU can't read or access call data.

They also took inspiration from Discord’s DAVE E2EE video calling protocol to improve how the encrypted video streams work without causing playback issues in browsers. For example, they send the first few bytes of each VP8 video packet unencrypted, so the browser can play the video smoothly, while keeping the rest fully encrypted.

Get Orange Me2eets

Cloudflare has provided a live demo for testing, exploring its features, and getting a feel for how it works. If you find Orange Me2eets useful or see potential for your use case, you can consider self-hosting it on your own infrastructure for full control and privacy.

Orange Me2eets

Suggested Read 📖

5 Best Open Source Video Conferencing ToolsDon’t trust the big tech with your data? Try these open source video conferencing tools for online meetings.It's FOSSAnkush Das

From It's FOSS News via this RSS feed

The rustificiation of Linux has been steadily progressing, with the most popular components being Red Hat's Nova GPU driver for NVIDIA GPUs and the Rust NVMe driver maintained by the kernel community.

As Rust continues to gain traction, new projects are popping up, each trying to integrate this memory-safe language into Linux, exploring new areas like GPU drivers, storage, and networking.

The newest of these efforts is Tyr, a GPU driver developed jointly by Collabora, Arm, and Google, aiming to bring Rust support to Mali CSF-based GPUs.

Tyr GPU Driver: What to Expect?

Taking its name from a Norse god, the Tyr GPU driver is a port of Panthor, a downstream kernel driver for ARM Mali CSF GPUs, and takes pointers from Nova and rust_platform_driver.rs (link).

At this stage, the driver is bare-bones with limited functionality. It can power on the GPU and perform basic device probing, but many advanced features are still under development.

Support is currently limited to the Rockchip RK3588 SoC, and future development aims to add full GPUVM support, enable MCU booting, and improve power management. Plans also include handling synchronization through VM_BIND and introducing a software scheduler for fair access to GPU resources.

Daniel Almeida from Collabora, who submitted this patch to the kernel, explained that:

At the current pace, I am fairly certain that we can achieve a working driver downstream in a couple of months, for a given definition of "working". In any case, reconciling this with upstream has been somewhat challenging recently, so this patch constitutes a change in the overall strategy that we have been using to develop Tyr so far.

You can visit Tyr's GitLab repo to dive into the source code and follow ongoing development and discussions.

Via: Phoronix

Suggested Read 📖

New Linux Kernel Drama: Torvalds Drops Bcachefs Support After ClashThings have taken a bad turn for Bcachefs as Linux supremo Linus Torvalds is not happy with their objections.It's FOSS NewsSourav Rudra

From It's FOSS News via this RSS feed

These past few years, the use of AI has been growing in coding, with tools once meant for code completion or syntax suggestions now serving as intelligent helpers, offering debugging help, generating boilerplate, and even helping design entire applications.

This has given rise to specialized tools like Claude Code, Anthropic's coding environment made for working with its AI model, Claude. With this, instead of just chatting, users give it goals like writing or fixing code, and the model figures out the steps to get it done, handling things like editing files and running tests along the way.

Now, Asterisk, a YC-backed startup, has released Claudia, a free, open source app that gives Claude Code a clean, visual interface.

Claudia: What is It?

Claudia running on Ubuntu 25.04.

Its main goal is to make working with Claude Code easier and more intuitive, where, instead of relying on command-line tools, it gives users a visual interface to manage projects, run sessions, build agents, and track progress.

Claudia includes features such as project and session management, custom AI agent creation, sandboxed execution environments, real-time usage analytics, and an integrated Markdown editor. It runs locally, ensuring that your data stays on your machine for maximum privacy and security, with no need for cloud storage or external servers.

It’s designed for developers who want the power of Claude’s code reasoning without the complexity of terminal commands. Claudia lowers the barrier to using such advanced AI tools, letting developers focus fully on solving complex problems and writing better, more efficient code.

Want to Try Claudia?

Currently, Claudia can only be built from source and supports Linux, Windows, and macOS. Before getting started, make sure to install the Claude Code CLI along with all required prerequisites and dependencies.

Claudia

Via: AIM

From It's FOSS News via this RSS feed

The AI-powered terminal space is heating up fast as developers look for smarter, more efficient ways to code and manage workflows. Traditional terminals are evolving beyond basic command lines into intelligent workspaces that can assist with coding, automating tasks, and streamlining collaboration.

Google recently entered this space with its Gemini CLI, an AI-powered command-line interface that looks to enhance productivity by offering context-aware code suggestions, simplifying complex commands, and enabling natural language interactions directly within the terminal.

Amid this rising competition, Warp 2.0 has now arrived, reimagining itself as an "Agentic Development Environment" (ADE), bringing AI agents directly into developer workflows.

🚧Warp is not FOSS! We cover it because it is available for Linux.

Warp 2.0: What's in Store?

If you’ve been following Warp’s development, this move might not come as a surprise. It’s a big one that fundamentally changes what Warp is capable of. With the introduction of AI agents, Warp shifts from being a terminal to something much smarter.

These agents can help write code, automate tasks, and even handle parts of your workflow, all within the terminal itself. You can run multiple agents at once, making it easier to juggle different tasks without breaking your flow.

Warp Drive is another major addition that helps you and your team keep important things in one place, like commands, prompts, and environment settings. The AI agents can also use this information to give smarter and more helpful suggestions while you work.

The Warp team hasn't forgotten the core terminal experience; there’s now more powerful command editing, seamless mouse support, and syntax highlighting to make working in the terminal easier and faster than earlier.

In the launch announcement, Zach Lloyd, CEO of Warp, noted:

The products on the market today, from AI IDEs to CLI coding agents, all miss the mark supporting this workflow. They bolt agents onto code editors through chat panels and bury them in CLI apps.

What’s needed is a product native to the agentic workflow; one primarily designed for prompting, multi-threading, agent management, and human-agent collaboration across real-world codebases and infrastructure.

You can go through the detailed blog put out by him to learn more about this major update.

Get Warp

The latest Warp builds are available for download on the official website (partner link), offering easy-to-install packages for Linux, Windows, and macOS.

Warp

For existing users, updating to the latest version is simple using the Warp app or their system’s package manager. If you encounter any issues or have questions, the official documentation should be your next stop.

Suggested Read 📖

11 Vibe Coding Tools to 10x Your Development on LinuxWant to vibe code and chill on your Linux system? Here are the tools you can explore.It's FOSSAbhishek Kumar

From It's FOSS News via this RSS feed

Since completing its ~$61 billion acquisition of VMware, Broadcom has wasted no time reshaping the virtualization services provider to fit its aggressive enterprise software strategy. Among the first and most controversial moves was the discontinuation of perpetual licenses, forcing customers into pricey bundled subscriptions that make no sense for many organizations.

Not long after, they axed the VMware Cloud Services Provider program, doing away with a crucial channel for third-party service providers and cloud partners.

And now, it seems Broadcom has employed a new tactic to make life difficult for perpetual license holders.

Broadcom is Strong-Arming Enterprises

Source: Ars Technica

Shared with Ars Technica is an audit letter sent to a VMware user in the Netherlands who holds a perpetual license. This notifies the company that it has been selected for a formal review of its VMware software usage and licensing compliance.

The letter, dated June 20, states that the audit will be conducted by Connor Consulting and may involve fieldwork, remote testing, and meetings with their accounting, licensing, and IT departments.

The employee who spoke with Ars Technica regarding this also stated that:

Because we are focusing on saving costs and are on a pretty tight financial budget, this will likely have impact on the salary negotiations or even layoffs of employees. Currently, we have some very stressed IT managers [and] legal department [employees] …

They further added that they aren’t sure if their employer exceeded its license limits, but if they did, the financial consequences could be huge.

So, is Broadcom now in the business of bankrupting others? Maybe, because that’s precisely how their licensing enforcement is starting to look.

There have also been reports of Broadcom sending cease-and-desist letters threatening audits, with some companies being targeted even though they had not used any VMware updates since their support contracts ended.

I must say, it looks bad, and things appear to be getting worse. While there have been growing calls to investigate Broadcom's aggressive and possibly anti-competitive licensing practices, not much has been done so far. For now, companies are left to deal with the consequences on their own.

Suggested Read 📖

Free as in Fraud? A $130M Aerospace Company Caught Exploiting Open Source TrialAn unnamed $130M company has been exploiting Xen Orchestra’s free trial for over a decade.It's FOSS NewsSourav Rudra

From It's FOSS News via this RSS feed

The process of Linux kernel development is prone to arguments, strong opinions, and the occasional dramatic standoff. Despite that, development has progressed steadily over the decades, driven by a diverse community of contributors who are as passionate as they are principled.

And every once in a while, one of these internal debates spills into public view. This time, it's Bcachefs, the COW filesystem that pitches itself as a filesystem that "doesn't eat your data".

No More Bcachefs in Linux?

Linus is not happy.

With the upcoming Linux kernel 6.17 release, Linus Torvalds has decided to drop Bcachefs support, owing to growing tensions between himself and Bcachefs maintainer Kent Overstreet. The decision follows a series of disagreements over how fixes and changes for it were submitted during the 6.16 release cycle.

Friction between them isn’t new. Late last year, Kent was called out for an offensive reply to Linux memory management (MM) developer, Michal Hocko, and this just seems like a conclusion to everything that’s happened since.

You see, recently, Kent filed a pull request to add a new feature called "journal-rewind" It was meant to improve bcachefs repair functionality, but it landed during the release candidate (RC) phase, a time usually reserved for bug fixes, not new features, as Linus pointed out.

Theodore Ts'o, a long-time kernel developer and maintainer of ext4, also chimed in, saying that Kent’s approach risks introducing regressions, especially when changes affect sensitive parts of a filesystem like journaling. He reminded Kent that the rules around the merge window have been a long-standing consensus in the kernel community, and it’s Linus’s job to enforce them.

After some more back and forth, Kent pushed back, arguing that the rules around the merge window aren’t absolute and should allow for flexibility, even more so when user data is at stake.

He then went ahead and resubmitted the patch, citing instances from XFS and Btrfs where similar fixes made it into the kernel during RCs. Linus merged it into his tree, but ultimately decided to drop Bcachefs entirely in the 6.17 merge window.

To which Kent responded by clarifying that he wasn’t trying to shut Linus out of Bcachefs' decisions, stressing that he values Linus’s input, and saying:

I don't want to be in that position. I'm just not going to have any sense of humour where user data integrity is concerned or making sure users have the bugfixes they need. Like I said - all I've been wanting is for you to tone it down and stopholding pull requests over my head as THE place to have that discussion. You have genuinely good ideas, and you're bloody sharp. It is FUN getting shit done with you when we're not battling.But you have to understand the constraints people are under. Not justmyself.

Currently, the matter seems to have reached its conclusion. But given the ever-changing nature of kernel development, there's always a chance that perspectives could shift. That said, I wouldn't hold my breath.

You can follow the full story by checking the Linux kernel 6.16-rc3 and Linux kernel 6.16-rc4 mailing list threads related to Bcachefs.

Via: Thorsten Leemhuis

From It's FOSS News via this RSS feed

Fedora is known for adopting new technologies and making bold decisions well before other major Linux distributions. This approach has made it a top choice among developers and power users. Heck, even Linus Torvalds uses it.

But that forward thinking sometimes comes with controversy, and the latest example is its plan to completely ditch 32-bit support. The proposal outlines a two-step phase-out: first, removing all 32‑bit libraries from the 64‑bit (x86_64) repositories, and later, stopping i686 builds entirely.

Fedora argues this will eliminate a growing maintenance burden, pointing out that other distros have already dropped 32-bit support. They are right in that regard, but many others, especially in the Linux gaming community, are pushing back.

Among the critics is the founder of Bazzite, Kyle Gospodnetich, who has voiced serious concerns about what this change means for his project’s future.

Bazzite's Founder Isn't Happy

Kyle communicated his point of view clearly.

Kyle argues that, while he understands the intention behind the change, it is simply too soon to drop 32-bit support. He warns this move would kill off projects like Bazzite entirely.

He points out that essential Steam use cases would break even if the required packages were rebuilt. Beyond the technical issues, Kyle warns of significant reputational damage to Fedora.

After he said all that, there has been a lot of back and forth between Fedora developers and community members. Some defend the move as necessary progress, while others continue to push to preserve key 32-bit components vital to gaming.

Kyle, at one point, even said that if the change were to go forward as it is written, then the best option would be to disband the Bazzite project.

Luckily, the situation is not a stalemate. The proposal to drop i686 support is still under discussion and has not yet been approved by the Fedora Engineering Steering Committee (FESCo).

Plus, Kyle agrees with what Fedora maintainer and FESCo member, Neal Gompa, has said:

If I assume that the Steam client isn’t getting ported to 64-bit x86 anytime soon, and that nobody develops any 32on64 thunking mechanism for Linux libraries, then we have to think about how far we may need to keep it.

We can put this off retiring i686 for quite a long time since each Fedora release is only supported for ~13 months. The last release we could reasonably maintain support for 32-bit x86 is Fedora 65 (released in October 2036 if I got my math right), since its EOL would be November 2037.

All in all, most people involved have handled the situation well. While there were a few offhand replies, the key parties seem to have reached a mutual understanding.

Suggested Read 📖

Fedora Looks to Completely Ditch 32-bit SupportFedora plans to drop 32-bit packages completely.It's FOSS NewsSourav Rudra

From It's FOSS News via this RSS feed

In a world full of Android smartphones, de-Googled phones remain a rare breed, with most major brands having little incentive to break free from the Google ecosystem. This leaves privacy-conscious users with few out-of-the-box options.

As a result, they often have to take the DIY route, unlocking bootloaders, flashing custom ROMs like GrapheneOS, CalyxOS, etc. Still, between compatibility issues, limited device support, and the complexity of flashing ROMs, the demand for de-Googled phones is growing.

That’s where niche players like Murena step in, who offers sustainable smartphones in collaboration with Fairphone with their own in-house developed, Android-based /e/OS.

Now, they’ve launched a brand-new device to match their growing software ecosystem.

📝 Murena Fairphone (Gen. 6): Key Specifications

I think the color options are pretty well done.

Based on Fairphone (Gen. 6), the Murena Fairphone (Gen. 6) is a device designed with longevity and sustainability at its core. It features a modular design with 12 user-replaceable parts, including the display, cameras, USB-C port, and even the battery—all swappable with just a standard screwdriver.

Under the hood, it’s powered by a Snapdragon 7s Gen 3 chipset, paired with 8 GB of RAM and 256 GB of internal storage, which is expandable via microSD up to 2 TB (SDXC).

Running on the latest /e/OS 3.0, the phone comes with a strong commitment to longevity, promising five years of software updates and security patches, ensuring that these devices remain secure, functional, and relevant for years to come.

Other notable features include:

Display: 6.31″ Full HD+ LTPO OLED (protected by Gorilla Glass 7i)Camera: 50 MP and 13 MP rear, 32 MP frontBattery: 4,415 mAh removable Li-ionNetwork Connectivity: 2G, 3G, 4G, and 5G (dual-sim)Wi-Fi: Wi-Fi 6E (802.11ax)Bluetooth: 5.4 + LENFC: Yes

I must say, Murena has been on a roll lately. The launch of /e/OS 3.0 marked a major milestone for them, bringing improved parental controls, new privacy tools, and a range of usability enhancements. Around the same time, they also introduced Murena Find, a privacy-focused search engine that now comes as the default on all Murena devices.

🛒 Getting the Murena Fairphone (Gen. 6)

Priced at €649 (excl shipping), the Murena Fairphone Gen 6 is available in three colors: Forest Green, White, and Black. You can purchase it from the official website, but note that shipping is currently limited to select countries—with the U.S. being the most recent addition.

The phone comes with a standard 2-year warranty, but you can extend this by an additional 3 years for a total of 5 years by registering your device on Fairphone's website.

Murena Fairphone (Gen. 6)

From It's FOSS News via this RSS feed

The Linux Foundation has been growing at a steady pace, expanding its portfolio of projects and attracting an increasing number of organizations to collaborate on open source initiatives.

Among those, the influx of Big Tech companies stands out; industry giants like Google, Microsoft, and Amazon are increasingly contributing their technologies, protocols, and resources to open source projects.

While not everyone may like this, the growing collaboration points toward an interesting future for the open source ecosystem. Anyhow, let’s move on to the topic at hand.

What's Happening: Back in April 2025, Google launched the Agent2Agent (A2A) protocol, an open standard designed to enable secure and intelligent communication between autonomous AI agents.

Months later, Google has now donated A2A to the Linux Foundation, which in turn has formed the Agent2Agent project in collaboration with AWS, Cisco, Google, Microsoft, Salesforce, SAP, and ServiceNow.

Announcing the move, Jim Zemlin, Executive Director of the Linux Foundation, added that:

We are happy to be the new home of the Agent2Agent Protocol project. By joining the Linux Foundation, A2A is ensuring the long-term neutrality, collaboration and governance that will unlock the next era of agent-to-agent powered productivity.

What to Expect: If you didn't know, the A2A protocol enables autonomous AI agents to communicate securely, discover each other’s capabilities, and collaborate across organizational and technological boundaries.

It lays the groundwork for building dynamic, multi-agent systems that can work together securely and effectively, regardless of who built them, how they're deployed, or where they operate.

With this transition, the protocol is set to operate under a vendor-neutral governance model, ensuring that no single company controls it. The Linux Foundation will oversee the project with a focus on maintaining A2A’s core principles: extensibility, security, and practical usability across real-world industries.

For more details, you can read the official press release.

Suggested Read 📖

WordPress Fork Next? Linux Foundation Launches FAIR Package ManagerThe Linux Foundation launches a decentralized WordPress infrastructure alternative.It's FOSS NewsSourav Rudra

From It's FOSS News via this RSS feed

What used to be a painful experience is now a thing of the past. I’m talking about Linux gaming. Thanks to a growing ecosystem of open source tools and community-driven innovation, Linux is no longer just for power users; it's a reliable platform for both mainstream users and gamers alike.

Even more so if one is into single-player gaming. With tools like Wine, Proton, and Lutris, running Windows games on Linux is easier than ever, and more developers now offer native Linux support right out of the box.

I am not counting those pesky multiplayer games with kernel-level anti-cheat because the developers of such games are too stubborn to support Linux.

A few weeks ago, Tsunoa Games, an indie developer out of Spain, was kind enough to send over a key for Katana Dragon, their latest game that recently added native Linux and Steam Deck support.

📋We were provided with free access to the game by the developer. Views are my own.

What Makes Katana Dragon Stand Out?

Join over 56,000 Linux Users on YouTube

If you haven’t heard of this one, Katana Dragon is a beautiful voxel-style ninja action RPG that has two protagonists, Shin and Nobi, who are twin ninja siblings on a quest to end a curse that threatens the land of Sogen — a place where ancient temples, hidden dungeons, and strange creatures dwell.

I went with Nobi for my playthrough on a Bazzite system, and after completing the tutorial, the Sensei gave them their first mission as real ninjas: To deliver a gift to the Heshu family, who are the guardians of the Haen Forest.

Along the way, I ran into a bunch of enemies eager to take me out, but thanks to the dash ability and my trusty katana, I was able to make quick work of those rascals. Moreover, if I left the character idle for a while, it would eventually just fall asleep on the ground. This is a small thing, but it adds a lot of personality to the game, if you ask me.

As I progressed through the game, I found out that the Heshu family had turned into ghosts, and even Shin and Nobi’s Sensei wasn’t spared. To fix this, Nobi and Sensei set off on a journey into the enchanted Haen Forest to uncover what went wrong.

Here, the game introduces the Shadow Clone ability, which adds a fresh layer to solving those tricky pressure plate puzzles. This reminded me of Naruto, who uses clones to trick enemies and make battles easier. It’s a clever mechanic that adds both fun and challenge to the gameplay.

The Cursed Seals and Gokairium pages.

Finally, let’s talk about the skills, equipment, and some other gameplay elements of Katana Dragon. As you progress, you unlock various skills that let you customize your combat style. There are dragon gems you can collect and add to your gear to boost your stats, and cursed seals that give you powerful buffs but come with risks.

The map system is straightforward and really helpful for navigating the world. There’s also something called the "Gokairium," which acts as an in-game wiki for learning about the different enemy types.

I did run into a few issues, like random stuttering and the settings menu showing gamepad controls even when the keyboard option was selected, but that’s expected from an early access game.

🎮 How to Play?

While still early in development, the developers of Katana Dragon have more content on the way, including new dungeons, ninja skills, over 20 new enemy types, as well as new skins, cursed seals, and accessories.

As this is a native Linux game, there's no need to mess with Proton; everything works out of the box on popular distros like Fedora, Bazzite, or Ubuntu.

Katana Dragon is available for $9.99 on Steam if you're looking to give it a try.

Katana Dragon

Suggested Read 📖

Is Linux Ready For Mainstream Gaming In 2025?Linux is quietly gaining ground on Windows in the gaming space. But how well does it actually perform? Here’s what I experienced.It's FOSS NewsSourav Rudra

From It's FOSS News via this RSS feed

The demise of the 32-bit (i686) architecture has been slow but well-known within the Linux ecosystem, as popular distributions have stopped providing 32-bit kernel packages, installer ISOs, and other related packages.

For Fedora, this started with the Fedora 31 release, where many 32-bit components were shed in favor of focusing development and support on 64-bit architectures and improving multilib compatibility.

Now, it looks like Fedora is ready to take the next major step when it comes to 32-bit support.

The End of 32-bit Support: What’s Proposed?

Starting with Fedora 44, support for the 32-bit architecture is planned to be discontinued in a two-step process where the first step involves removing 32-bit libraries from the 64-bit (x86_64) repositories, a change that can be reversed if any issues arise.

The second step will permanently end the building of any i686 (32-bit) packages, making the discontinuation of 32-bit support final.

For those looking to run 32-bit applications, the developers mention that Wine’s WoW64 configuration will be required. This setup enables 32-bit Windows applications to run on 64-bit-only systems without the need for native 32-bit Linux libraries.

As for the reasoning behind this move, here's an excerpt from the change proposal:

Many projects have already been officially dropping support for building and / or running on 32-bit architectures, requiring either adding back support for this architecture downstream in Fedora, or requiring packaging changes in a significant number of packages to adapt to this dropped support.

By dropping support for the i686 architecture entirely, this additional - and growing - maintenance burden is eliminated.

In the discussion thread for it, many have opposed this move, citing concerns that it could break older games, disrupt Steam, affect OBS game capture, and create problems for downstream distributions like Bazzite.

Fortunately, this is still just a proposal in the feedback-gathering phase. It will only move forward if approved by the Fedora Engineering Steering Committee (FESCo), which seems unlikely given the strong community pushback.

💬 Do you think Fedora will proceed with this? Add your thoughts in the comments below!

From It's FOSS News via this RSS feed

A smooth and intuitive user onboarding experience is crucial for any modern operating system, as it’s the very first impression users get after booting up for the first time. This can make or break their overall perception of it.

Mainstream OSes like Windows and macOS both have nice initial setups that take users through essential steps like language selection, timezone settings, account creation, and privacy settings.

While I'm not a fan of the telemetry prompts and pre-installed bloat these systems often push users to accept and install, the concept of a guided first-run experience itself is neat.

On the Linux side, GNOME provides a comparable first-run experience, but not every distro comes with it out of the box. That’s where KDE’s new initial setup tool for Plasma comes in.

KDE Plasma’s User Onboarding is About to Get Better

Source: Kristen McWilliam (Merritt)

Okay, so the tool itself isn't entirely new; it has been around in KDE's repository for over four years now. It's called KDE Initial System Setup, or "KISS" for short. Two months ago, KDE developer Kristen McWilliam set out to revive it by giving it much-needed TLC.

The project initially started out as a way to provide KDE Plasma with a basic first-run setup experience, something that could guide users through essential steps like creating a user account, setting the system language, and keyboard layout.

In her blog, Kristen shared the current state of the undertaking, noting that the tool now successfully builds and runs, with new features like user account creation, language and keyboard layout selection, basic CI, and the addition of ECM logging.

Kristen revived the project two months ago.

For the future, there are plans for the tool to auto-run on first boot and in live sessions, allow authentication without a user prompt, have better, more developer-friendly documentation, and get a new name.

I like where this is headed. If it’s ready before the upcoming Windows 10 EOL in October, it could be a real game changer for KDE Plasma and help retain new users. I hope other major desktop environments like Cinnamon, Xfce, and Budgie take note of this and consider overhauling their user onboarding experience as well.

Via: OMG! Ubuntu

From It's FOSS News via this RSS feed

Void is a new, AI-powered code editor that banks on freedom, privacy, and transparency to offer a developer-first alternative to popular proprietary coding tools like Cursor and GitHub Copilot.

Unlike those platforms, Void is fully open source and gives users complete control over how AI runs on their hardware. They can choose which models to connect to, manage their data, and define what the editor can access.

Created by Andrew and Mathew Pareles, Void is a Y Combinator-backed project that aims to be the go-to open source alternative to Cursor. The first beta was released back in January, and it’s seen regular updates since.

In this article, I’m taking a quick look to see how it’s coming along.

Void: Overview ⭐

I tested Void's AI features using LM Studio, running the Gemma 3 12B model locally.

Thanks to its Visual Studio Code (VS Code) foundation, Void provides a familiar and lightweight coding experience right away, with support for themes, extensions, and standard key bindings. Core features like the built-in terminal, Git integration, and language server support are also there.

Our VS Code tips still apply here. 😉

Plus, Void is flexible enough to run AI models locally via LM Studio, Ollama, and even through proprietary providers like OpenAI, Anthropic, and Google.

⭐ Key Features

Full control over prompts and data.Fast, native editing even in huge files.Understands the whole codebase, no configuration needed.

Initial Impressions 👨‍💻

While I’m not a coder myself, I’ve done my best to showcase what Void has to offer. I tested it on Fedora 41 and connected it to a local LM Studio server running Google's Gemma 3 12B model.

For the demo, I loaded Dosage’s codebase, which is mostly JavaScript and features a clean, well-organized structure.

From the beginning, using Void felt snappy and responsive. The AI-powered features feel like a natural part of the editor, not something that was haphazardly added on top of VS Code.

If you’re into vibe coding, then this might be what you’ve been looking for.

LM Studio and Void's AI Models page.

Before I could start using the AI features, I had to set up a model through LM Studio. After a bit of trial and error, I got it working using LM Studio’s local server option under the "Developer" tab. Void automatically detected the installed models, making the process surprisingly smooth.

As shown in the video here and the first screenshot in this article, Void integrated nicely with the Gemma 3 12B model, handling prompts like "Summarize my codebase" and "Is the code good?", returning consistent, useful outputs.

It felt surprisingly smooth for a local setup, though the output was a bit slow since I was using the Nouveau GPU driver. I’m confident performance would improve with the proprietary NVIDIA driver.

I then checked out the built-in terminal, running simple commands like uname -a and echo "I really need to get back into coding". Everything worked smoothly, with no fuss or strange bugs. It felt just like using a regular terminal.

Void supports most VS Code themes and many extensions.

If you’re considering switching from VS Code, Void makes it pretty easy. Most of your extensions, themes, and settings should still work, so you don’t have to start from scratch. The interface feels familiar, and the usual features are right where you expect them to be.

📥 Get Void

Binaries for Linux can be found in Void's GitHub repository (pick the latest release under Assets), with additional packages being made available for Windows and macOS on the official website.

Keep in mind that Void is still in active development, so using it comes with the usual risks of pre-release software.

Void

I let you try and decide if Void can be on your list of vibe coding tools on Linux.

From It's FOSS News via this RSS feed