TechyMagThings

A lot has been made about the increase of automated traffic on the Internet, with the past years LLM-related crawlers having quite literally swarmed the picture here. Not only does this drive up traffic, it also increases load on web servers, whose owners find themselves faced with increased hosting costs. This recently led to The-Numbers.com going offline for a while as automated traffic was quite literally destroying their bottom line.

This saga is covered by [Stephen Follows], who had a chance to talk with the founder and CEO of the site, [Bruce Nash], after the site went basically offline for a few months. Since the website both licenses data for commercial purposes as well as offering the free access on its website, there were accusations of this being a ‘rug pull’.

The site was started in 1997, as a static HTML site on Geocities where [Bruce] provided box office analyses for investment purposes. Since that beginning traffic was generally polite, with human visitors and usually well-behaved search engine crawlers. Then around 2024 the first wave of scraper bots arrived, followed by a larger wave around December of 2025.

Despite implementing a few mitigations, such as LLM-targeted text, the increased traffic and the resulting load on a site architecture that was never designed for this ultimately led to a collapse. One of the major sources of traffic turned out to be from so-called ‘prediction markets’, like Polymarket, whose bots absolutely hammered the site.

Fortunately for [Bruce] and his team they do not rely on the free website for income, but they have had to massively rework the site’s architecture to bring back a semblance of the original features. As noted in the article, the amount of crawling traffic by these LLMs and ‘agentic AI’ tools is logarithmically more than that for search engines, which makes this a major challenge.

Issues like these is why services such as Cloudflare are offering blocking features for such automated traffic. After all, unless such traffic is of use to you, you may as well treat it like a DDoS attack and cut it off at the root.

Thanks to [Ben] for the tip.

The sport of fencing requires keeping score, just like so many other similar pastimes. When their club’s existing scoring rig broke, [jc0025] stepped up to build a scoring box of their own, using the typical tools of the maker trade.

The brains of the operation is an Arduino Nano, running the venerable ATmega328P. It’s set up to drive a pair of 8×8 WS2812B addressable LED panels. It’s also hooked up to a pair of fencing socket blocks, which hook up to the lamé (jacket), weapon, and guard of each player for electronically scoring hits. The Arduino is thus programmed to respond to various conditions, lighting the LEDs in turn. For example, the tip of one player’s weapon hitting the other player’s lamé will fire a colored light, allowing the hit to be scored. Meanwhile, a tip hitting the floor will fire a white light, indicating off-target. There’s also a buzzer for sonic indication, as well. Everything is wrapped up in a tidy 3D-printed housing, while power is courtesy of a USB-C charger hooked up to the unit.

If you’re looking to replace some old broken fencing scoring gear on the cheap, this project is probably a good place to start. We’ve featured some other great scoreboard projects over the years, too. Meanwhile, if you’re whipping up your custom own gear for your local sporting club, we might like to hear about it on the tipsline.

Silicon-based CMOS camera sensors are cheap and plentiful, but they’re rarely used to their full potential: they can detect a greater range of wavelengths in the infrared spectrum than they can in the visible spectrum, but in most cameras this is blocked by an IR-cut filter. [Project 326]’s infrared camera system reverses this: it records infrared images in color while blocking out visible light.

The system uses three USB webcams, each with its IR cut filter removed and replaced with a different dichroic IR band-pass filter. One filter is centered at 750 nm, one at 850 nm, and one at 940 nm. There is no band overlap; in testing, each camera only detected an infrared flashlight tuned to its own filter wavelength. The original cameras didn’t hold the sensors in a consistent position, so [Project 326] designed new housings. Using three lenses, each with distinct aberrations, introduced some difficulties in alignment. [Project 326] originally intended to use a pair of beam-splitting prisms with only one lens, but this proved too difficult to align using 3D-printed frames.

A Raspberry Pi records a separate monochromatic stream from each camera, which can then be processed into a composite color video. The first frames need to be manually aligned, but afterwards a script can apply the alignment to the rest of the video. Finally, the channels are mapped to colors, with the precise mapping being freely changeable. There were some few unexpected issues: each camera has its own, not terribly precise, local oscillator, and they drifted apart by about one or two frames per minute. Parallax error, on the other hand, was less severe than might be expected: at close range it’s noticeable, but by a distance of 35 meters, it represents less than one pixel of distortion.

The resulting images look great, and it’s easy to forget that they’re being captured without the use of any visible light. We’ve seen a similar technique (though extending into the visible range) used to recreate the surreal effect of Aerochrome film.

Would you buy a broken device, fix it, then return the repaired item? [Team Repair] hopes you will.

Many people would rather repair what they have than have to get a new device and deal with all the annoyances of shopping and getting used to a new item. The problem is, most people don’t have any experience fixing their broken electronics and are intimidated by the process. [Team Repair] has found that helping people through the process the first time is a big confidence boost, especially when they have a test piece to work with.

[Team Repair] started with workshops and now has a “Fixers Club” subscription box for 8-14 year olds where they will send a broken device every three months to fix. After the device is repaired, it gets sent back into Team Repair, or iFixit in the US, to be broken again and sent to the next fixer. For grownups, there is a similar smartphone repair kit, but it’s currently waitlisted. If you’re in the UK, they even have a classroom program if you want to recruit larger groups into your fixer corps.

Curious about other ways to get into repairing your own things? We don’t recommend starting with smart rings, but learning analog repairs like zippers or turning a laptop into a desktop could be good places to start?

Key to understanding something like epilepsy is to be able to record highly transient events in biological tissues. Generally this is done using light sheet microscopy, which provides effectively a 2D ‘slice’ of the tissue in question, but to observe a brief event in a larger biological system you need to be able to rapidly change the layer and focus between the virtual layers. This is what [Bingxi Liu] et al. al did using adaptive optics with an electrically tunable lens (ETL) in order to capture seizures in the brain of zebrafishes.

Their system can capture a volume of 499 × 499 × 150 μm³ at 4 volumes per second, which is large enough to fit optically transparent zebrafish larva into. The optical setup is shown in the above image, with the design based on the OpenSPIM platform for selective plane illumination microscopy.

Here the 488 nm laser provides the illumination (excitation) of the layer, while the 543 nm laser is for calibration purposes. The ETL is thus in the imaging path that allows for capturing by a digital camera, while a beam splitter directs part of the captured data to a Shack-Hartmann wave front sensor (SHWFS), which is part of the adaptive optics system.

After a seizure was induced in the zebrafish larva using the drug pentylenetetrazol the results were recorded using this system. It showed the seizure’s origin in the posterior brain, with subsequent propagation to the anterior before subsiding gradually over tens of seconds.

This system should be quite useful even outside of seizure research, as there are a lot of 3D systems in biology where having a relatively high-speed microscopic capture can be very revealing.

The standard computer mouse is a perfectly useful peripheral if your hands work. If you’ve got some trouble in that area, you might appreciate an alternative input solution. To that end, [Varun Adinath Patil] created a neat hands-free solution for moving a cursor around a screen.

The build is based on the Neuro PlayGround Lite, a board built for physiological signal acquisition in the Feather form factor. It’s hooked up to an IMU sensor—both a MPU6050 or BMI270 work—which tracks head movements to allow the cursor to be panned around the screen. Other biological signals are then used to activate other standard mouse functions. Clenching the jaw fires off a left click, while a triple blink fires a right click. Clicking and dragging is achieved by a double-blink. The jaw muscles are sensed via EMG signals picked up with gel electrodes on the skin, while the blinks are detected via EOG signals via the same contact points.

Commercial solutions in this realm exist, but it’s great to see how such a device can be built from the ground up. We’ve looked at other neat applications of head-tracking before, too. If you’re working on your own innovative accessibility tools, don’t hesitate to let us know via the tipsline.

Community-led open source project hosting site Codeberg has formally announced that projects whose code is largely or fully machine-generated through LLMs and other ‘AI’ tools will no longer be welcome. This follows on the heels of a similar ban on cryptocurrency-related projects.

The community vote was on two issues, the first being the notion that scraping of project code for the use in LLMs should be forbidden, which was a motion that easily passed. The second motion was on disallowing projects whose code was substantially generated by LLMs like Claude, OpenAI Codex, and similar. This motion passed with 358 in favor versus 144 against.

In the earlier linked blog post the reasoning behind especially this second issue is expanded upon, covering not only ‘license whitewashing’, but also the direct and indirect hardware costs, with the expanding ‘AI’ datacenter hyperscaling having massively increased hardware costs for Codeberg over the past years, as the costs have been largely externalized.

Also covered is also the aspect of these LLM-based tools destroying the OSS community, which is something that is backed up by recent studies. Even if we ignore that such LLM-tools are destroying the cognitive abilities of its users, there’s an argument to be made that if LLM-scraping is disallowed, then it’s consistent to also not allow LLM-generated code.

In the Terms of Use you can see these changes, both for LLMs and for cryptocurrency projects.

Thanks to [mk-fg] for the tip.