OpenBSD 7.8 released 22 Oct 2025, 2:24 pm

Like clockwork, every six months, we have a new OpenBSD release. OpenBSD 7.8 adds support for the Raspberry Pi 5, tons of improvements to sleep, wake, and hibernate, the TCP stack can now run in parallel on multiple processors, and so much more. DRM has been updated to match Linux 6.12.50, and drivers for the Qualcomm Snapdragon DRM subsystem and Qualcomm DisplayPort controller were added as well.

The changelog is, as always, long and detailed, so head on over for the finer details. OpenBSD users will know how to upgrade, and new users can visit the download page.

What about the icons in pifmgr.dll? 22 Oct 2025, 2:17 pm

Raymond Chen has another great post about some of the classic icons from Windows 95, this time focusing on pifmgr.dll. In this file, there are a variety of random-seeming icons, and it turns out they’re random for a reason: they were just a bunch a fun, generic icons intended for people to use when creating PIF files.

The icons in pifmgr.dll were created just for fun. They were not created with any particular programs in mind, with one obvious exception. They were just a fun mix of icons for people to use for their own homemade shortcut files.

↫ Raymond Chen at The Old New Thing

For those of us who didn’t grow up with Windows, or who, god forbid, are too young to know, PIF or personal information files are effectively shortcuts to DOS programs for use in a multitasking environment. A PIF file would not only point to the relevant DOS executable, but also contain information about the environment in which said executable was supposed to run. Their history goes back to IBM’s TopView, and Microsoft later embraced and adapted them for use in Windows.

Understanding driver updates through Windows Update 22 Oct 2025, 2:05 pm

Microsoft has published a set of short questions and answers about driver updates through Windows Update, and there’s one tidbit in there I found interesting.

Driver dates might look old, but that is not true. The driver date is descriptive info set by the driver provider and can be any date they choose. When determining which driver to install, Windows Update uses targeting information set by the provider inside the driver files to determine the best driver. This lets the device provider promote the best driver, regardless of the chosen date. 

↫ Microsoft knowledge base article

Whenever I do have to fiddle with Windows machines, I always wondered about why some drivers in Windows Update would show some seriously old dates. It turns out the answer is as obvious as it always tends to be: OEMs.

KDE Plasma 6.5 released 21 Oct 2025, 5:28 pm

KDE is on a roll lately, and keeps on rolling with today’s release of KDE Plasma 6.5. As the project itself notes, this release focuses on relatively small improvements, refinements, and other niceties, without making any massive changes. With Linux desktops taking accessibility more seriously lately than ever before, I want to focus on the accessibility improvements first.

The Orca screen reader now announces caps lock state changes, and screen readers will now describe the Shortcuts and Autostart pages more optimally. There’s also a new grayscale colour filter for people sensitive to colours, developers have done Plasma-wide pass to eliminate bright flashes in the UI, and the desktop zoom feature will now follow the text insertion point as it moves around the UI. Keyboard navigation in various parts of Plasma have been improved, and a few other small changes have been to improve accessiblity.

Other changes include rounded bottom window corners (which can be turned off), automatic and scheduled theme and wallpaper transitions (e.g. from light to dark), and a new and improved applications permissions settings panel. A small new feature that will be a massive time saver for me is the ability to favourite items in your clipboard history, so they remain available over time. I reuse certain copied bits of text all the time, and I can’t wait to start using this little addition.

Remote desktop has also received a ton of love in Plasma 6.5. You can now share your clipboard, and you no longer need to create dedicated RDP user accounts; you can just log in with your normal account credentials as you would expect you could. Plasma’s Discover application, used for application and update management, has seen major work to improve its performance – very welcome, for sure. Of course, there’s a ton of other changes, too.

KDE Plasma 6.5 will find its way to your distribution soon enough.

Intel, AMD to bring memory tagging to x86, at some point 21 Oct 2025, 3:21 pm

Now that ARM’s memory tagging, used extensively by Android ROMs such as GrapheneOS and now also by Apple, is becoming the new norm to aid in improving memory safety, the x86 world can’t sit idly by. As such, Intel and AMD have announced a ChkTag, x86’s version of memory tagging.

ChkTag is a set of new and enhanced x86 instructions to detect memory safety violations, such as buffer overflows and misuses of freed memory (use-after-free). ChkTag is designed to be suitable for hardening applications, operating system kernels, hypervisors for virtualization, and UEFI firmware. ChkTag places control in the software developers’ hands to balance their security needs with operational elements that often become prominent when deploying code. For example, ChkTag provides instruction-granular control over which memory accesses are checked. Compilers can offer optimizations and new language features or intrinsics. ChkTag prepares x86 for a future with increasing amounts of code written in memory-safe languages running alongside code in other languages. Furthermore, ChkTag loads tags from linear/virtual memory that can often be committed on demand.

↫ Intel and AMD’s announcement

It’s important to note that ChkTag – why not just call it CheckTag – isn’t ready yet, nor is there any indication when it will be included in any processors from Intel and AMD. The goal is to catch certain memory safety problems in hardware. According to Intel and AMD’s shared announcement, developers will have fine-grained control over the feature, allowing them to tap into the functionality in whatever way they deem necessary or valuable for their software in specific circumstances.

My fear is that Intel and AMD will use this feature as a product differentiator, restricting it to either more expensive processors or to Xeon/Threadripper processors, thereby fracturing the market. This would inevitably lead to spotty support for the feature across the x86 landscape, meaning most ordinary consumer won’t benefit from it at all.

This is how much Anthropic and Cursor spend on Amazon Web Services 21 Oct 2025, 2:49 pm

I can exclusively reveal today Anthropic’s spending on Amazon Web Services for the entirety of 2024, and for every month in 2025 up until September, and that that Anthropic’s spend on compute far exceeds that previously reported. 

Furthermore, I can confirm that through September, Anthropic has spent more than 100% of its estimated revenue (based on reporting in the last year) on Amazon Web Services, spending $2.66 billion on compute on an estimated $2.55 billion in revenue.

↫ Ed Zitron

These numbers do not even include what the company spends on Google’s services. Going through all the numbers and reporting, Zitron explains that the more “successful” Anthropic becomes, the bigger the gap between income from paying customers and its spending on Amazon and Google services becomes. It’s simply unsustainable, and the longer we keep this scam going, the worse the consequences will be when the bubble pops.

Sadly, nobody will go to jail once hell breaks loose.

Cartridge chaos: the official Nintendo region converter and more! 21 Oct 2025, 2:25 pm

This post is a combination of looks at several oddities among my pile of NES and Famicom cartridges. Why, for example, do I have a copy of Gyromite when I don’t have a R.O.B.? Did I miss something interesting in my MMC blog post? And while it is the Japanese release of Kid Niki: Radical Ninja, is my Kaiketsu Yanchamaru being a little too radical? Who put the ram in the rama-lama-ding-dong? Some of these questions will be answered!

↫ Nicole Branagan at Nicole Express

A well-written post with tons of weird NES nerdery. Branagan delivers, every time.

Microsoft breaks USB input in Windows Recovery Environment 20 Oct 2025, 9:03 pm

With official support for Windows 10 having officially ended a few days ago, let’s take a look and see how its successor, Windows 11, is doing.

Microsoft released the first Patch Tuesday update (KB5066835) for Windows 11 25H2 this past week and it is probably fair to say that it has been a rough start for the new feature update. Despite the announcement of a wide rollout wherein the new version is now available for download for everyone, the company has already confirmed large-scale issues.

First up, Microsoft was forced to issue an emergency workaround as the update broke localhost auth and following that the company today has confirmed another problem where recovery can become impossible if you happen to use a USB keyboard or mouse.

↫ Sayan Sen at Neowin

Yes. This is a real thing. This latest round of patches makes it entirely impossible to navigate the Windows Recovery Environment with USB keyboards and mice. Since it’s 2025, USB is probably the protocol through which most people connect their keyboard and mice (although to be fair, some laptops probably still default to internal PS/2 for their touchpads). This means that if you run into a problem with Windows 11 that requires you to access the Windows Recovery Environment – perhaps OneDrive did too many lines of cocaine again – you can’t actually do anything inside of it.

There’s no fix yet, so you either remove the offending patches, hope your PC still has a PS/2 port and you still have PS/2 peripherals, or hope Windows 11 won’t fall over and die until Microsoft releases a fix for the issue. Of course, people still using Windows 10, people who aren’t installing every single Windows 11 update as they become available, and people using real operating systems have nothing to worry about.

You can’t help but wonder, though – with Microsoft pushing “AI” so hard, how many of these recent faceplants are the result of Microsoft engineers frantically trying to meet code quotas using Copilot?

Servo 0.0.1 released 20 Oct 2025, 8:48 pm

Today, the Servo team has released new versions of the servoshell binaries for all our supported platforms, tagged v0.0.1. These binaries are essentially the same nightly builds that were already available from the download page with additional manual testing, now tagging them explicitly as releases for future reference.

↫ Servo’s official blog

Servo is making steady progress, and that’s awesome news. Every month a whole slew of new features and improvements make their way into this new browser engine, and I’m fairly confident Servo is our best shot at regaining some independence from Google and Apple in the web browser space. Other efforts are either too limited in scope, targeting only a specific niche, already being eaten alive by massive corporations, written in non-memory safe languages, run by people whose code I wouldn’t even trust to flush my toilet, or any combination thereof.

Servo is it, folks. Our best shot.

“I remember taking a screen shot of a video, and when I opened it in Paint, the video was playing in it!” 20 Oct 2025, 8:37 pm

In older versions of Windows, if you had a video playing, took a screenshot, and pasted that screenshot into Paint, you could sometimes see the video continue to play inside Paint. What kind of sorcery enabled this to happen? A few of you will realise instantly why this used to happen: render surfaces. Back in at least the Windows 9x days, playing video involved drawing solid green where you wanted the video to go (the video player window), rendering the video pixels to a surface shared with the graphics card, and then have the graphics card replace said green pixels with the video pixels from the shared surface.

This approach has a whole array of benefits, not least of which is that it allowed you to render the video on a thread separate from the main user interface, so that if the main interface was sluggish or locked up, the video would keep rendering properly. You could also create two shared surfaces to render multiple frames at once, thereby eliminating tearing. Knowing this, it should be obvious what’s going on with the screenshot and Pain story.

Now, when you load the image into Paint or any other image viewer, Windows sends those green pixels to the video card, but if the media player is still running, then its overlay is still active, and if you put Paint in the same place that the media player window is, then the green pixels in Paint get changed into the pixels of the active video. The video card doesn’t know that the pixels came from Paint. Its job is to look for green pixels in a certain region of the screen and change them into the pixels from the shared surface.

If you move the Paint window to another position where it doesn’t overlap the media player, or if the media player isn’t playing a video, you will see the bitmap’s true nature: It’s just a bunch of green pixels.

↫ Raymond Chen at The Old New Thing

I’ve never had this particular oddity happen, but I do have vague memories of video player windows rendering tons of green artifacts whenever something went wrong with the video player, the file it was trying to play, or whatever else, and I guess the cause of those green artifacts is the same. In modern operating systems, graphics rendering of the UI is done entirely on the GPU, with only the final composition being sent to your display.

As such, the green screen effect no longer occurs.

The early Unix history of chown() being restricted to root 20 Oct 2025, 7:57 pm

Chris Siebenmann with another interesting look at a tiny detail of UNIX history.

A few years ago I wrote about the divide in chown() about who got to give away files, where BSD and V7 were on one side, restricting it to root, while System III and System V were on the other, allowing the owner to give them away too.

[….]

The answer is that the restriction was added in V6, where the V6 chown(2) manual page has the same wording as V7. In Research Unix V5 and earlier, people can chown(2) away their own files; this is documented in the V4 chown(2) manual page and is what the V5 kernel code for chown() does. This behavior runs all the way back to the V1 chown() manual page, with an extra restriction that you can’t chown() setuid files.

↫ Chris Siebenmann

The deeper levels of this particular rabbit hole need more exploring, though, as eventually Siebenmann hits a roadblock when trying to figure out why, exactly, the restriction was added, and why certain versions chose to not adopt the new restriction. This may be part of the lore of UNIX we won’t uncover, until one of the people involved speaks up.

Windows 11, now with even more “AI” where you don’t want it 17 Oct 2025, 9:32 pm

Microsoft has posted a blog post about detailing its latest round of additions to Windows 11, and as will surely not surprise you, it’s “AI”, all the time, whether you like it or not. I’m not even going to detail most of these “features”, as I’m sure most of them will just become yet another series of checkboxes on whatever debloating tool you prefer. Still, there’s one recurring theme running throughout Microsoft’s recent “AI” marketing that really stands out, and this blog post is no different:

Until now, the power of AI has often been gated behind your skill at prompting.  The more context you provide and detail you share, the richer response you receive in return. But typing it out can be tedious and time consuming, especially if it takes multiple tries to get it right.

With 68% of consumers reporting using AI to support their decision making, voice is making this easier.

↫ Yusuf Mehdi at the Windows Blogs

“You’re holding it wrong” has become a recurring meme whenever someone places the blame for a shit product on its users, but we’re really starting to see this line of thinking explode with “AI” tools now. If you’re getting bad, wrong, or downright made up results out of your text generator – which happens all the time – the problem isn’t that the text generator is shit; no, the problem is that the user is shit at manipulating and coercing it into generating the right string of words.

This is a major problem for “AI” companies, as the obtuseness of input and the inevitable shoddiness of results is most likely putting users off using them, and if there’s one thing these companies needs, it’s users. All of them are hemorrhaging money without any realistic paths towards profitability, so there’s a mad scramble to convince and trick people into using “AI” tools, and every single recent effort by Microsoft regarding Windows and Office is 100% geared towards this goal. That’s why nothing is sacred, and everything from Notepad to Paint, from the the Windows Start menu to context menus, from the Explorer file manager to your Windows command line is getting Copilot buttons and sparkly icons: Microsoft has to be able to brag about “AI” user numbers to keep the scam going.

As the bubble gets bigger and bigger, and as we come closer and closer to that satisfying pop, you can expect ever more places in Windows to get “AI” features. I can’t wait for the sparkle icon to show up when formatting a disk, installing a driver through Device Manager, or during a kernel panic. I can’t wait for the blue screen of death to open a Copilot chat that advises you to do something utterly unrelated.

You can do it, Microsoft.

A deep dive into the Silicon Graphics Indigo² IMPACT 10000 17 Oct 2025, 7:57 pm

This beautiful purple slab is the Silicon Graphics Indigo² (though, unlike its earlier namesake, not actually indigo coloured) with the upper-tier MIPS R10000 CPU and IMPACT graphics. My recollection was that it worked at the time, but I couldn’t remember if it booted, and of course that was no guarantee that it could still power on. If this machine is to stay working and in the collection, we’re gonna need a Refurb Weekend.

↫ Cameron Kaiser at Old Vintage Computing Research

Out of all the retro UNIX workstations of old, the machines from SGI are both the most popular, the most well-known, and thus, also some of the most expensive. Yet, at the same time, everything up until the very last generation or two of MIPS IRIX workstations, generally do not seem to be particularly rare either. The community around SGI’s machines and IRIX is also quite thriving still, much more so than the communities of the other commercial UNIX variants. Still, the odds of me completing my collection of final-generation commercial UNIX workstations are low, exactly because of just how rare and stupidly expensive the SGI Tezro is.

As always, Cameron Kaiser goes into a level of detail few other people in the world do when it comes to rare or special computers, and this article about the Silicon Graphics Indigo² is no exception. Detailed photographs, an in-depth history of the machine, detailed descriptions of the hardware, the various fixes that needed to be performed, getting it back up and running, and everything else. There’s really nobody else writing these kinds of articles.

The weekend’s here, so sit back, relax, and have fun.

NLnet sponsors development of WPA3 support for OpenBSD 17 Oct 2025, 2:17 pm

The NLnet foundation has sponsored a project to add WPA3 support to OpenBSD, support which in turn can be used by other operating systems.

This project delivers the second open-source implementation of WPA3, the current industry standard for Wi-Fi encryption, specifically for the OpenBSD operating system. Its code can also be integrated by other operating systems to enable modern Wi-Fi encryption, thereby enhancing the diversity and resilience of the global IT ecosystem.

↫ NLnet foundation announcement

WPA3 support in Linux seems to be the only other open source implementation of WPA3, so this is great news not only for OpenBSD, but also for other operating systems who rely on BSD network drivers through compatibility layers, like Haiku. FreeBSD, meanwhile, is planning to build its own WPA3 implementation, so they, too, might benefit form the work that’s going to be done through OpenBSD.

October is listed as the start of this project, so work is probably already underway.

An initial investigation into WDDM on ReactOS 16 Oct 2025, 9:46 pm

One of the problems the ReactOS project continually has to deal with is that Windows is, of course, an evolving, moving target. Trying to be a Windows-compatible operating system means you’re going to have to tie that moving target down, and for ReactOS, the current focus is on being compatible with Windows Server 2003 “or later”. This “or later” part is getting a major boost in a very crucial area.

The history of ReactOS spans a wider range than the lives of many of the people who work on it today. Incredible individuals have come and gone from the project with vastly different goals for what they want to see developed. In recent years, better hardware support has emerged as one of those goals. As ReactOS gazes towards the world of Vista and beyond, a few questions about how hardware works emerge. Vista introduced massive overhauls to how hardware drivers are written and maintained. Gradually we’re trying to handle many of these overhauls with great success. Today we talk about WDDM, or the Windows Display Driver Model.

An initial investigation into WDDM on ReactOS

There’s a ton of technical details in the blog post, but the end result is that ReactOS can now tentatively load some WDDM drivers. For instance, ReactOS can run NVIDIA’s Windows 7 driver now, and the example used an NVIDIA GTX 1070. Of course, we’re looking at basic 2D display output only and no 3D acceleration, so don’t expect to be running any 3D games on ReactOS any time soon.

Still, this is a pretty massive step forward for ReactOS, but of course, a ton more work remains to be done, as is always the case for ReactOS. I do have to say – the fact that WDDM support is now on the table and progress is being made here is great news. ReactOS is not even remotely close to being an alternative to Windows, but even if it never gets there, it’s a great showcase for what talented, determined developers can do, and they deserve recognition for that.

How to turn Liquid Glass into a solid interface 16 Oct 2025, 3:27 pm

Apple’s new Liquid Glass interface design brings transparency and blur effects to all Apple operating systems, but many users find it distracting or difficult to read. Here’s how to control its effects and make your interface more usable. Although the relevant Accessibility settings are quite similar across macOS, iOS, watchOS, and tvOS, I separate them because they offer different levels of utility in each. I have no experience with (or interest in) a Vision Pro, so I can’t comment on Liquid Glass in visionOS.

↫ Adam Engst at TidBITS

An incredibly detailed article showing exactly how to change the relevant settings, and exactly what they do, for each of Apple’s relevant platforms. I have a feeling quite a few of you will want to bookmark this one.