This week Jonathan Bennett and and Dan Lynch chat with Anthony Annunziata about Open Source AI and the AI Alliance. We get answers to our burning AI questions, and talk about the difficulty of defining what Open Source means for these large models.
Continue reading “FLOSS Weekly Episode 804: The AI Alliance — Asimov Was Right”
The Braun TS2 radio was a state-of-the art tube set in 1956. Today it still looks great, but unsurprisingly, the one that [Manuel Caldeier] has needed a little tender loving care. The table radio had a distinct style for its day and push-buttons. However, the dial glass and the speaker grill needed replacement. Even more interesting, the radio has a troublesome selenium rectifier, giving him the perfect chance to try out his new selenium rectifier solid-state replacement.
The radio is as good-looking inside as it is outside. You can tell that this isn’t his first restoration, as he has several tricks to test things at different stages of the project.
While the radio looked good, it smelled of smoke, which required a big effort to clean. The dial glass was intact enough for him to duplicate it in a graphic program and print it on a transparent adhesive sticker. With a deep breath, he removed the original markings from the glass so he could add the sticker to it. That didn’t work because the label needed cutouts. So now he is waiting for a piece of acrylic that will have the art UV printed on it.
We want to see the next part as we imagine the radio sounds as good as it looks when it is working. If you want to know more about the rectifier replacement, we covered that earlier. Even years later, Braun would have a clean aesthetic.
Orbital mechanics is a fun subject, as it involves a lot of seemingly empty space that’s nevertheless full of very real forces, all of which must be taken into account lest one’s spacecraft ends up performing a sudden lithobraking maneuver into a planet or other significant collection of matter in said mostly empty space. The primary concern here is that of gravitational pull, and the way it affects one’s trajectory and velocity. With a single planet providing said gravitational pull this is quite straightforward to determine, but add in another body (like the Moon) and things get trickier. Add another big planetary body (or a star like our Sun), and you suddenly got yourself the restricted three-body problem, which has vexed mathematicians and others for centuries.
The three-body problem concerns the initial positions and velocities of three point masses. As they orbit each other and one tries to calculate their trajectories using Newton’s laws of motion and law of universal gravitation (or their later equivalents), the finding is that of a chaotic system, without a closed-form solution. In the context of orbital mechanics involving the Earth, Moon and Sun this is rather annoying, but in 1772 Joseph-Louis Lagrange found a family of solutions in which the three masses form an equilateral triangle at each instant. Together with earlier work by Leonhard Euler led to the discovery of what today are known as Lagrangian (or Lagrange) points.
Having a few spots in an N-body configuration where you can be reasonably certain that your spacecraft won’t suddenly bugger off into weird directions that necessitate position corrections using wasteful thruster activations is definitely a plus. This is why especially space-based observatories such as the James Webb Space Telescope love to hang around in these spots.
Continue reading “Lagrange Points And Why You Want To Get Stuck At Them”
If you remember a time when TV news sets universally incorporated a room full of clattering wire service teleprinters to emphasize the seriousness of the news business, congratulations — you’re old. Now, most of us get our news piped directly into our phones, selected by algorithms perfectly tuned to rile us up on whatever the hot-button issue du jour happens to be. Welcome to the future.
If like us you long for a simpler way to get your news, [Andrew Schmelyun] has a partial solution with this dot-matrix news feeder. It’s part of his effort to detox a bit from the whole algorithm thing and make the news a little more concrete. He managed to chase down a very old Star Micronics printer with a serial interface, which he got on the cheap thanks to the previous owner not being sure if it worked. It did, at least after some cleaning, and thanks to a USB-to-serial and the efforts of Linux kernel hackers through the ages, was able to echo output to the printer from a Raspberry Pi Zero W.
From there, getting a daily news feed was as simple as writing some PHP code to mine the APIs of a few selected services. We’re perplexed and alarmed to report that Hackaday is not among the selected sources, but we’re sure this was just a small oversight that will be corrected in version 2. The program runs as a cron job so that a dead-tree version of the day’s top stories is ready for [Andrew]’s morning coffee.
We’ve seen similar news printers before; we particularly like this roll-feed paper version. But for a seriously retro feel, we’d love to see this done on a real teletype.
Remember Duracell’s PowerCheck? The idea was that a strip built into the battery would show if the battery was good or not. Sure, you could always get a meter or a dedicated battery tester — but PowerCheck put the tester right in the battery. [Technology Connections] has an interesting video on how these worked and why you don’t see them today. You can see it below.
Duracell didn’t invent the technology. The patent belonged to Kodak, and there were some patent issues, too, but the ones on the Duracell batteries used the Kodak system. In practice, you pushed two dots on the battery, and you could see a color strip that showed how much capacity the battery had left. It did this by measuring the voltage and assuming that the cell’s voltage would track its health. It also assumed — as is clearly printed on the battery — that you were testing at 70 degrees F.
The temperature was important because the secret to the PowerCheck is a liquid crystal that turns color as it gets hot. When you press the dots, the label connects a little resistor, causing the crystals to get warm. The video shows the label taken apart so you can see what’s inside of it. The resistor isn’t linear so that’s how it changes only part of the bar to change color when the battery is weak but not dead.
It is a genius design that is simple enough to print on a label for an extremely low cost and has virtually no components. PowerCheck vanished from batteries almost as suddenly as it appeared. Some of it was due to patent disputes. But the video purports that normal people don’t really test batteries.
Watch out for old batteries in gear. Of course, if you want to really test batteries, you are going to need more equipment.
Most of us have made paper airplanes at one time or another, but rather than stopping at folded paper, [VirgileC] graduated to 3D printing them out of PLA. Then the obvious question is: can you cast one in cement? The answer is yes, you can, but note that the question was not: can a cement plane fly? The answer to that is no, it can’t.
Of course, you could use this to model things other than non-flying airplanes. The key is using alginate, a natural polymer derived from brown seaweed, to form the mold. The first step was to suspend the PLA model in a flowerpot with the holes blocked. Next, the flowerpot gets filled with alginate.
After a bit, you can remove the PLA from the molding material by cutting it and then reinserting it into the flower pot. However, you don’t want it to dry out completely as it tends to deform. With some vibration, you can fill the entire cavity with cement.
The next day, it was possible to destroy the alginate mold and recover the cement object inside. However, the cement will still be somewhat wet, so you’ll want to let the part dry further.
Usually, we see people print the mold directly using flexible filament. If you don’t like airplanes, maybe that’s a sign.
If you’re looking for an intriguing aerial project, [DilshoD] has you covered with his unique twist on modular airships. The project, which you can explore in detail here, revolves around a modular airship composed of individual spherical bodies filled with helium or hydrogen—or even a vacuum—arranged in a 3x3x6 grid. The result? A potentially more efficient airship design that could pave the way for lighter-than-air exploration and transport.
The innovative setup features flexible connecting tubes linking each sphere to a central gondola, ensuring stable expansion without compromising the airship’s integrity. What’s particularly interesting is [DilshoD]’s use of hybrid spheres: a vacuum shell surrounded by a gas-filled shell. This dual-shell approach adds buoyancy while reducing overall weight, possibly making the craft more maneuverable than traditional airships. By leveraging materials like latex used in radiosonde balloons, this design also promises accessibility for makers, hackers, and tinkerers.
Though this concept was originally submitted as a patent in Uzbekistan, it was unfortunately rejected. Nevertheless, [DilshoD] is keen to see the design find new life in the hands of Hackaday readers. Imagine the possibilities with a modular airship that can be tailored for specific applications. Interested in airships or modular designs? Check out some past Hackaday articles on DIY airships like this one, and dive into [DilshoD]’s full project here to see how you might bring this concept to the skies.
