HN user

ckocagil

869 karma
Posts3
Comments172
View on HN
Hy3 13 days ago

Isn't Q8 way overkill these days? I see many graphs showing Q4 or Q5 having less than %1 deviation. Nvidia's NVFP4 Qwen quantization should be even better due to its better training methods.

Late by a decade or more (JSR310 was released in 2014), but still a good development. I've tried convincing colleagues to use js-joda in the past but they thought they were keeping it simple by sticking to moment.js. They weren't.

Finally there's a new stepping (i.e revision) for the RP2350 die that fixes the E9 bug that has plagued the chip so far, causing excessive input currents on GPIO pins which made internal pull-up resistors useless and made high impedance inputs impossible.

GPIO pins are now all 5V tolerant too!

I made my DEs somewhat pretty but never got too deep into ricing. Then I switched to i3-gaps. But I still spent a lot of time configuring and comparing terminal emulators, shell prompts, completion, vim plugins, iommu/vfio...

At one point I got sick of it all and decided to switch back to Windows/MacOS, not configure anything too heavily, install tools when I need them, and use all tools as vanilla as possible. Besides not having to maintain all this stuff I also realized I could then ssh into any machine and not miss my customized tools.

All the "AI software engineer" hype is about one thing: suppressing developer wages. Steve Jobs used to do this covertly, to some extent, with his "no cold call" anti-poaching agreements.

It's a multi-prong attack. One angle is the interest rates and mass layoffs. Another angle is AI. There are more ridiculous ones, such as this "study" that claims 10% of devs do literally no work: https://x.com/yegordb/status/1859290734257635439 - I expect many such studies to be conducted and generously funded by the industry in the coming years.

The big irony is how all this talk will lead to a decrease in the software engineer supply. Why would anyone choose a career that's about to die?

For any consumer use to play back audio I can't image a scenario where TL071 wouldn't be enough. And you'd rarely even need an op-amp when integrated solutions are available.

The problem with the common CO2 sensor modules is they don't have DC accuracy. Meaning they rely on the device being present in place where it regularly (e.g at least once a week) gets exposed to fresh air, which the module sets as its baseline. This works because fresh air has roughly the same CO2 concentration everywhere.

Hopefully this method doesn't have the same restriction.

Performance. JS would need to run 60 times a second without hiccups. But JS is single-threaded, therefore any time something clogs your main thread you'll skip a frame. Plus, as incredible as modern JS engines are they're still not as efficient as the native code, which leads to extra CPU usage and that spells trouble for battery life.

Nice project. A low DA capacitor Cf in parallel with the gain setting resistor Rf can average out more noise if needed. It can even deal with the mains noise if that remains an issue. Another potential improvement: addition of an optical chopper wheel to deal with the entire system's 1/f noise. The downside is this would require higher sampling rate which would then get demodulated + filtered externally. Would also limit Cf to higher frequencies.

I had posted this a few years back but it didn't get noticed back then. Posting again because this video is exemplary in its clarity and it's a topic the HN crowd might like. This is also one of the best science education videos I've seen and I keep coming back to it every few years.