More Information

The Federal Communications Commission has released a notice today designating any consumer routers manufactured outside the US as a security risk. The rule states that new foreign-made product models for network routers will land on the Covered List, a set of communications equipment seen as having an unacceptable risk to national security. Previously purchased routers can still be used and retailers can still sell models that were approved by the prior FCC policies. In an exception to the usual rule, routers included on the Covered List can continue to receive updates at least through March 1, 2027, although the date could potentially be extended.

The move stems from a goal in the White House’s 2025 national security strategy that reads: “the United States must never be dependent on any outside power for core components—from raw materials to parts to finished products—necessary to the nation’s defense or economy.” The notice from the FCC states that companies can apply for conditional approval for new products from the Department of War or the Department of Homeland Security. However, that requires the businesses to provide a plan for shifting at least some of their manufacturing to the US in order to receive that conditional approval.

Few, if any, brands known for consumer-grade routers currently build products stateside. It seems likely this sweeping provision could face legal challenges from and cause confusion for the many companies that have production facilities overseas. In addition to Chinese tech giants like TP-Link, US companies will also be affected. NetGear, Eero and Google Nest are all headquartered domestically but have manufacturing in Asia. At least some of that manufacturing activity happens in regions like Taiwan that have historically been on good terms with the US. Until the sector sorts out this new restriction, don’t expect to see any new router models on store shelves.

We don’t usually speculate on the true identity of the hackers behind these projects, but when [TN666]’s accoustic drone-detector crossed our desk with the name “Batear”, we couldn’t help but wonder– is that you, Bruce? On the other hand, with a BOM consisting entirely of one ESP32-S3 and an ICS-43434 I2S microphone, this isn’t exactly going to require the Wayne fortune to pull off. Indeed, [TN666] estimates a project cost of only 15 USD, which really democratizes drone detection.

The key is what you might call ‘retrovation’– innovation by looking backwards. Most drone detection schema are looking to the ways we search for larger aircraft, and use RADAR. Before RADAR there were acoustic detectors, like the famous Japanese “war tubas” that went viral many years ago. RADAR modules aren’t cheap, but MEMS microphones are– and drones, especially quad-copters, aren’t exactly quiet. [TN666] thus made the choice to use acoustic detection in order to democratize drone detection.

Of course that’s not much good if the ESP32 is phoning home to some Azure or AWS server to get the acoustic data processed by some giant machine learning model.  That would be the easy thing to do with an ESP32, but if you’re under drone attack or surveillance it’s not likely you want to rely on the cloud. There are always privacy concerns with using other people’s hardware, too. [TN666] again reached backwards to a more traditional algorithmic approach– specifically Goertzel filters to detect the acoustic frequencies used by drones. For analyzing specific frequency buckets, the Goertzel algorithm is as light as they come– which means everything can run local on the ESP32. They call that “edge computing” these days, but we just call it common sense.

The downside is that, since we’re just listening at specific frequencies, environmental noise can be an issue. Calibration for a given environment is suggested, as is a foam sock on the microphone to avoid false positives due to wind noise. It occurs to us the sort physical amplifier used in those ‘war tubas’ would both shelter the microphone from wind, as well as increase range and directionality.

[TN] does intend to explore machine learning models for this hardware as well; he seems to think that an ESP32-NN or small TensorFlow Lite model might outdo the Goertzel algorithm. He might be onto something, but we’re cheering for Goertzel on that one, simply on the basis that it’s a more elegant solution, one we’ve dived into before. It even works on the ATtiny85, which isn’t something you can say about even the lightest TensorFlow model.

Thanks to [TN] for the tip. Playboy billionaire or not, you can send your projects into the tips line to see them some bat-time on this bat-channel.

The Z80 has been gone a couple of years now, but it’s very much not forgotten. Still, the day when new-old-stock and salvaged DIP-40 packaged Z80s will be hard to come by is slowly approaching, and [eaw] is going to be ready with the picoZ80 project.

You can probably guess where this is going: an RP2350B on a DIP-40 sized PCB can easily sit on the bus and emulate a Z80. It can do so with only one core, without breaking a sweat. That left [eaw] a second core to play with, allowing the picoZ80 to act as a heck of an accelerator, memory expander, USB host, disk emulator– you name it. He even tossed in an ESP32 co-processor to act as a WiFi, Bluetooth, and SD-card controller to use as a virtual, wirelessly accessible disk drive.

The onboard ram that comes with an RP2350B would be generous by 1980s standards, but [eaw] bumped that up with an 8 MB SPRAM chip–accessed in 64 pages of 64 kB each, naturally. If more RAM than a very pricey hard drive wasn’t luxury enough, there’s also 16 MB of flash memory available. That’s configured to store ROM images that are transferred to the RAM at boot– the virtual Z80 isn’t grabbing from the flash at runtime in [eaw]’s architecture, because apparently there are limits to how much he wants to boost his retro machines.

There are already drivers to use in certain Z80 systems. You can of course configure it as a bare Z80 with no machine-specific emulation, or set up the picoZ80 with the “persona” of a classic Z80 machine. So far [eaw] has tried this on an RC2014 homebrew computer, as well as Sharp MZ-80A– which we’ve seen here before, in miniature–and Sharp MZ-700. The Sharp drivers are still works in progress, after which the Amstrad PCW8256/Tatung TC01 is apparently next. We’ve seen Amstrad PCWs here a time or two as well, come to think of it.

If somehow you missed it, the venerable Z80 only hit EOL in 2024, so supplies won’t be drying up any time soon. This hack is really more about the quality-of-life addons this allows. Come back in a decade, and we’ll see if the RP2350 lasts longer than the stack of NOS Z80s.

“Tavajoh! Tavajoh! Tavajoh!” a man’s voice announces, before going on to narrate a string of numbers in no apparent order, slowly and rhythmically. After nearly two hours, the calls of “Attention!” in Persian stop, only to resume again hours later.

The broadcast has been playing twice a day on a shortwave frequency since the start of the US-Israel attack on Iran on February 28.

According to Priyom, an organization which tracks and analyses global military and intelligence use of shortwave radio, using established radio-location techniques, the broadcast was first heard as the US bombing of Iran began. It has since played on the 7910 kHz shortwave frequency like clockwork—at 02.00 UTC and again at 18.00 UTC.

Over the weekend, Priyom said it had identified the likely origin of the broadcast. Using multilateration and triangulation techniques, the group traced the signal to a shortwave transmission facility inside a US military base in Böblingen, southwest of Stuttgart, Germany.

The site lies within a restricted training area between Panzer Kaserne and Patch Barracks, with technical operations possibly linked to the US army’s 52nd Strategic Signal Battalion, headquartered nearby.

That identification narrows the field, but it does not reveal who is behind the transmissions or who they are meant for.

The two-hour-long transmission is divided into five to six segments, each lasting up to 20 minutes. Each opens with “Tavajoh!” before shifting into a string of numbers in Persian, sometimes punctuated with an English word or two. Five days into the broadcast, radio jammers were heard attempting to block the frequency. The following day, the transmission shifted to a different frequency—7842 kHz.

Radio communication experts believe the broadcast is likely part of a Cold War–era system known as number stations.

The Return of the Numbers

Number stations are shortwave radio broadcasts that play strings of numbers or codes that sound random—like the one now heard in Iran. “It is an encrypted radio message used by foreign intelligence services, often as part of a complex operation by intelligence agencies and militaries,” says Maris Goldmanis, a Latvian historian and avid numbers stations researcher.

Number stations are most commonly associated with espionage. “For intelligence agencies, it is important to communicate with their spies to gather intelligence,” says John Sipher, a former US intelligence officer who served 28 years in the CIA’s National Clandestine Service. “This is not always possible in person due to political constraints or conflict. This is where number stations come in.”

While the use of number stations can be traced back to the First World War, they gained prominence during the US-Soviet Cold War. As espionage grew more sophisticated, governments used automated voice transmissions of coded numbers to communicate with agents, Goldmanis says. Citing declassified KGB and CIA documents, he adds that number stations were widely used during this period, often as Morse code transmissions and, in many cases, as two-way communications, with agents reporting back using their own shortwave transmitters.

“Nowadays, you have various satellite and encrypted communications technologies,” Sipher says. “But during the Cold War and even before that, governments had to find ways to do this without being noticed, and broadcasting coded messages was one way to communicate with your assets discreetly.”

The apparent randomness of the numbers means they can be understood only with a codebook, Sipher adds. “Nobody can make heads or tails of it or understand what it says unless you have the codebook that can give you hints to decrypt the code,” he says, noting that such systems must be set up and coordinated in advance.

A Signal Without a Sender

While the likely origin of the signal may now be clearer, its purpose and intended recipient remain unknown.

Because the broadcasts are encrypted and designed to be covert, those details may remain unclear for years, Goldmanis says. The structured nature of the transmission—its fixed schedule and consistent use of frequencies—further suggests it is part of a planned operation.

As the Trump administration phases out the use of animal experimentation across the federal government, a biotech startup has a bold idea for an alternative to animal testing: nonsentient “organ sacks.”

Bay Area-based R3 Bio has been quietly pitching the idea to investors and in industry publications as a way to replace lab animals without the ethical issues that come with living organisms. That’s because these structures would contain all of the typical organs—except a brain, rendering them unable to think or feel pain. The company’s long-term goal, cofounder Alice Gilman says, is to make human versions that could be used as a source of tissues and organs for people who need them.

For Immortal Dragons, a Singapore-based longevity fund that’s invested in R3, the idea of replacement is a core strategy for human longevity. “We think replacement is probably better than repair when it comes to treating diseases or regulating the aging process in the human body,” says CEO Boyang Wang. “If we can create a nonsentient, headless bodyoid for a human being, that will be a great source of organs.”

For now, R3 is aiming to make monkey organ sacks. “The benefit of using models that are more ethical and are exclusively organ systems would be that testing can be meaningfully more scalable,” Gilman says. (R3’s name comes from the philosophy in animal research known as the three R’s—replacement, reduction, and refinement—developed by British scientists William Russell and Rex Burch in 1959 to promote humane experimentation.)

New drugs are often tested in monkeys before they’re given to human participants in clinical trials. For instance, monkeys were critical during the Covid-19 pandemic for testing vaccines and therapeutics. But they’re also an expensive resource, and their numbers are dwindling in the US after China banned the export of nonhuman primates in 2020.

Animal rights activists have long pushed to end research on monkeys, and one of the seven federally funded primate research facilities across the country has signaled it would consider shutting down and transitioning into a sanctuary amid growing pressure. The US Centers for Disease Control and Prevention is also winding down monkey research, part of a bigger trend across the government to reduce reliance on animal testing.

As a result, Gilman says, there aren’t enough research monkeys left in the US to allow for necessary research if another pandemic threat emerges. Enter organ sacks.

Organ sacks would in theory offer advantages over existing organs-on-chips or tissue models, which lack the full complexity of whole organs, including blood vessels.

Gilman says it’s already possible to create mouse organ sacks that lack a brain, though she and cofounder John Schloendorn deny that R3 has made them. (For the record, Gilman doesn’t like the term “brainless” to describe the organ sacks. “It’s not missing anything, because we design it to only have the things we want,” she says.) Gilman and Schloendorn would not say how exactly they plan to create the monkey and human organ sacks, but said they are exploring a combination of stem-cell technology and gene editing.

It’s plausible that organ sacks could be grown from induced pluripotent stem cells, says Paul Knoepfler, a stem cell biologist at the University of California, Davis. These stem cells come from adult skin cells and are reprogrammed to an embryonic-like state. They have the potential to form into any cell or tissue in the body and have been used to create embryo-like structures that resemble the real thing. By editing these stem cells, scientists could disable genes needed for brain development. The resulting embryo could then be incubated until it grows into organized organ structures.

The first full moon of spring 2026 is on its way, and with it, an early Easter. April’s Pink Moon is scheduled for the first day of April, and while it’s not a lunar eclipse like the full moon in March, it should still light up the sky.

The best time to view the full moon is the evening of April 1. Per The Old Farmer’s Almanac, peak illumination occurs at 10:12 p.m. ET. That’s well after dark for much of the US, and since the moon is set to rise at around 8 p.m. local time in all time zones, most of the US should get a chance to see it at peak illumination. 

The only ones left out are those on the West Coast, where the moon won’t rise until around an hour after peak illumination. It doesn’t matter much in the grand scheme since the moon will still be completely full. If you miss the full moon due to inclement weather, the moon will still be mostly full in the two days leading up to and after April 1, so you’ll have plenty of chances to see at least a mostly full moon.

Happy Easter!

The Pink Moon doesn’t have any special characteristics like January’s supermoon or last June’s micromoon. (More of those will come later in 2026.) There is still some cultural significance for this year’s Pink Moon. In Christianity, the first full moon that takes place after the spring equinox determines the calendar dates for Easter. That particular full moon is known as the Paschal Moon.

Easter is always observed on the first Sunday after the first full moon of spring. That means the date for Easter this year is April 5. 

Since the holiday doesn’t have a fixed date like Christmas, it’s commonly referred to as the “movable feast” and can take place anywhere between March 22 and April 25. The dates are based on the fact that Christianity recognizes the spring equinox as March 21 every year, even though the astronomical date varies slightly, making March 22 the earliest day Easter can occur. The moon cycle is 29.5 days, and when you do the math, the latest you can go is April 25. The next time Earth is scheduled to have an Easter on April 25 is 2038.

The launch of Crimson Desert wasn’t smooth for everyone, especially Intel Arc GPU users. While the game was a AAA release that showed refreshing levels of polish, it didn’t support Intel Arc graphics at all.

If you looked for more details and stumbled upon the FAQ, Pearl Abyss simply told you to seek a refund.

And as expected, this didn’t go down so well with the gaming community. Now, the studio is changing course and has confirmed that Intel Arc support is officially in the works, marking a major shift in stance.

The backlash clearly worked

The controversy was picked up quickly after launch, with players calling out the lack of support. This was even more surprising since Intel was working with the studio during development, and reportedly even offered drivers and engineering help for Arc GPU support.

But following the immediate backlash, the developers issued an apology.

What went wrong in the first place?

At launch, Crimson Desert simply would not run on Intel Arc GPUs, throwing up an “unsupported hardware” error. It wasn’t a minor bug or performance issue; what was surprising was the complete lack of compatibility. This affected both discrete Arc cards and Intel’s integrated graphics.

The decision raised eyebrows almost immediately, considering how widely Intel iGPUs are used across laptops and PCs. The good news is that Pearl Abyss has now committed to fixing it. While it has confirmed active development for compatibility updates, performance optimization, and a “smooth and stable gameplay experience” on Arc GPUs, there’s still no clear timeline for when this support will actually roll out.