gsquaredxc

Even better: All of the under 18 users are probably on the free plan. It’s fairly likely that this demographic is disproportionately expensive.

I somewhat agree and somewhat disagree. If your students haven’t completely disabled GoGuardian, they are only working within the system. At least when I was a student working on this problem, I spent some time writing some software that would selectively disable and enable GoGuardian features, but my peers were not. I’d imagine similar exists today, a “good enough” approach beats out a technically sound one. “Unblocked” games or websites rely on a brute-force “what are they going to do about it” ideology that doesn’t work when the person you are disagreeing with isn’t a 50 year old man that lacks motivation. If it can’t be explained in a minute, it’s probably unlikely any more than one in one hundred actually do it. As hopeful as I’d like to be about this, I’d doubt the next generation gets adept at anything more than lying about their age online.

Unrelated to security, modern chips are designed with JavaScript in mind. ARM’s FJCVTZS is the most famous example. Broadly speaking, a custom chip for JavaScript would likely not be to improve security, only performance. ARM did try out something similar (in the broadest sense of the word) for Jazelle and later ThumbEE, which haven’t been around for decade(s) now.

Delay is frustrating, but not a violation. Just because you reverse engineered GPLv2 code doesn’t make that in violation of GPLv2, many people have done that for reasons that rhyme with yours over the decades. GPLv2 also doesn’t have any clause about security patches or if you have a requirement for a certain timeline. GPLv2 isn’t a “you agree to not be an asshole” license.

Tarballs certainly aren’t in violation of the GPL, via gnu.org own source distribution methodology. It would also be a very interesting legal claim that git metadata is required, and likely something already hashed out when the government decided that metadata isn’t the same as the underlying data.

They still make printed textbooks. Google’s lawyers I’m sure are well funded enough to make this case.

That’s an ambitious interpretation. The preferred form for making modifications to the Linux kernel also involves having a build cache, but that’s clearly not required to remain GPL compliant.

Starlink is going to be a cellular company, except instead of maintaining cheap metal frames, they have to physically launch the antennas to space every 5 years.

You can reason it out (in a way that might make Fermi proud) pretty easily: a very large AC unit can be powered by a very small (relative) solar panel, where the solar panel is maybe 20% efficient. Thus, solar irradiation is substantially higher than AC power usage.

Wouldn’t someone that has to pay for an AC’s power bill be more aware of the cost-benefit analysis of energy efficiency over someone that doesn’t?

> Here I should note that Python is not the best choice for CPU-bound software. I want to take the opportunity to learn Zig.

For optimizing CPU-bound operations in Python, there’s some low hanging fruit with numba. I would recommend this as a 5-minute solution to you limiting your algorithm because regular Python is too slow. I regularly tell people that if their Python program is slow enough to take several minutes, you could probably learn numba before it finishes.

There's some good answers here but something important is missed. Militaries spend a lot of time having soldiers shoot rounds down the range to ensure everyone is up to snuff. The engineering corps is similar but better. You can fund target practice for your army engineers by having them go build public infrastructure. It's a win for the military (practical experience), win for locals (public infrastructure), and a win for the taxpayer (two for one deal).

Even weirder: the Kickstarter campaign says it’s 4 MSPS per channel. 100kHz bandwidth with 4 MSPS per channel just doesn’t make sense. However, they have “verified” 400kHz on their Kickstarter. Not sure who verified it, but it’s verified.

The Kickstarter does have product photos of the back in a gif but be forewarned: they don’t include any chip designations.

The specs on the physical device are quite weak too. 100kHz scope, no mention of # of bits. Power supply isn’t programmable, just +-5V at 200mA. 5V PWM and “sine and triangle” is the function generator.

It’s definitely misleading to say this replaces $1000+ worth of equipment because I’m not sure you can buy equipment with this poor of specs.

A 100kHz scope, 2W fixed dual rail power supply, and a function generator that is anything but arbitrary means at the very least projects are constrained heavily. I think part of the joy of learning is going off of the beaten path a bit, and I think this is limiting enough to prevent that.

I’ve been on the lookout for cheaper stuff to recommend to friends looking to work their way down from classical software engineering towards embedded and I think this is relatively disappointing in that regard. Second hand seems like the way to go economically but I haven’t yet looked into proper sourcing for equipment.

I would say that there’s quite substantial value perhaps not to the general public but to the general HN user from their reports. I was very excited to see them have a public scan of an LFP battery because the previous scans they did of batteries were very informative on which brands to trust and the level of inconsistency that can exist. Their analysis of BYD’s battery design is very useful to someone who is looking to understand LFP and the levels of quality control necessary to have safe batteries. The reports are a bit of a “How it’s Made” but written for the target audience of engineers and has similar utility.