The various Raspberry Pi camera modules have become the default digital camera hacker’s tool, and have appeared in a huge number of designs over the past decade. They’re versatile and affordable, and while the software can sometimes be a little slow, they’re also of decent enough quality for the investment. Making a Pi camera can be annoying though, because different screens, lenses, and modules have their own mounting requirements. [Jacob David C Cunningham] has a solution here, with a modular Raspberry Pi camera, as an experimentation platform for different screens and lenses.

It takes the form of a central unit that holds the Pi and its support components, and front and rear modules for the screens or displays. Examples are given using the HQ and non-HQ modules, as well as with round or rectangular displays.

When designing a camera for 3D printing it’s a very difficult task, to replicate or exceed the industrial design of commercial cameras. Few succeed, and we’d include ourselves among that number. But this one comes close; it looks like a camera we’d like to use. We like it.

hackaday.com/2026/03/08/experi…

As pointed out by Tom’s Hardware, it’s been 26 years since the introduction of the gigahertz desktop CPU. AMD beat Intel to the punch by dropping the 1 GHz Athlon chip on March 6th of 2000, and partnered with Compaq and Gateway (remember them?) to deliver pre-built machines featuring the speedy silicon just a week later. The archived press release announcing the availability of the chip makes for some interesting reading: AMD compares the accomplishment with Chuck Yeager breaking the sound barrier, and mentions a retail price of $1,299 for the CPU when purchased in 1,000 unit quantities. In response Intel “launched” their 1 GHz Pentium III chip two days later for $990, but supply problems kept it out of customer’s hands for most of the year.

Speaking of breaking a barrier, Mobile World Congress took place this week in Barcelona, where TechCrunch reports there was considerable interest in developing a sub-$50 smartphone. The GSM Association’s Handset Affordability Coalition is working with major telecom carriers in Africa and as of yet unnamed hardware partners to develop the low-cost 4G device with the hopes of bringing an additional 20 million people online. While the goal is worthy enough, industry insiders have pointed out that the skyrocketing cost of memory will make it particularly challenging to meet the group’s aspirational price point.

Swapping out busted ports is a breeze on the new ThinkPads.
While we’re big fans of affordable hardware at Hackaday, we’re less enthusiastic when it comes at the cost of repairability. It seems that won’t be a problem with Lenovo’s new T14 and T16 ThinkPads however, as earlier this week iFixit announced they were giving the laptops a provisional repairability score of 10 out of 10. As impressive as this sounds, there’s a bit of a caveat here: Lenovo apparently achieved this milestone by working closely with iFixit to identify pain points that could be improved.

Of course, this doesn’t invalidate the work both companies put into these new machines, but you do have to wonder if it didn’t put a thumb on the scale. To address this there’s an Editor’s Note at the top of the post denying that any preferential treatment was given while scoring.

Although we’re thrilled to see a manufacturer other than Framework actually put effort into making their laptops cheaper and easier to repair, it’s a shame that things have gotten to the point that repairability is now considered a special feature. We’re not just talking about computers either; modern cars are notoriously difficult to work on, and electrics doubly so. Which is what makes the Aria EV so appealing.

Developed by students at the Eindhoven University of Technology in the Netherlands, the electric car is designed to be as repairable as possible. Before you get too excited, the idea isn’t to try and get the car to market. In fact, the team cautions that the vehicle isn’t technically street legal. Rather it serves as an technical demonstrator and test bed for concepts that one day the major players might include in their own vehicles, such as using multiple smaller battery packs that are easier to service than one gargantuan array of cells.

Finally, we’re not quite sure how long it’s been around, but we’ve been having a blast browsing through Famelack recently. It allows you to watch free Internet TV streams from all over the world right in your browser. The About page mentions several open source projects being used under the hood, such as Three.js, which powers the slick 3D globe used to select which country you want to tune into. Perhaps most notably however, it’s using the collection of streams curated by the iptv-org project, a valuable resource to keep in mind for future projects. If you end up watching anything particularly noteworthy, let us know in the comments.

See something interesting that you think would be a good fit for our weekly Links column? Drop us a line, we’d love to hear about it.

hackaday.com/2026/03/08/hackad…

With the RAM and storage crisis hitting personal computing very hard – along with new software increasingly suffering the effects of metastasizing ‘AI’ – more people than ever are pining for the ‘good old days’. For example, using that early 2000s desktop PC with Windows 98 SE might now seem to be a viable alternative in 2026, because it couldn’t possibly make things worse. Or could it? As a reality check, [SteelsOfLiquid] over on YouTube gave this setup a whirl.

The computer of choice is a very common Dell Dimension 2100, featuring a zippy 1.1 GHz Intel Celeron, 256 MB of DDR1, and a spacious 38 GB HDD. Graphics are provided by the iGPU in the Intel i810 chipset, all in a compact, 6.9 kg light package. As an early Windows XP PC, this gives Windows 98 SE probably a pretty solid shot at keeping up with the times. At least the early 2000s, natch.

Of course, there is a lot of period-correct software you can install, such as Adobe Photoshop 5, MS Office 97 (featuring everyone’s beloved Clippy), but a lot of modern software also runs, with the Retro Systems Revival blog documenting many that still run on Win98SE in some manner, including Audacity 2.0. This makes it totally suitable for basic productivity things.

YouTube in Netscape 4.5 on Windows 98. (Credit: Throaty Mumbo, YouTube)
Gaming on Win98 is naturally limited to games from around that early 2000s time period or before, but the gaming library even for just Win98 and MS-DOS is pretty massive, so as long as you’re fine not playing the latest and greatest games, this is also pretty easy.

Where things get dicey is of course with using the modern Internet, as you need a modern browser and support for the latest TLS encryption features to not have many websites throw a hissy fit. Using Frog Find and similar proxies that target retro computing help here, fortunately.

Previously we covered ways that you can use Discord even on Windows 95 and Windows NT 3.1, others have ported .NET applications to Windows 9x, got Win98 up and running on a 2020-era system, and you can totally use modern YouTube in even the Netscape 2.x browser using an NPAPI plugin.

Although there are many arguments to be made for using at least a Windows version with an NT kernel over the 9x one, it’s hard to deny that software Back Then was less complex, less resource-hungry and still got all the things done. Maybe it is worth another look, before the AI Crisis forces us all back on Windows XP systems like the one featured in this video.

youtube.com/embed/fD4M34xjbAw?…

hackaday.com/2026/03/08/how-us…

A long winter has a way of making a lot of us northerners a little bit squirrly. In [Build N Pulsejets]’s case, squirly enough to mount a home-made propane-powered pulse-jet to a kids’ kick scooter and take to the frozen lake for a rip.

Okay, well, it started as a kid’s scooter, but after trying it on the ice sans pulsejet, [Build N] decided that his cabin fever wasn’t quite bad enough to risk using it in stock configuration. Before mounting the 180 lb thrust (800 N) pulse jet he’d built in a previous video, a few modifications would be needed. Namely, a trio of scrap metal skis and a goodly amount of metalwork to mount them, and the pulse jet. Even on ice, with relatively little friction, the mass of maker and a full propane tank meant the acceleration wasn’t great, but he did get it over 44 mph (77 km/h) on the snowmobile drag strip. (Yeah, snowmobile drag racing’s a thing in the frozen north. Those of you sipping mai tais in the tropics are probably pretty jealous right now, huh?)

These pages have been no stranger to pulse jets, given that they’re probably the easiest engine to build at home. We’ve seen them mounted on everything from go karts, to Swedish snomobiles, and even tea kettles. Actually, we’ve seen two of those. No points for guessing what nation the tea kettle builds hail from.

youtube.com/embed/-lkSPgqQXek?…

hackaday.com/2026/03/08/pulse-…

There’s a well-known movie trope in which a hacker takes control of the traffic lights in a city, causing general mayhem or creating a clear getaway path. Unlike many Hollywood representations of hacking, this is actually possible in principle; many cities install Emergency Vehicle Preemption (EVP) systems in their traffic signals to turn them green when an emergency vehicle is approaching. To see what it would actually take to control one of these, [xssfox] reverse-engineered a Strobecom II EVP system.

Most EVP systems, particularly older ones, use a strobing infrared light to alert a traffic signal to an approaching emergency vehicle. To avoid misuse, vehicles often encode a vehicle ID in the infrared signal. There have been some claims that a Flipper Zero can trigger these systems, but none that were well-verified, and probably with good reason; anyone actually trying this against a live system is courting serious legal trouble. To see whether this was actually possible [xssfox] obtained real hardware and tried to reverse-engineer the infrared protocol.

There are two main manufacturers for optical EVP systems: GTT Opticom and Tomar Strobecom. [xssfox] managed to buy a Tomar power supply which handled the processing for signal transmission, and which worked with Opticom systems. Looking at the output of this revealed that it encoded data by skipping pulses, which should be simple enough for Flipper Zero to replicate.

To reverse-engineer the Strobecom protocol, [xssfox] managed to buy a Strobecom optical signal processor, which would normally detect an emergency signal. This worked by modulating the length of infrared pulses. After some brute-forcing, a transmitter using an Arduino Nano and an infrared LED managed to activate the preemption signal, and even to transmit a vehicle ID. It seems that Strobecom systems, at least, are fairly demanding in terms of the signals they accept; signals had to be precisely timed, and in at least some systems, a valid vehicle ID would be needed to change the light.

If you’d like to learn more, we’ve gone into the technology of North American traffic signals before.

youtube.com/embed/wodYdRVp_rY?…

hackaday.com/2026/03/08/spoofi…

The restored PET/CBM 3032. (Credit: Drygol, retrohax.net)
The Commodore CBM 3032 is a successor to the original Commodore PET 2001, yet due a conflicting trademark issue with Philips these first European PETs were called ‘CBM’ instead. Hence the labeling on the CBM 3032 that [Drygol] had in for a restoration, which would have been produced somewhere between 1979 and the cessation of its manufacturing a few years later. This former machine of the University of Szcezecin in Poland had languished in a basement until a local demoscene group came across it and wanted to use it, after a restoration.

Although at first glance from just the front it didn’t look too shabby, problems were apparent from just a walkaround, including rusty and buckled paneling, showing that the time spent in storage had not done it any favors. Internally there was decades worth of dust, along with a dodgy potentiometer, cold joints and some PCB-level bodges that may or may not have been there from the factory.

The main case was disassembled by drilling out the rivets to gain full access to every nook and cranny, allowing for a good cleaning and repainting prior to putting in fresh rivets. On the PCB side of things, a potentiometer and an LM340KC-12 linear regulator in a TO-3 package had to be replaced, after which the system managed to boot reliably once in every three attempts.

Fixing this took basically cleaning all contacts and IC sockets, as well as refurbishing the keyboard, with corrosion and the occasional broken trace causing a lot of grief. Ultimately the system was restored and ready to be put into demoscene service.

hackaday.com/2026/03/08/restor…

Around the thirteenth century CE, European society was in the midst between transitioning from Roman numerals to the Arabic numerals that we use today. Less remembered are the Cistercian numerals, which [BigCrimping] used for their most recent project in the form of a rather unique clock.

The Cistercian numeral system was developed by the Cistercian monastic order in the 13th century, forming a rather unique counterpoint to the Arabic numeral system. Although Arabic numerals are already significantly more compact than Roman numerals, Cistercian numerals up the ante by being capable of displaying any number between 1 and 9,999 with a single glyph.

Although for a simple 24-hour clock you don’t need to use more than a fraction of the possible glyphs, there is the complication of the Cistercian numerals not having a zero glyph, but that invites an even better take. For the version that [BigCrimping] made there are namely two glyphs that encode date and time, with the left glyph a counter for blocks of two hours and the right for seconds from 1 through 7200.

The clock is based around MAX6969 LED drivers and an ESP32 MCU on a custom PCB, with the design files including the 3D-printed enclosure available in the repository.

youtube.com/embed/COuOULkGAXk?…

hackaday.com/2026/03/08/buildi…

You’ve got to be ambitious to target a legend. If there’s one thing the folks at Hermeus Aerospace are, though, it’s ambitious: not only do they plan on their Quarterhorse unmanned arial vehicle (UAV) to outfly the SR-71 blackbird, they’re hoping to do it in record time. They took one big step closer to that goal in March 2026, when Quarterhorse 2.1 took off for the first time from Spaceport America.

The F-16-sized prototype is actually the second first flight Hermeus can brag of in the past year– version one first flew in May 2025. They’re iterating fast. Version 2.1 is hoped to prove a key part of the engine design for v2.2, which is the plane Humerus hopes to use to break the SR-71’s air-breathing speed record of Mach 3.3 from 1976. They’re hoping the next prototype can actually hit mach 5, which would be amazing if they pulled it off. Of course when exactly v2.2 will fly will depend largely on how this current model does in its test envelope.

This Quarterhorse hasn’t yet broken the sound barrier, but it certainly will. With the same F100 engine as the F-15 and F-16 fighters, it’s got the thrust, and one look tells you it has the aerodynamics. Of course an F100 can’t fly at Mach 5 — not on its own — but the F100 isn’t purely stock. It’s actually a component in Hermeus’ Chimera engine, which combines the F100 with a pre-chiller to actively cool the incoming supersonic air so the engine doesn’t melt at high speeds, and a ramjet stage that bypasses the engine entirely. That would make the Chimera a turboramjet engine; starting with an old and well-known turbine stage seems like a good move and is arguably a hack.

It would work like this: the engine takes off on turbine, the chiller kicks in when the aircraft goes supersonic, and the turbine is bypassed completely at high mach. This is how they hope to break the SR-71’s record: as well-designed as the J-57 engine was in that plane, it only pushed bleed air into the afterburner, rather than bypassing its turbine stage entirely, so was limited by the need to not melt said turbines. In some ways, the Chimera reminds us of a cheaper, simpler SABER engine. Of course as ambitious as breaking a 50 year old speed record might be, Hermeus’ goals are downright humble compared to the single-stage-to-orbit dreams the SABRE was meant to allow.

It remains to be seen just how fast Quarterhorse 2.1 will be able to go. Notably, at least as it was first unveiled, the aircraft doesn’t have any kind of shock cone on the inlet. It’s unlikely that the pre-chiller makes that unnecessary; it is more likely that either 2.1 is going to be restricted to low mach numbers where such things aren’t necessary, or it will be fitted later. Either way, we look forward to following the test program, at least as much as it is made public. Check out footage from the test flight in the video embedded below.

youtube.com/embed/apuGYFkTc5g?…

hackaday.com/2026/03/07/quarte…

You are at war. Trains are key to keeping your army supplied with fuel, ammunition, food, and medical supplies. But, inexplicably, your trains keep blowing up. Sabotage? Enemy attack? There’s no evidence of a bomb or overt enemy attack. This is the situation the German military found itself in during World War II. As you can see in the video below, the hidden bomb was the brainchild of a member of Britain’s SOE.

The idea was to put plastic explosive inside a fake plastic lump of coal. They hand-painted each one, and the color had to match the exact appearance of local coal. Paint and coal dust helped with that. The bomb had to weigh the correct amount as well.

The coal was safe until it got quite hot, so resistance fighters could easily carry the coal and surreptitiously drop the bomb anywhere coal is stored. Eventually, it will be put in a boiler, and at the right temperature, it will do its job. There’s some actual footage of a test in the second video below.

As the CIA notes, the idea actually dates back to the US Civil War. [Thomas Edgeworth Courtenay] built coal “torpedos” in the 1860s. (In those days, a torpedo could refer to any kind of bomb.) Probably the biggest impact was to tie up soldiers to guard coal stocks. However, in 1864, the USS Chenango’s boiler exploded in New York, and although the Union denied it, [Courtenay] was convinced it was one of his coal torpedoes that had done the trick. Later that year, Greyhound, the personal steamer of Major General Benjamin Butler, exploded right after taking in fresh coal. The CIA also mentions how coal bombs were also produced by the OSS, and even the Axis powers had their own version.

While we are no fans of war, we have to admit we are always fascinated with war technology. Even if that means microwave death rays. Certainly, hiding explosives in coal qualifies as a wartime hack.

youtube.com/embed/xdkkR4SCXL4?…

youtube.com/embed/BqMh-KBIHrE?…

hackaday.com/2026/03/07/hiding…

During the 1990s the Chornobyl Nuclear Power Plant – formerly the Chernobyl NPP – continued operating with its remaining three RBMK reactors, but of course the 1970s-era automation with its very limited SKALA computer required some serious modernization. What was interesting here is that instead of just replacing this entire Soviet-era mainframe with a brand-new 1990s one, the engineers responsible opted to build a new system – called DIIS – around it. This is detailed in a recent video by the [Chornobyl Family] on YouTube.

This SKALA industrial control system was previously detailed in a video, covering this 24-bit mainframe computer and its many limitations. It wasn’t quite a real-time control system, but it basically did what it was designed to do. Since at the time it was not clear for how long these three RBMKs would be kept running, they didn’t want to go overboard with investments either.

Ultimately Unit 2 only was active until 1991 due to a turbine fire, Unit 1 until 1996 and Unit 3 was shutdown for the last time in 2000, so this a sensible decision. During those years, an auxiliary information-measurement system (DIIS) was the big upgrade, which got bridged into SKALA via a Ukrainian-made SM-1210 minicomputer, with the latter connected to an 80386 PC which itself was connected to an ARCnet hub.

Best part of this DIIS upgrade was that it made it possible to run modeling algorithms for the reactor core based on measurements, without having to send data all the way over to the central control office in Moscow. Now reactor parameters could be visualized in real-time, and adjustments made via the same PRIZMA program’s magnetic tapes of the SKALA system as before.

Although the result was a bit of an odd mixture of 1970s Soviet mainframe design, 1980s-derived Ukrainian mainframe design and 1990s Intel computing power, it worked well enough to bring the ChNPP to the very doorstep of the 21st century with no issues worthy of note. Definitely a testament to the engineers who hacked this upgrade together and made it work so smoothly.

youtube.com/embed/86iGeVGZYQ4?…

hackaday.com/2026/03/07/how-th…

A commercial Delrama prism-based anamorphic lens for large cameras. (Source: Mathieu Stern, YouTube)
Although to the average person a camera lens is just that bit of glass you stick on the front of the camera to make stuff appear in focus, there’s a whole wide world out there of lens designs and modifications with enough variety to make your head spin. Some of these designs make a big impact, while others fade away again, sometimes at the whims of film makers and photographers. Prism-based anamorphic lenses are an oddity that recently [Mathieu Stern] got his hands on. (Video, embedded below.)

During the 1950s and 1960s there was a bit of a competition between anamorphic formats, which use special lenses that ‘squeeze’ a larger image so that widescreen movies could be recorded on standard 35 mm film. By using the same lens for recording and playback, the result was a mostly distortion-free image. Here the Technirama format by Technicolor who teamed up with Dutch company De Oude Delft (‘Old Delft’) to produce the prism-based Delrama lenses that fit on existing lenses for cameras and projectors.
The last gasp of the Delrama anamorphic lenses. (Credit: Mathieu Stern, YouTube)a
Despite having a clearly superior, distortion-free image than the cylindrical lenses of the competition, Technirama got pushed out of the commercial market, leaving De Oude Delft to try and interest the consumer market for Delrama with 8 and 16 mm adapters. These latter are the ones that [Mathieu] got his hands on and tried out with a DSLR camera.

Troublesome with these Delrama adapters is that their silver mirrors tend to degrade over time, and they also turned out to be rather fragile, which are both things that made consumers sour on them. Another challenge was the fixed four meter focus that’s great when you’re using it with a projector, but terrible for up-close shots. All of these issues resulted in Delrama fading from the market by the 1970s until all that remains are these remnants of a format that once was used to film some of the biggest Hollywood movies.

youtube.com/embed/jOMCzMHFhto?…

hackaday.com/2026/03/07/the-tr…