Monthly Archive:: March 2020
OpenBSD aims to be a secure operating system. In the past few months there were quite a few security errata, however. That’s not too unusual, but some of the recent ones were a bit special. One might even say bad. The OpenBSD approach to security has a few aspects, two of which might be avoiding errors and minimizing the risk of mistakes. Other people have other ideas about how to build secure systems. I think it’s worth examining whether the OpenBSD approach works, or if this is evidence that it’s doomed to failure. I picked a few errata, not all of them, that were interesting and happened to suit my narrative.
Honda has done what no other car maker is doing, and returned to analogue controls for some functions on the new Honda Jazz. While most manufacturers are moving to touchscreen controls, identifying smartphone use as their inspiration – most recently seen in Audi’s latest A3 – Honda has decided to reintroduce heating and air conditioning controls via a dial rather than touchscreen, as in the previous-generation Jazz. Unlike what the introduction states, Honda joins fellow Japanese car maker Mazda in not just blindly using touchscreens for everything inside cars. This is a good move, and definitely takes some guts, since I’ve seen countless car reviewers – including my standout favourite, Doug DeMuro – kind of blindly assuming that any car without 100% touchscreen control is outdated, without questioning the safety consequences. Good on Honda.
A work-in-progress patch series was posted over the weekend for adding variable refresh rate support into Mutter for X.Org and Wayland. This includes checking for VRR support from connected monitors using the DRM properties, support for activating VRR, and the ability to toggle the VRR support via a DBus API. The VRR support isn’t advertised to Wayland clients at the moment for the lack of an upstream Wayland protocol around VRR. I can’t wait for Mutter and Kwin to adopt and integrate support for variable refresh rates, so seeing these first patches is good news.
When AMD introduced its Ryzen 4000 mobile CPUs at CES, the company made bold claims of game-changing performance. Coming off of years of underwhelming laptop chips, AMD promised it had optimized Ryzen 4000 for mobile computing. Now we’ve tested those claims in AMD’s Ryzen 9 4900HS chip, an 8-core, 7nm chip with Radeon Vega cores. We’re stunned at the CPU’s impressive tour de force that defeats just about every Intel 8th- and 9th-gen laptop CPU we’ve ever seen. Just open up your YouTube feed and you’ll see pretty much every PC hardware channel staring at disbelief in just how good AMD’s Ryzen 4000 mobile processors really are. This isn’t just a “kind of good enough” processor – the top of the line model is faster than or equal than Intel’s top of the line processor at both single core and multicore workloads, while using slightly more than half the power. It’s all well and good for AMD to roundly run circles around Intel in the server and desktop/workstation space, but the laptop space is where the real money and mindshare can be found. This new line of AMD mobile processors is simply stunning.
Earlier this evening, Linus released Linux 5.6, which contains our first release of WireGuard. This is quite exciting. It means that kernels from here on out will have WireGuard built-in by default. And for those of you who were scared away prior by the “dOnT uSe tHiS k0de!!1!” warnings everywhere, you now have something more stable to work with. The last several weeks of 5.6 development and stabilization have been exciting, with our codebase undergoing a quick security audit, and some real headway in terms of getting into distributions. WireGuard is probably the biggest new feature in 5.6, announced earlier today.
You may have seen dark rumors around the Web that Microsoft is about to kill off the classic Control Panel. Rest assured, friend, we were as horrified as you are—but on more careful inspection, this seems not to be the case. That’s one of the many downsides of being at the mercy of closed operating systems like Windows or macOS – as a user, you’re not really in control, and your platform landlords can decide to remove vital functionality or features on a whim, and there’s nothing you can do about it. If you haven’t done so yet, I’d highly suggest start looking at open source alternatives before it’s too late, because I feel the noose is only going to tighten more, not less.
Here you’ll find my complete set of posts covering the Amiga Machine Code course. The course consists of twelve letters and two disks, that can be found here. The letters are available as PDF’s in their original Danish language as well as translated to English. Some light reading for the weekend.
If you have music on a collection of MiniDisc media and want to finally copy the data off onto modern media (or the cloud!), here are simple instructions for some different solutions. Why would you stop using MiniDisc though?
Ars Technica reports on a story from the early 2000s 2020: When software and operating system giant Microsoft announced its support for inclusion of the exFAT filesystem directly into the Linux kernel back in August, it didn’t get a ton of press coverage. But filesystem vendor Paragon Software clearly noticed this month’s merge of the Microsoft-approved, largely Samsung-authored version of exFAT into the VFS for-next repository, which will in turn merge into Linux 5.7—and Paragon doesn’t seem happy about it. Yesterday, Paragon issued a press release about European gateway-modem vendor Sagemcom adopting its version of exFAT into an upcoming series of Linux-based routers. Unfortunately, it chose to preface the announcement with a stream of FUD (Fear, Uncertainty, and Doubt) that wouldn’t have looked out of place on Steve Ballmer’s letterhead in the 1990s. This is some “get the facts” level of tripe. You’d think that in 2020, we’d be spared this sort of nonsense, and I’m sad I’m even spending precious bits on this one – but at least we get the name of Paragon out so you can avoid them like the plague.
AMD has filed at least two DMCA notices against Github repos that carried “stolen” source code relating to AMD’s Navi and Arden GPUs, the latter being the processor for the upcoming Xbox Series X. The person claiming responsibility for the leak informs TorrentFreak that if they doesn’t get a buyer for the remainder of the code, they will dump the whole lot online. I’d love to hear the backstory behind this hack. For a company like AMD, such a hack must’ve been an inside job, right? While I know I shouldn’t be surprised anymore by just how lacking security can be at even the most prominent technology companies, I just can’t imagine it being very easy to get your hands on this documentation and code without some form of inside help.
For those managing to get their hands on a recently released Loongson 3A4000/3B4000 or even older Loongson 3 MIPS64 processors, improving the support is on the way with the upcoming Linux 5.7 kernel. Queued as part of the MIPS architecture work for Linux 5.7 are a number of Loongson improvements, in particular for the Loongson 3 series. The Loongson processors are pretty much impossible to come by outside of China, and gained some fame as the platform of choice for Richard Stallman.
Apple has released macOS 10.15.4, watchOS 6.2, and iOS, iPadOS and tvOS 13.4. Earlier today, Apple continued its tradition of updating all of its operating systems at once. The day brought major new feature releases to iOS, iPadOS, macOS, watchOS, and tvOS. The iOS, iPadOS, and tvOS updates are numbered 13.4, Apple Watches got watchOS 6.2, and Macs saw the release of macOS Catalina 10.15.4. You know where to get them.
Ever wondered what’s it like to run Android without Google’s services and applications? Well, get a Huawei device. A smartphone UI isn’t much use without apps, of course, and here is where Huawei hits its first hurdle. Huawei has its own store called AppGallery, which it claims is the third largest in the world based on its more than 400 million monthly active users. The vast majority of those users will be in China, of course, where the Google Play Store has never been included alongside AppGallery. If you buy a Mate 30 Pro now anywhere in the world, though, AppGallery is what you get out of the box. To be blunt, it is not great. I wouldn’t call it barren — there is support from major US companies like Microsoft, Amazon, and Snap. You can’t get Chrome, of course, but Opera is there if you want something with desktop sync. But a huge amount of its content is aimed at China, with other big Western names like Facebook, Slack, Netflix, and Twitter missing, which puts the Mate 30 Pro in a more precarious app situation than even the diciest days of Windows Phone. Huawei has announced a $1 billion plan to help stock AppGallery’s shelves, but it has its work cut out. A bigger problem is that even if you can get popular applications installed, they often won’t work properly because the device lacks the Google Mobile Services. It’s an incredibly hard situation for Huawei to be in.
Google announced its decision to drop support for the User-Agent string in its Chrome browser. Instead, Chrome will offer a new API called Client Hints that will give the user greater control over which information is shared with websites. We’ve talked about this earlier this year, but I want to highlight it again since it’s very important this initiative doesn’t devolve into Google and Chrome shoving this alternative down the web’s throat. Deprecating user agent strings is a good thing, but only if the replacement is a collective effort supported by everyone.
Update: the WebKit blog post has been updated with a clarification: Web applications added to the home screen are not part of Safari and thus have their own counter of days of use. Their days of use will match actual use of the web application which resets the timer. We do not expect the first-party in such a web application to have its website data deleted. That’s definitely a relief, and good thing they cleared this up. Original continues below: On the face of it, WebKit’s announcement yesterday titled Full Third-Party Cookie Blocking and More sounds like something I would wholeheartedly welcome. Unfortunately, I can’t because the “and more” bit effectively kills off Offline Web Apps and, with it, the chance to have privacy-respecting apps like the prototype I was exploring earlier in the year based on DAT. Block all third-party cookies, yes, by all means. But deleting all local storage (including Indexed DB, etc.) after 7 days effectively blocks any future decentralised apps using the browser (client side) as a trusted replication node in a peer-to-peer network. And that’s a huge blow to the future of privacy. I’m sure that’s entirely a coincidence for a company that wants to force everyone to use their App Store, the open web be damned.
When Apple CarPlay and Android Auto first started rolling out, initial evidence suggested these technologies held promise to reduce distracted driving. These systems funneled the most important features from our phones onto the infotainment screen, curbing motorists’ desire to reach for their handhelds. Yet, it looks like these mirroring technologies may not be nearly as safe as initially hoped. A new study from the UK’s IAM Roadsmart, an independent road safety organization, paints a far bleaker picture. The stark findings showed that drivers using one of the smartphone mirroring systems in a car displayed reaction times slower than someone who’d used cannabis. In fact, these motorists’ reaction times were five times slower than someone driving with the legal limit of alcohol in their system. This shouldn’t come as a surprise to anyone with more than two braincells to rub together. These systems are based on touch screen technology, and touchscreens without any tactility are simply not suited for use while operating a motor vehicle. Touchscreens are far more distracting than plain old tactile buttons in a fixed order that you learn over time and can feel, and it blows my mind that no safety regulations heavily curtailing their use to parked situations has been enacted just yet.
The Counterpoint program launcher was supplied with the Amstrad PC5086 and other Amstrad PCs from that era. It acts as a user-friendly front end, replacing the full GUIs (Windows 2.0, or GEM) supplied with previous models. The Amstrad-branded version opens with a warning that it should only be used on Amstrad computers. However it appears to run successfully in non-Amstrad environments, such as the virtual machine used to make these screenshots. I love discovering user interfaces I’ve never known about this before, and this one fits the bill just right. Wild UI experimentation was the norm during the late ’80s and early ’90s, before we all settled on what we’re all using now. Digging into the past and learning from even relatively obscure footnotes such as these is fascinating.
It came out much later in March than we expected, but yesterday Google launched the second developer preview for Android 11, the next big version of Android due out at the end of the year. Despite the coronavirus disrupting just about every part of normal life, Google posted the same schedule it did with Preview 1, indicating that the plan is still to have a preview release every month. Anyway, here are the important new things in this release. As always, an excellent look at the new features by Ars. We’re still early on in Android 11’s development cycle, though, so everything is still very much subject to change.
Usually, x86 tutorials don’t spend much time explaining the historical perspective of design and naming decisions. When learning x86 assembly, you’re usually told something along the lines: Here’s EAX. It’s a register. Use it. So, what exactly do those letters stand for? E–A–X. I’m afraid there’s no short answer! We’ll have to go back to 1972… I love digital archeology.
We have seen earlier that Microsoft’s designers are working on a new Start Menu for Windows 10 (not Windows 10X) and now Panos Panay has posted a video celebrating 1 billion Windows 10 installations which appears to confirm that the changes and more are on the way. There’s finally hopefully going to be a modern replacement for Explorer, and context menus seem to be modern and thus consistent too. The already mentioned updated Stert menu is coming, too.