«Il est en train de mourir»: dans le Pas-de-Calais, les images insoutenables de l’interpellation de Zakariyya
«Mediapart» révèle les images de l’interpellation brutale de Zakariyya dans le Pas-de-Calais. On le voit faire plusieurs malaises sous les mains de policiers. Selon nos informations, le parquet de Béthune a ouvert une enquête pour violences volontaires par personne dépositaire de l’autorité publique. ⌘ Read more

Flattened Image Tree 1.0 Specification For Embedded Linux Systems
The Flattened Image Tree “FIT” 1.0 specification was recently finalized for this container format used by U-Boot on embedded systems for providing various boot components like DTBs, the Linux kernel image, and more into a single file… ⌘ Read more

Security updates for Thursday
Security updates have been issued by AlmaLinux (dovecot, fence-agents, freeipmi, git-lfs, image-builder, kernel, libsoup, osbuild-composer, and python-tornado), Debian (apache2, libdatetime-timezone-perl, lrzip, tzdata, and wireshark), Fedora (dovecot, forgejo-runner, gh, gnutls, krb5, nano, pdns, pyOpenSSL, squid, vim, and xorg-x11-server-Xwayland), Mageia (graphicsmagick, kernel-linus, krb5-appl, libexif, libtiff, nano, nginx, ntfs-3g, opam, perl-Net-CIDR-Lite, perl-Starlet, perl-Starma … ⌘ Read more

A Kid With a Fake Mustache Tricked an Online Age-Verification Tool
To stop children from bypassing its age checks, Meta is revamping its age-verification tools with an AI system that analyzes images and videos for “visual cues,” such as height and bone structure. ⌘ Read more

ReactOS Unifies Installation Media, Introduces GUI Installer and New ATA Driver
jeditobe writes: Developers of ReactOS told Phoronix that the project has introduced a unified BootCD, replacing its previously separate installation media and LiveCD images. The new image combines the traditional text-mode installer with a LiveCD mode in a single medium. Within this unified BootCD, the updated LiveCD … ⌘ Read more

Security updates for Wednesday
Security updates have been issued by AlmaLinux (corosync, dovecot, image-builder, python-tornado, resource-agents, and systemd), Debian (openjdk-11, openjdk-17, and pyjwt), Fedora (pdns, pyOpenSSL, and squid), Slackware (hunspell), SUSE (alloy, avahi, bubblewrap, cmctl, coredns, curl, dpkg, firefox, golang-github-prometheus-prometheus, grafana, libpng12, PackageKit, sed, and xen), and Ubuntu (docker.io-app, nghttp2, python-django, and python-mako). ⌘ Read more

Retina Scan for Diabetes Could Also Reduce Deaths During Pregnancy in Developing Countries
This week Bill Gates wrote a blog post about a special camera from medtech startup Remidio, which delivers high-resolution images of a patient’s retina in seconds. The camera plugs into a phone running an AI system that watches for early signs of diabetes — all without needing a blood draw, e … ⌘ Read more

Sanctioned Chinese AI Firm SenseTime Releases Image Model Built for Speed
With US restrictions limiting its access to advanced tech, SenseTime is doubling down on open source with a new model optimized to run on Chinese-made chips. ⌘ Read more

South Korea Police Arrest Man For Posting AI Photo of Runaway Wolf
South Korean police arrested a man accused of spreading an AI-generated image of an escaped wolf, after the fake photo reportedly misled authorities and disrupted the real search operation. The BBC reports: South Korean police have arrested a man for sharing an AI-generated image that misled authorities who were searching for a wolf that had br … ⌘ Read more

@bender@twtxt.net Yeah, it’s crazy!

Somehow, your link 403s here, but I just searched it. At least it has the diagonals at the lower two steps. However, the two upper platforms also suffer from the plastic covers, it appears (I cannot tell the material from the low quality images I found). Maybe it is aluminium? I think some joints use machine bolts, though (but again, not enough detail visible).

Happy ladder climbing!

无需科学上网，国内丝滑使用最新GPT-Image-2！
相信大家这两天都被最新的「Image-2」图像模型刷屏了吧？ 就在昨天，椒图AI已经在第一时间上线该模型，欢迎大家前往椒图AI切换模型体验！ 根据官方排行榜给出的最新消息，Image-2的能力甚至超越了大香蕉，实测也是相当惊艳！ 使用地址：www.jiaotuai.cn 像下图这种真实的直播间截图↓ ⌘ Read more

ChatGPT 图像 2.0 正式发布了，这次有点强
前几天在这个帖子里分享了几张图片，就是灰度到的 ChatGPT Images 2.0，足以以假乱真的截图（AI生成）： 今天 ChatGPT Images 2.0 正式发布了，效果…非常赞啊。 只需要极短的 Prompt，就可以生成非常高质量的图片，而且很多图片，越来越感受不到是 AI 生成了，足以 ⌘ Read more

RADV Driver Enables Host Image Copy By Default For RDNA2 & Newer
Introduced back in 2023 with Vulkan 1.3.258 was VK_EXT_host_image_copy to copy data between host memory and images on the host processor without needing to stage the data through a CPU-accessible buffer. This direct CPU-to-GPU image data transfer path can reduce memory usage during asset loads and all around more efficiency and performance. Finally now the RADV open-source Radeon driver is enabling support by default… ⌘ Read more

Arch Linux now has a reproducible container image
Robin Candau has announced
the availability of a bit-for-bit reproducible container image for
Arch Linux:

The bit-for-bit reproducibility of the image is confirmed by digest
equality across builds ( podman inspect --format '{{.Digest}}' <image>) and by running diffoci
to compare builds. We provide d … ⌘ Read more

GIMP 3.2.4 Released With A Fix For Its XCF Code That Has Existed Since 1999
Following last month’s GIMP 3.2 feature release that was followed by the GIMP 3.2.2 point release at the end of March, out now is GIMP 3.2.4 to ship more fixes to users of this open-source alternative to Adobe Photoshop and other imaging applications… ⌘ Read more

As an enjoyer of delightfully bad graphic design, found on most Czech village center cork boards, I’m sad to see the stolen clipart and badly cropped watermarked stock images, gradually replaced with AI slop.

This is far from a serious rant, but generating images of my kind being telepathically hit with sharp rocks, surely gives me a right to complain.

So far these seem the most prominent slop categories, seem to be…

Architecture slop:

1. find a sketch of what an old building looked like
   
   
2. generate an AI version, without correcting any of the perspective errors - this one is diagonally levitating
   
   
3. generate a recreation of the buildings demise - after going through the AI, for the second time, it is now a completely different building
   
   

Moralizing slop:

History slop:

OpenAI’s Big Codex Update Is a Direct Shot At Claude Code
OpenAI is updating Codex with more agent-like capabilities, positioning it as a more direct rival to Anthropic’s Claude Code. Some of the new features include the ability to operate macOS desktop apps, browse the web inside the app, generate images, use new workplace plug-ins, and remember useful context from past tasks. The Verge reports: Codex will now … ⌘ Read more

[$] Tagging music with MusicBrainz Picard
Part of the “fun” that comes with curating a self-hosted music library is tagging
music so that it has accurate and uniform metadata, such as the band names, album titles,
cover images, and so on. This can be a tedious endeavor, but there are quite a few
open-source tools to make this process easier. One of the best, or at least my
favorite, is MusicBrainz Picard. It is
a cross-platform music-tagging application that pulls information from the
well-curated, … ⌘ Read more

Security updates for Tuesday
Security updates have been issued by Debian (gdk-pixbuf, gst-plugins-bad1.0, and xdg-dbus-proxy), Fedora (chromium, deepin-image-viewer, dtk6gui, dtkgui, efl, elementary-photos, entangle, flatpak, freeimage, geeqie, gegl04, gthumb, ImageMagick, kf5-kimageformats, kf5-libkdcraw, kf6-kimageformats, kstars, libkdcraw, libpasraw, LibRaw, luminance-hdr, nomacs, OpenImageIO, OpenImageIO2.5, photoqt, python-cryptography, rawtherapee, shotwell, siril, swayimg, vips, and webkitgtk), Red Hat (firefox an … ⌘ Read more

Mark Zuckerberg Is Reportedly Building an AI Clone To Replace Him In Meetings
According to the Financial Times, Meta is developing an AI avatar of Mark Zuckerberg that could interact with employees using his voice, image, mannerisms, and public statements, “so that employees might feel more connected to the founder through interactions with it.” The Verge reports: Meta may start allowing creators … ⌘ Read more

当「豆包AI生成」出现在一张重要照片上，会发生什么？
我最近开始认真对待图片上的水印，是因为一件有点特别的事情。 为什么要去水印 说起来，就在前不久。去参加亲人的葬礼（不要害怕，喜丧），当时请的人一条龙服务，也包括遗像。 是的，你想到了吧，遗像的右下角就明晃晃的写着「豆包AI生成」几个大字。 真的很大，因为照片是放大打印出来的，比豆包原版的水印字体还要 ⌘ Read more

TIL that SSH actually stands for Secure Snake Home, a massively multiplayer snake game playable via the SSH protocol: ssh snakes.run

Of course, no one else was online when I was playing, so…

Image

Christina Koch looking at Earth is my new wallpaper:

(Sorry, forgot where I originally found the image. Some NASA photo collection.)

Security updates for Thursday
Security updates have been issued by Debian (firefox-esr, postgresql-13, and tiff), Fedora (bind, bind-dyndb-ldap, cef, opensc, python-biopython, python-pydicom, and roundcubemail), Slackware (mozilla), SUSE (ckermit, cockpit-repos, dnsdist, expat, freerdp, git-cliff, gnutls, heroic-games-launcher, libeverest, openssl-1_1, openssl-3, polkit, python-poetry, python-requests, python311-social-auth-app-django, and SDL2_image-devel), and Ubuntu (dogtag-pki, gdk-pixbuf, linux, linux-aws, … ⌘ Read more

Planet Labs Tests AI-Powered Object Detection On Satellite
BrianFagioli writes: Artificial intelligence has now run directly on a satellite in orbit. A spacecraft about 500km above Earth captured an image of an airport and then immediately ran an onboard AI model to detect airplanes in the photo. Instead of acting like a simple camera in space that sends raw data back to Earth for later analysis, the satellite perfo … ⌘ Read more

Chrome 148 Will Start ‘Lazy Loading’ Video and Audio to Improve Performance
“Google has announced that it’s currently testing a new feature for Chrome 148 that could speed up day-to-day browsing,” reports PC World:

[T]he browser can intelligently postpone the loading of certain elements. Why load all images at the start when it can instead load images as you get close to them while scrolling? Chrome and … ⌘ Read more

Gentoo Releases Experimental Images Using GNU/Hurd
Following an April Fools’ Day tease of Gentoo claiming they were going to switch to GNU Hurd as their primary kernel moving forward, they have now acknowledged the joke but in fact also announcing there are now experimental Gentoo GNU/Hurd images available… ⌘ Read more

CEO of America’s Largest Public Hospital System Says He’s Ready To Replace Radiologists With AI
Mitchell H. Katz, MD, president and CEO of NYC Health + Hospitals, said hospitals could already replace many radiologists with AI for some imaging tasks – if regulators allowed it. He argued the technology presents an opportunity to simultaneously cut costs and expand access. Radiology Busin … ⌘ Read more

@lyse@lyse.isobeef.org For reasons I can’t fully explain, we have a bunch of courses in the area, most in public parks (they integrate nicely since they can be built with the existing landscape, only adding some yellow baskets, concrete starting pads, and maybe signs).

In my experience, the main difference between a disc golfer and a frisbee thrower is that the disc golfer will often have a bag full of different shapes of discs (including drivers of varying ranges and/or putters). Even in my small bag, I’ve got some long range drivers (a Beast, a Cheetah, a Valkyrie, and a Wraith), my aforementioned MRV (Mid-Range Vector), an ultralight Aero (which feels similar to a “standard” frisbee), and 2 “rubber” putters (softer plastic, less “bouncy”).

Image

@bender@twtxt.net Thanks, and you are correct: MRV stands for Mid-Range Vector (I think) as it is a stable mid-range driver (it says so on the disc):

Image

@lyse@lyse.isobeef.org Thank you for the suggestions. I will probably do some of that when I have time. For the thumbnails, I’m also thinking about trying the loading=“lazy” img attribute. Top on my mind is actually understanding why the big images don’t load. Maybe my VPS’s network connection is saturated, for example. I’ve never needed to worry about such things until now. I’m looking forward to spending some time on it.

4Chan Mocks $700K Fine For UK Online Safety Breaches
The UK regulator Ofcom fined 4chan nearly $700,000 (520,000 pounds) for failing to implement age checks and address illegal content risks under the Online Safety Act, but the platform mocked the penalty and signaled it won’t pay. A lawyer representing the company responded with an AI-generated cartoon image of a hamster, writing in a follow-up post on X: “In the only coun … ⌘ Read more

‘Pokemon Go’ Players Unknowingly Trained Delivery Robots With 30 Billion Images
More than 30 billion images captured by Pokemon Go players have helped train a visual mapping system developed by Niantic. The technology is now being used to guide delivery robots from Coco Robotics through city streets where GPS often struggles. Popular Science reports: This week, Niantic Spatial, part of the tea … ⌘ Read more

@lyse@lyse.isobeef.org Thanks for letting me know. HTML checkers seem happy now. I’m not sure what to do about the images not loading. The photos have three sizes (thumbnail, photo page, and original if you click the img tag on the photo page); can you at least see the smaller two sizes? Maybe I will do some experimental fetches and/or start measuring things on my web server.

Meta To Charge Advertisers a Fee To Offset Europe’s Digital Taxes
Meta will begin charging advertisers a 2-5% “location fee” to offset digital services taxes imposed by several European countries, including the UK, France, Italy, Spain, Austria, and Turkey. Reuters reports: The fee, for image or video ads delivered on Meta platforms including WhatsApp click-to-message campaigns and marketing messages together w … ⌘ Read more

After US-Israel Attacks, 90 Million Iranians Lose Internet Connectivity
CNN reports that images from Iran’s capital “have shown cars jammed along Tehran’s street, with heavy traffic on major roads after today’s wave of attacks by the US and Israel.” And though Iran has a population of 93 million, the attacks suddenly plunged Iran into “a near-total internet blackout with national connectivity at 4% of or … ⌘ Read more

FreeBSD 14.4-RC1 Adds Emacs, Vim & More To DVD Images
For those on the current FreeBSD 14 series with no immediate plans to move to FreeBSD 15 that debuted at the end of 2025, FreeBSD developers have been preparing for the release of FreeBSD 14.4. Released overnight was the first release candidate of FreeBSD 14.4… ⌘ Read more

Google Launches Nano Banana 2 Model With Faster Image Generation
Google has launched Nano Banana 2 (Gemini 3.1 Flash Image), a faster, more realistic image generation model that becomes the default across Gemini, Search, Lens, and Flow. TechCrunch reports: The new Nano Banana 2 retains some of the high-fidelity characteristics of the Pro model but produces images faster. The company says you can create images … ⌘ Read more

Firefox 148 Now Available With The New AI Controls, AI Kill Switches
Firefox 148 introduces granular AI controls and a global “AI kill switch” that allows users to disable or selectively manage the browser’s AI features. Phoronix reports: Among the AI features that can be toggled individually are around translations, image alt text in the Firefox PDF viewer, tab group suggestions, key points in link preview … ⌘ Read more

AOMedia Open Audio Codec “OAC” Aims To Be The Successor To Opus
While the Alliance For Open Media “AOMedia” is most known for developing the AV1 open video codec, the associated AV1 Image File Format (AVIF), and the next-generation AV2, they are now working on the Open Audio Codec (AOC)… ⌘ Read more

Pinterest Is Drowning in a Sea of AI Slop and Auto-Moderation
Users say Pinterest has become flooded with AI-generated images and heavy-handed automated moderation, with artists reporting wrongful takedowns and their hand-drawn work mislabeled as “AI modified.” As the company doubles down on AI features and layoffs, longtime users argue the platform’s creative ecosystem is being undermined. 404 Media reports: “I … ⌘ Read more

Google’s AI Music Maker Is Coming To the Gemini App
Google is bringing its Lyria 3 AI music model into the Gemini app, allowing users to generate 30-second songs from text, images, or video prompts directly within the chatbot. The Verge reports: Lyria 3’s text-to-music capabilities allow Gemini app users to make songs by describing specific genres, moods, or memories, such as asking for an “Afrobeat track for my mot … ⌘ Read more

WordPress Gets AI Assistant That Can Edit Text, Generate Images and Tweak Your Site
WordPress has started rolling out an AI assistant built into its site editor and media library that can edit and translate text, generate and edit images through Google’s Nano Banana model, and make structural changes to sites like creating new pages or swapping fonts.

Users can also invoke the assistant by taggin … ⌘ Read more

99% of Adults Over 40 Have Shoulder ‘Abnormalities’ on an MRI, Study Finds
Up to a third of people worldwide have shoulder pain; it’s one of the most common musculoskeletal complaints. But medical imaging might not reveal the problem – in fact, it could even cloud it. From a report: In a study published in JAMA Internal Medicine this week, 99 percent of adults over 40 were found to have at least one … ⌘ Read more

ByteDance Suspends Seedance 2 Feature That Turns Facial Photos Into Personal Voices Over Potential Risks
hackingbear writes: China’s Bytedance has released Seedance 2.0, an AI video generator which handles up to four types of input at once: images, videos, audio, and text. Users can combine up to nine images, three videos, and three audio files, up to a total of twelve fi … ⌘ Read more

Google Chrome 145 Released With JPEG-XL Image Support
Back in 2022 Google deprecated and then removed JPEG-XL image support from the Chrome/Chromium browser codebase and now in 2026 it’s back. Last month I wrote about JPEG-XL decoding merged back to Chromium/Chrome and that has rolled out today as part of today’s Chrome 145 stable debut… ⌘ Read more

GNOME’s Glycin 2.1 Beta Enables JPEG 2000 Support By Default
GNOME’s Rust-based and sandboxed Glycin image loading library focused on safety now supports JPEG 2000 images by default… ⌘ Read more

OpenIndiana Is Porting Solaris’ IPS Package Management To Rust
OpenIndiana as the open-source project built atop Illumos that is continuing to maintain and advance the former OpenSolaris code is working on a big ambitions of modernizing the Image Packaging System (IPS) package management solution. As part of that they are working to move from a C and Python codebase over to Rust… ⌘ Read more

Libcamera 0.7 Released - GPU Acceleration Support For SoftISP Can Deliver 15x Performance
Libcamera 0.7 was published today for this modern software library for image signal processors (ISPs) and embedded cameras under Linux. The standout change with libcamera 0.7 is initial plumbing for GPU acceleration in the software ISP “SoftISP” for delivering better performance than just CPU-based… ⌘ Read more

Mesa NVK Driver Temporarily Disabling Support For Larger Pages Due To Bug
Upstreamed to the Nouveau open-source kernel driver in Linux 6.19 was support for larger pages and with that compression support available with the larger page sizes. Subsequently the Mesa NVK open-source Vulkan driver began making use of the larger pages and compressed image support dependent upon the larger page sizes as it should help with performance. But for now it’s being temporarily disabled due to a discovered issue… ⌘ Read more