Phoronix reports on a presentation about trying FreeBSD on modern Framework laptop from last week's Open Source Summit hosted by the Linux Foundation: With FreeBSD having worked on improving its laptop support over the past two years with some big changes and ongoing efforts for making a nice KDE desktop experience on FreeBSD, FreeBSD Foundation's Executive Director has been trying to daily drive FreeBSD on laptops...

With the Framework Laptop, the touchscreen "just worked" as did other basic functionality from the KDE desktop on FreeBSD, including peripherals like a wireless mouse. Among the challenges were Zoom failing for video calls but eventually working, the web camera took steps to enable, and Microsoft Teams only partially worked. With the help of online resources, ultimately she was able to succeed in her journey of running FreeBSD daily on a laptop.

The Web Serial API lets websites write to (and read from) serial devices using JavaScript, including USB and Bluetooth devices with virtual serial ports. And this week's Firefox 151 release introduced support for the Web Serial API on desktop.

"Most folks won't use this API," acknowledges Mozilla's blog, "but for our community of builders and tinkerers, it unlocks the ability to use Firefox to communicate directly with compatible hardware devices like microcontrollers, development boards, and other serial-connected devices..." With Firefox's browser engine, Gecko, now supporting Web Serial, users can now connect, code, configure, and control compatible hardware directly from the browser in many workflows, often without additional software or complicated setup...

As part of this week's launch, Adafruit, one of the internet's most beloved open-source hardware communities, is collaborating with us to test and validate what browser-based hardware development can look like in Firefox with Web Serial support... With Web Serial support in Firefox 151, Adafruit's browser-based hardware workflows now work directly in Firefox as well, with no additional software or complicated setup required for many projects. We invite you to give it a try...

We want the web to be open, flexible, and shaped by the diversity of people building on it. If you're wiring up your first board, experimenting with hardware projects, or dusting off an old electronics kit, give Adafruit and Web Serial in Firefox a try. Build something amazing. Make something useful. Tell us what works. Tell us what breaks. Most of all, make it your own.
Mozilla's "Hacks" blog demonstrates with an Adafruit ESP32-S2 based board "where messages sent from web code can be directly displayed on the device over Web Serial."

And Mozilla engineer Alex Franchuk even built a handheld device that changes a web page's CSS properties.

The It's FOSS blog has news about the Linux Vendor Firmware Service, which gives hardware vendors a secure portal to upload firmware updates "which can then be downloaded and installed by users through clients such as GNOME Software or fwupdmgr." (Originally developed in 2015 by GNOME maintainer Richard Hughes...) The issue, however, obviously, had been funding with the largest contributors being the usual suspects, Framework and Open Source Framework Foundation, at $10K a year. Recently, however, Lenovo and Dell joined suite as Premier sponsors, which is the highest tier at $100K a year each, making the project more sustainable and manageable.

These companies contributing makes a lot of sense, considering they are two of the bigger computer companies which offer Linux by default in some cases, especially with Lenovo's ThinkPads being the Linux users' favorite for decades. And now... HP has followed suit as a Premier sponsor, also providing $100K a year, right alongside Dell and Lenovo...

The question still remains, however, where are the other vendors? What are they waiting for... This major move by these three companies should not only be seen as a sign of relief and wider acceptance of the usage of Linux, but as a beacon for other vendors to follow, who ought to make their hardware more accessible to the open-source community.

Linus Torvalds spoke this week at the Linux Foundation's Open Source Summit North America, reports ZDNet — and described how AI is impacting Linux kernel development: "In the last six months, we've seen a lot more commits," Torvalds noted, estimating that "the last two releases, it's been about 20% more commits than we had in the previous releases over many years.... The real change that happened in the last six months was that the AI tools actually got good enough for a lot of people... we're seeing a definite uptick in just development on pretty much all fronts...."

On the positive side, he framed AI-discovered bugs as "short-term pain" with long-term benefits: "When AI finds a bug in any source code... long term is you found a bug, we fixed it, that the end result is better for it." After all, he continued, "I think finding bugs is great, because the real problem is all the bugs you didn't find..." For small teams or solo maintainers, he said, flood-style AI bug reports can cause real burnout, especially when "it's a bug report, and when you ask for more information, the person has done a drive-by and doesn't even answer your questions anymore."
The AI news site Techstrong notes this quote from Torvalds. "I have a love-hate relationship with AI. I actually really like it from a technical angle, I love the tools, I find it very useful and interesting, but it is definitely causing pain points." The chief challenge with AI is that it forces people to change how they work, he found. People get into a rut, and AI challenges their norm. The Linux security mailing list got the brunt of this new wave of AI-generated commits. Not all bugs are security issues, but when "people think that when they find a bug with AI, the first reaction seems to sometimes be let's send it to the security list, because this may have security implications," Torvalds said. As a result, the security list — watched over by a small group of maintainers — was overrun by duplicate entries...

The Linux project learned to manage the bug influx with a set number of tools to sort out and deprioritize the obvious drive-by reports (ones where the person submitting the report won't even answer any questions). One tool, Sashiko, reviews all the patches submitted on the mailing list. "Sometimes the review is not great, but quite often it finds issues and it asks questions and says, 'Hey, what about this issue?'" he said.
Linux also updated their documentation, partly just to address "an uptick in bug and security reports from discoveries made in full or in part with AI."

The Free Software Foundation announced this week that "its global call for free software supporters to organize LibreLocals this May resulted in free software supporters organizing forty-six LibreLocal events on six continents thus far." (And new dates and locations are being added daily.) The FSF invited free software supporters to organize in-person community meetups in their area during May 2026, or LibreLocal month, to bring people together to swap ideas, learn from each other, and celebrate free software. People were encouraged to organize events grounded in freedom to help spread the free software philosophy.... "The success of these LibreLocals speaks to how many people globally are interested in free software and ready to build community, and it demonstrates the strength of our movement" [said FSF executive director Zoë Kooyman]. "People getting together like this also proves how computer freedom and digital rights are on people's minds. When we reject freedom-restricting software and promote software that respects user rights, it helps further so many other basic rights...."

The FSF has financially supported some of the events, but notes organizers are going above and beyond to create noteworthy events by any measure, and is impressed with the global network taking shape. "The energy we feel from all organizers is extremely motivating and we look forward to seeing LibreLocal events spread even wider over the next years! We want to support these initiatives even more, so we'll be looking to build a network of sponsors for future iterations as we work towards May 2027," says Heshan de Silva-Weeramuni, FSF program manager... William Goodspeed, the organizer behind the Beijing LibreLocal, reported that their meetup was double the size of last year's, and a number of very rich collaborative projects have emerged among the attendees.

Discussing the value of connecting people, de Silva-Weeramuni notes: "Free software supporters know that connecting with each other leads them to learn, experiment, and create great things that protect our individual and shared rights. The extraordinary contributions that free software has made to the world were born through such collaborations between like-minded people towards a freer society. This same global spirit of collectively building a better future is one of the inspiring things that we have once again seen unfold through this year's many LibreLocals."

Qualys's Threat Research Unit (TRU) has discovered and published a logic flaw in Linux kernel "that permits an unprivileged local user to disclose sensitive files and execute arbitrary commands as root on default installations of several major distributions." Friday their blog pointed out "The bug has resided in mainline Linux since November 2016 (v4.10-rc1)."

"Upstream patches and distribution updates are already available." Working exploits are circulating publicly, and administrators should apply vendor kernel updates without delay. During ongoing research into Linux kernel privilege boundaries, TRU identified a narrow window in which a privileged process that is dropping its credentials remains reachable through ptrace-family operations even though its dumpable flag should have closed that path. By pairing this window with the pidfd_getfd() syscall (added in v5.6-rc1, January 2020), an attacker can capture open file descriptors and authenticated inter-process channels from a dying privileged process and re-use them under their own uid. The primitive is reliable and turns any local shell into a path to root or to sensitive credential material [including host private keys under /etc/ssh ]

CVE-2026-46333 is local-only, but the impact is severe... Any unprivileged shell on a vulnerable host is enough to read /etc/shadow, exfiltrate SSH host private keys, or execute arbitrary commands as root through hijacked dbus connections to systemd. In practice, the distinction between an unprivileged foothold and full host compromise collapses: a phished developer account, a constrained CI runner, a low-privilege service account, or a shared multi-tenant host all become direct paths to root. With the vulnerable code shipping in mainline kernels since v4.10-rc1 (November 2016), the historical exposure spans nine years of enterprise fleets, cloud images, and container hosts.

Qualys followed responsible disclosure throughout. Qualys reported the vulnerability privately to the upstream Linux kernel security contact on 2026-05-11. Over the following three days the kernel security team developed and reviewed the fix, CVE-2026-46333 was assigned, and the patch was committed publicly on 2026-05-14. We then engaged the linux-distros mailing list, the standard pre-disclosure channel for downstream coordination. A short time later, an independent exploit derived from the public kernel commit appeared.... Qualys is releasing the complete advisory today because the underlying technique is novel, the public picture is now incomplete and uneven, and independent researchers have already achieved local root and published exploit material. Doing so gives defenders, detection engineers, and downstream maintainers a single authoritative reference for the flaw, the race against do_exit(), the role of pidfd_getfd(), and the four exploitation case studies.

NASA plans to open competition for the contract to operate JPL for the first time in nearly a century, meaning Caltech's historic role managing the iconic deep-space lab could come to an end when its current agreement expires in 2028. According to JPL, Caltech has managed the lab since the its inception in the 1930s, and has done so for NASA since the agency was established in 1958. Space.com reports: According to the JPL statement, Caltech has been preparing for this possible transition since last summer, so the news "comes as no surprise." But the potential change is part of a larger shakeup for the agency. Earlier this morning, NASA announced a major reorganization, which is separate from the JPL news. "To support the agency's ambitious short- and long-term goals, NASA is taking action to increase specialization at centers and integrate mission directorates, elevating delivery of technically excellent work," the agency said in a statement today.

JPL is NASA's lead center for the robotic exploration of Mars and other deep-space locales. The agency has worked with JPL through Caltech as a manager for nearly 70 years. Though JPL still counts as one of NASA's field centers, it's run as a contracted FFRDC (federally funded research and development center). This status has allowed the lab to function slightly differently than other NASA centers; it has a unique sort of independence, though NASA has always had significant oversight of the lab. "As an FFRDC, JPL operates under a special contractual and governance framework designed to ensure that its work is performed in the public interest and aligned with national priorities," NASA has stated. "The FFRDC model enables NASA to retain access to this depth of capability while maintaining a clear separation between government decision-making authority and contractor execution responsibilities."

Opening up the competition for institutions beyond Caltech to operate JPL could mean significant changes for everything from day-to-day mission management to big NASA science programs. Until now, JPL and Caltech have been heavily intertwined, with mission personnel, scientists, leadership, and others working closely "across the pond" between JPL and Caltech. JPL mission and program meetings often include Caltech employees and sometimes even take place on its Pasadena campus.

Trump Mobile confirmed that a third-party platform exposed customers' personal data to the open internet. The data included names, email addresses, mailing addresses, phone numbers, and order IDs. TechCrunch reports: Chris Walker, a spokesperson for the Trump-branded phone maker, told TechCrunch that the company is investigating the exposure and has not found evidence that content or financial information spilled online. The company said there was no breach of Trump Mobile's network, systems, or infrastructure. Walker said that the exposure was linked to a third-party platform provider that supports "certain Trump Mobile operations." He did not name the provider.

[...] On Wednesday, two YouTubers who ordered Trump Mobile's phone said a researcher alerted them that their personal information was exposed online. The YouTubers Coffeezilla and penguinz0 said they tried to alert Trump Mobile of the exposure after the researcher also tried but to no avail. Walker said Trump Mobile is evaluating whether it needs to notify customers of the exposure of their personal data. Further reading: Trump Phones Start Shipping - But Were There Really 600,000 Preorders?

An anonymous reader quotes a report from Billboard: Spotify and Universal Music Group (UMG) announced a licensing deal for recorded music and publishing rights, enabling Spotify to launch generative AI music models in the future. With this deal, Spotify's models will allow fans to create covers and remixes of their favorite songs from participating artists and songwriters signed to UMG. The new deal was announced on Thursday (May 21) as part of Spotify's Investor Day presentation, and the company touts that it will open up additional revenue streams on top of what artists already earn on Spotify and will provide new discovery opportunities for participating UMG talent. These AI products will eventually become available to premium users as a paid add-on. It is unclear when they are set to launch. "We recognize there's a wide range of views on use of generative music tools within the artistic community," the announcement read. "Therefore, artists and rightsholders will choose if and how to participate to ensure the use of AI tools aligns with the values of the people behind the music."

Spotify also announced a feature called "Reserved" that will set aside concert tickets for Premium subscribers it identifies as an artist's most dedicated fans. "Getting concert tickets today can feel like a race you're set up to lose," Spotify wrote in a post on Thursday. "You show up at the right time, refresh endlessly, and still miss out. Too often, the experience is stressful, unpredictable, and disconnected from what should matter most: whether real fans actually get tickets. We think there's a better way."

BrianFagioli writes: Flipper Devices has finally revealed Flipper One, a Linux-powered cyberdeck that sounds less like a gadget and more like an attempt to rebuild portable ARM computing from the ground up. Unlike Flipper Zero, which focuses on offline protocols like RFID and Sub-1 GHz radio, Flipper One is all about networking, modular hardware, SDR experimentation, local AI, and upstream Linux kernel support. The company says it wants to build "the most open and best-documented ARM computer in the world," complete with zero vendor BSP dependency and as few binary blobs as possible. That alone is enough to get Linux folks paying attention.

The hardware itself is loaded with nerd bait: dual Gigabit Ethernet, Wi-Fi 6E, M.2 expansion for SSDs and 5G modems, GPIO add-ons, HDMI 2.1, and a dual-processor architecture pairing a Rockchip RK3576 with a Raspberry Pi RP2350 microcontroller. Flipper Devices is even developing its own small-screen Linux UI framework because squeezing KDE onto tiny touchscreens is miserable. The company openly admits the project is financially and technically terrifying, which honestly makes this announcement feel more believable than most startup hardware pitches. Whether Flipper One succeeds or not, it is one of the most ambitious Linux hardware projects in years.

An anonymous reader quotes a report from Ars Technica: Google on Wednesday published exploit code for an unfixed vulnerability in its Chromium browser codebase that threatens millions of people using Chrome, Microsoft Edge, and virtually all other Chromium-based browsers. The proof-of-concept code exploits the Browser Fetch programming interface, a standard that allows long videos and other large files to be downloaded in the background. An attacker can use the exploit to create a connection for monitoring some aspects of a user's browser usage and as a proxy for viewing sites and launching denial-of-service attacks. Depending on the browser, the connections either reopen or remain open even after it or the device running it has rebooted.

The unfixed vulnerability can be exploited by any website a user visits. In effect, a compromise amounts to a limited backdoor that makes a device part of a limited botnet. The capabilities are limited to the same things a browser can do, such as visit malicious sites, provide anonymous proxy browsing by others, enable proxied DDoS attacks, and monitor user activity. Nonetheless, the exploit could allow an attacker to wrangle thousands, possibly millions, of devices into a network. Once a separate vulnerability becomes available, the attacker could use it to then compromise all those devices.

"The dangerous part here is that you can just have a lot of different browsers together that you can in the future run something on that you figure out," said Lyra Rebane, the independent researcher who discovered the vulnerability and privately reported it to Google in late 2022 in an interview. He said using the exploit code Google prematurely published would be "pretty easy," although scaling it to wrangle large numbers of devices into a single network would require more work. In the thread of Rebane's disclosure to Google, two developers said in separate responses that it was a "serious vulnerability." Its severity was rated S1, the second-highest classification.

Since its reporting 29 months ago, the vulnerability remained unknown except to Chromium developers. Then on Wednesday morning, it was published to the Chromium bug tracker. Rebane initially assumed the vulnerability was finally fixed. Shortly thereafter, he learned that, in fact, it remained unpatched. While Google removed the post, it remains available on archival sites, along with the exploit code. Google representatives didn't immediately respond to an email asking how and why it published the vulnerability and if or when a fix would become available. The exploit works by abusing Chromium's Browser Fetch API to open a service worker that remains persistently active. A malicious website can trigger it through JavaScript, creating a connection that can be used "for monitoring some aspects of a user's browser usage and as a proxy for viewing sites and launching denial-of-service attacks," reports Ars.

Depending on the browser, those connections "either reopen or remain open even after it or the device running it has rebooted," effectively turning the device into part of a "limited botnet."

Google is giving its iconic search box its first major redesign since 2001. The new design incorporates, you guessed it, artificial intelligence, "getting bigger and more interactive so that people can ask even longer questions and upload photographs and videos into queries," reports the New York Times. "In addition, people can ask follow-up questions with a chatbot on Google's main search page." From the report: The company will also offer digital assistants, known as agents, to automate searches so that someone who may be apartment hunting can be notified of a new listing without opening a real estate site like Zillow. The search features will be powered by a new artificial intelligence model, Gemini 3.5 Flash. Google said the model had improved on creating software code and performing autonomous tasks, worked faster and was less expensive to run than comparable models.

[...] Google is also bringing one of A.I.'s biggest breakthroughs -- software coding -- to search. When people research complex topics like astrophysics, Gemini can build interactive graphics and simulations behind the scenes to provide a deeper answer than its previous listing of websites. Google said it was introducing an alternative to the agents powered by Anthropic's Claude Code and OpenAI's Codex. Called Gemini Spark, the service is embedded in Gmail, Docs and other Google products, where it can turn meeting notes spread across emails and chats into a single document. It can also read and draft emails. "The open web is on its way out," says Richard Kramer, a financial analyst with Arete Research. "With A.I., Google is reducing everyone to raw data providers."

Microsoft is turning Azure Linux into a general-purpose, Fedora-based cloud distribution available to all Azure customers, while also productizing Flatcar as Azure Container Linux for immutable container hosts. "When Microsoft joined the Linux Foundation, there was this big conspiracy theory that somehow the Linux Foundation was undermining open source in partnership with Microsoft, and now you announce that you're shipping a Linux distribution," Jim Zemlin, the Linux Foundation's CEO, said in response to Microsoft's surprise announcement. "That's amazing." ZDNet reports: Until now, [Lachlan Everson, Microsoft's Principal Program Manager on Azure's open-source team] noted, "we had Azure Linux only available to third-party customers through AKS specifically, and that was Azure Linux 3.0." Going forward, this will be ACL. Everson emphasized that Azure Linux 4.0 is the culmination of years of internal usage and the evolution of the earlier Mariner distribution. "So we've been running Azure Linux for many years internally, and we got through to 3.0, and we only allowed it on as a container host on AKS. What we've done is make it a general-purpose, so this is all the learnings that we've had in the heritage of Mariner."

Under the hood, Azure Linux 4.0 is based on Fedora Linux and is delivered as an open distribution on GitHub. This code is available now. Yes, Red Hat knows that Microsoft has done this. Everson continued, "So, we made a decision to use Fedora as an upstream, so it's using RPMs in the Fedora ecosystem. Microsoft curates the packages and the supply chain to fit Azure's cloud platform." Microsoft also created "it to be purpose-built for Azure, which integrates vertically into all of our infrastructure to give you the best Azure Linux experience on Azure." While Azure Linux will ship as a VM image, Microsoft is already preparing a developer-friendly path onto Windows desktops: "And as of today, we have it as a VM image for your VM host on Azure. We're going to announce WSL images as well."

While developers will be able to run Azure Linux locally through WSL, Microsoft is not positioning it as a traditional desktop Linux. Asked whether he could run it on his laptop, Everson said: "I will be able to run it on my laptop, or what have you. Yes, on Windows 11." However, when pressed about a desktop experience, Everson was clear that there are "no plans" for a graphical environment. "It's optimized for server-side in the cloud," he said, adding that even on a developer machine, users should expect a lean environment. "Minimal packages, yeah. The idea is that we offer you a consistent experience to do your development on your machine, and that you can take your workloads as you develop them on your machine and run them with VS Code. You can run your applications on that, and know that the platform is the same that you're running on the cloud, so that you have that kind of consistency between environments."

Flatcar itself remains the upstream project, but Microsoft is packaging it for Azure customers. Everson described Flatcar as "purpose-built, immutable, secure by default, production-ready operating system, and Azure Container Linux is the productization of that, but we're still investing in the upstream Flatcar ecosystem and pulling that downstream into a productized exterior experience just for container workloads, so it's a container hosting in AKS." To underscore the immutable model, he added that "Everything's baked in, so there is no package manager. We bake the bits into the immutable, and they're in the immutable version. So Azure Container Linux is the immutable version. So you shouldn't be changing any system packages or any application packages. Anything that you need to change is customer workloads run in containers."

An anonymous reader shared this report from Electrek: When Fisker Inc. filed for Chapter 11 bankruptcy in June 2024, it left roughly 11,000 Ocean SUV owners holding the keys to vehicles that cost them anywhere from $40,000 to $70,000 — and that were rapidly losing the software brains that made them work. No more over-the-air updates. No more connected services. No more warranty. The manufacturer was dead.

What happened next is one of the most remarkable stories in the history of the electric vehicle industry. Instead of accepting that their cars would become rolling paperweights, Fisker Ocean owners organized, reverse-engineered their vehicles' proprietary software, hacked into CAN bus networks, built open-source tools on GitHub, and effectively stood up a volunteer-run open-sourced car company from the ashes of Fisker... Within months of the bankruptcy filing, thousands of Ocean owners formed the Fisker Owners Association (FOA) — a nonprofit that quickly grew to 4,000 members and began operating as something between a car club, a tech startup, and an independent automaker. The FOA hired independent tech experts who began reverse-engineering Fisker's proprietary software patches. Members taught each other how to flash firmware. They organized bulk purchases of replacement parts — negotiating the price of key fobs down from roughly $1,000 each to a fraction of that through coordinated group buys. They hosted free global key fob pairing events, saving each owner $100 to $250...

What started as desperate troubleshooting has evolved into a genuine open-source ecosystem around the Fisker Ocean. On GitHub, a developer named MichaelOE reverse-engineered the API behind Fisker's official "My Fisker" mobile app and built a Home Assistant integration that exposes every cloud API value as a sensor — with all the app's buttons available as Home Assistant controls... [Community members have also been systematically mapping CAN bus files.]
The article noes this "is not an isolated incident. Nikola also filed for bankruptcy, leaving its owners in a similar bind. Canoo and Arrival are headed for liquidation auctions..." Consumer advocates are now pushing for structural changes: mandatory software escrow funds that would keep vehicle software running even if the manufacturer disappears, open-source mandates in bankruptcy proceedings, and shared repair data requirements... European automakers, meanwhile, are moving in a different direction entirely — Volkswagen, BMW, Mercedes-Benz, and eight suppliers signed a memorandum in 2025 to develop a shared open-source automotive software platform....

The Fisker Owners Association has proven that a dedicated community can keep orphaned EVs on the road. But they shouldn't have had to... [O]wners shouldn't need to become hackers and parts brokers and quasi-manufacturers just to keep driving the cars they already paid for.

Today Amazon ends support for first- and second-generation versions of Kindles and Kindle Fire tablets, along with the Kindle Touch, the 9.7-inch Kindle DX, and other devices released in 2012 or earlier.

Owners can continue reading ebooks that they've already downloaded, and they can also still sideload books using a USB cable (from, for example, Project Gutenberg). And PCMag points out that "There are plenty of e-stores where you can buy DRM-free novels legally, such as ebook.com and Smashwords. If you want to try this process for free, public-domain repositories such as the one at Standard Ebooks are a great place to start." (eBook files can be converted for the Kindle with the open source tool Calibre.)

New ebooks can no longer be purchased directly from Amazon. But most of Amazon's affected devices "have not received firmware updates for over a decade," notes the blog OMG Ubuntu, "and most lost on-device access the Kindle Store." Some Kindle owners are taking things even further: You can unlock the firmware of older devices to add extra functionality (custom screensavers, epub support) or run entirely different software. On the hardware hacks side, some choose to turn old Kindles into photo frames or online dashboards.
TechCrunch offers some caveats about jailbreaking: This process allows users to install custom fonts, new screensavers, alternative reading apps, and even third-party tools that expand the Kindle's functionality... [I]t's important to note that jailbreaking a Kindle might violate Amazon's terms of service. In many jurisdictions, jailbreaking isn't considered a criminal offense for personal use, but it may become a crime if it involves copyright infringement, illegal software distribution, or the sale of modified devices. Many Kindle owners who opt to jailbreak view it as a method to gain control over a device they purchased that is still functional, rather than being forced to buy a new device. However, jailbreaking is technical and carries risks, including the possibility of rendering the device unusable if something goes wrong. It also isn't possible on every Kindle model or firmware version, so before proceeding, Kindle owners should first spend some time researching if their device is compatible.
Alternately, PCMag notes, "If you're feeling particularly virtuous, you can donate your old Kindle to a local library or send it back to Amazon free of charge via its electronic recycling program."

Former Microsoft programmer Keith Curtis "wrote and self-published After the Software Wars to explain the caliber of free and open source software," according to his entry on Wikipedia, "and why he believes Linux is technically superior to any proprietary OS."

He's also KeithCu (long-time Slashdot reader #925,649), and has written a blog post on "How I added an LLM-based grammar checking + TeX math import to LibreOffice." : At Microsoft, I spent five years working on the text components RichEdit and Quill, and came to understand the "physics" of word processing: the file formats, data structures, and algorithms that provided fast access to text and properties, independent of the length of the file. Selecting one million characters to make them bold took about the same time as changing one character, because of the clever data structures (piece tables) and algorithms in these engines...

When I decided to add a real-time AI grammar checker to [LibreOffice plugin] WriterAgent, I knew what I was getting into, but I underestimated the trickery of LibreOffice's UNO.
His site shares the surprises he encountered, one by one. (Starting with "the office suite throws a bunch of initialization variables at your constructor. If your Python __init__ method doesn't handle them, the code fails to map the call, the stack misaligns, and the program dies.") There's sentence casing issues, duplicate words, and foreign-language syntax — all culminating in new features for "a LibreOffice extension (Python + UNO) that adds generative AI editing to Writer, Calc, and Draw..."

"If you want to try it out, the repo is here... Let's make LibreOffice and the free desktop AI-native!"

The blog It's FOSS has an update on the Fedora AI Developer Desktop Initiative, a proposed platform for AI/machine learning workloads on Fedora. It's now been blocked "after two Fedora Council members retracted their earlier approval votes." The initiative was proposed by Red Hat engineer Gordon Messmer, aiming to deliver an Atomic Desktop with accelerated AI workload support, covering developer tools, hardware enablement, and building a community around AI on Fedora... At the May 6 council meeting, the members unanimously voted to approve this new initiative. After which a short, lazy consensus window was left open until May 8 to accommodate absent members, after which the decision was to be ratified.

But that last bit never happened, as council member Justin Wheeler (Jflory7) was the first person to change their vote to -1... ["While I strongly support leveraging AI to establish Fedora as a leading platform, completely rearchitecting our kernel strategy is a massive structural shift. It requires explicit alignment with our legal and engineering stakeholders before we commit the project to this path."] Following that, fellow council member Miro HronÄok (churchyard) put in his -1, saying that he had originally assumed the proposal was purely additive and therefore uncontroversial. But seeing the community's response, he realized that he was mistaken about that. As an elected representative, he felt the need to reflect on this major proposal before signing it off.

Over 180 replies have piled up in the proposal's discussion thread, with many well-known Fedora contributors pushing back on things like kernel policy, proprietary software, and project identity. Hans de Goede from the packaging team called out the proposal's emphasis on CUDA support as going against Fedora's foundational commitment to free software, arguing that open alternatives like AMD's ROCm and Intel's oneAPI should be the focus instead.

AMD says FSR 4.1 will finally bring its newer hardware-accelerated upscaling technology to older Radeon GPUs. "The rollout will begin in July with RDNA3- and 3.5-based GPUs, which include the Radeon RX 7000 series, as well as integrated GPUs like the Radeon 890M and Radeon 8060S," reports Ars Technica. "In 'early 2027,' support will also be extended to the RDNA2 architecture, which includes the Radeon RX 6000 series, integrated GPUs like the Radeon 680M, and the Steam Deck's GPU. This would also open the door to supporting FSR 4 on the PlayStation 5 and Xbox Series X and S, all of which also use RDNA2-based GPUs." From the report: [AMD Computing and Graphics SVP Jack Huynh's] short video presentation didn't get into performance comparisons, but did mention that AMD had to work to get FSR 4's superior hardware-backed upscaling working on its older graphics architectures. RDNA4 includes AI accelerators that support the FP8 data format in the hardware, and porting FSR 4 to older GPUs meant getting it running on the integer-based INT8 hardware in the RDNA3 and RDNA2-based GPUs.

This may mean that FSR 4.1 running on an RDNA3 or RDNA2-based GPU may come with a larger performance hit relative to RDNA4 cards, or that image quality may differ slightly. Modders have already worked to get FSR4 working on INT8-supporting GPUs, and the older GPUs reportedly see a 10 to 20 percent performance hit relative to FSR 3.1 running on the same hardware. AMD's official implementation may or may not improve on these numbers.

[...] Any games that support FSR 4 should be able to support FSR 4.1 running on Radeon 7000-series cards; users will presumably be able to install a driver update in July that enables the new feature. Games that support the older FSR 3.1 can also be forced to use FSR 4 in the Radeon graphics driver.

ArXiv says it will ban authors for one year if they submit papers containing AI-generated slop, such as hallucinated citations, placeholder text, or chatbot meta-comments left in the manuscript.

"If generative AI tools generate inappropriate language, plagiarized content, biased content, errors, mistakes, incorrect references, or misleading content, and that output is included in scientific works, it is the responsibility of the author(s)," said Thomas Dietterich, chair of the computer science section of ArXiv, on X. "We have recently clarified our penalties for this. If a submission contains incontrovertible evidence that the authors did not check the results of LLM generation, this means we can't trust anything in the paper." 404 Media reports: Examples of incontrovertible evidence, he wrote, include "hallucinated references, meta-comments from the LLM ('here is a 200 word summary; would you like me to make any changes?'; 'the data in this table is illustrative, fill it in with the real numbers from your experiments.'" "The penalty is a 1-year ban from arXiv followed by the requirement that subsequent arXiv submissions must first be accepted at a reputable peer-reviewed venue," Dietterich wrote.

Dietterich told [404 Media] in an email on Friday morning that this is a one-strike rule -- meaning authors caught just once including AI slop in submissions will be banned -- but that decisions will be open to appeal. "I want to emphasize that we only apply this to cases of incontrovertible evidence," he said. "I should also add that our internal process requires first a moderator to document the problem and then for the Section Chair to confirm before imposing the penalty."

A growing number of writers are leaving Substack for alternatives most people haven't heard of like Ghost, Beehiiv, Patreon, and Passport. The reason, writes The Verge's Emma Roth, is the "platform's increased focus on social features as well as a pricing model that puts a chokehold on their business." From the report: Sean Highkin, the creator of the NBA-focused publication The Rose Garden Report, tells The Verge that he makes "significantly more money" after switching from Substack to Ghost last April. "When I first joined up, [Substack] gave me a big push and featured me and funneled a lot of traffic to me, which led to a good amount of growth," Highkin says. "But once I wasn't one of the 'new recruited talent' they could tout, they stopped featuring me and I saw my growth stagnate." Highkin now pays $2,052 per year using Ghost and an add-on called Outpost, compared to $4,968 per year on Substack. The Rose Garden Report's subscriber base has grown 22 percent since the end of 2024, Highkin says. [...]

Substack launched in 2017 as a platform that allows writers to create their own newsletters and manage paying subscribers. Unlike some of its biggest rivals, Substack takes a 10 percent cut of total subscription revenue. That tax may not seem substantial at first, but it quickly adds up as creators gain subscribers and begin charging more for their subscriptions. A calculator on Substack's own website estimates that for a newsletter charging $10 per month with 400 subscribers, the total monthly cost -- including the platform's 10 percent cut and credit card processing fees -- would add up to $636. That cost jumps to $15,900 per month with 10,000 subscribers and skyrockets to $79,500 per month for 50,000 members -- nearly $1 million per year.

Many Substack rivals charge a flat monthly fee, rather than a commission. Ghost, an open-source platform for blogs and newsletters, starts at $15 per month with 1,000 members for website creation, email newsletter capabilities, and a custom domain. Beehiiv, a creator platform with tools for launching a newsletter, website, and podcast, is free for up to 2,500 subscribers with limited access to certain features, like a built-in ad network, while its other plans vary in price based on subscriber count. A person with 10,000 subscribers, for example, will pay $96 per month for Beehiiv's "Scale" plan. There's also Kit, a newsletter platform that offers a tiered pricing model similar to Beehiiv, costing $116 per month with 10,000 subscribers on its "Creator" plan. It's not just the 10% fee critics are complaining about; they also argue the platform offers limited customization and third-party integrations compared to some of the mentioned alternatives, heavily promotes its own branding and social features, and makes creators more dependent on its ecosystem.

Beehiiv founder Tyler Denk argues that creators should be able to build their own brands without the platform taking center stage: "We don't want to take credit for the work of our content creators." While writers can export subscribers, content, and some payment relationships, they cannot take Substack "followers" or Apple-managed iOS billing data with them.