Hacker Newsnew | past | comments | ask | show | jobs | submit | Palomides's commentslogin

modern tls libraries on classic macs has been available for at least 6 years, and was pretty easily possible sooner if anyone bothered, no AI needed

I work on embedded linux machines that go a decade plus in active use with micro sd cards, it's not a problem at all

I suspect most of the "corruption" RPi users experience is likely bad power (shitty undersized USB power bricks) glitching the CPU and causing the filesystem to corrupt. The SD card is likely doing exactly as it's being told by the host system - write (invalid) data to blocks, which it does.

I had more than my fair share of SD failures running pis off their official power supplies.

It really is a high failure rate.


"Enterprise" Boot SD's in servers love to die.

More than likely due to being in a warm box...


Do pis have an official power supply?

Yes. https://www.raspberrypi.com/products/power-supply/

I have always bought the official power supply, yet of the two Pi 3s that I have connected to my TV at home, the one that runs Kodi/OSMC to serve up movies has never suffered any filesystem corruption, despite it suffering repeated power outages without warning (it's low enough to the ground that the toddler has often pulled the power cable out, and I haven't been able to find a better place to keep it). The other one, bought at the same time from the same source, is meant to be the one the kids use for programming in Snap!, a variant of MIT Scratch that doesn't have the artificial limitations that Scratch has. (Snap! functions — the exclamation mark is part of the name — can take lists as input, whereas Scratch functions can only take numbers or strings, giving it an artificial ceiling; also, Snap! functions can take functions as input and return functions, making higher-order programming possible, while Scratch deliberately does not allow that). However, that Raspberry Pi has suffered two SD card failures in the past year; it's been down more often than it's been up, and I've ended up using either my laptop or my wife's laptop for my kids' programming lessons.


I've had enough SD and micro-SD cards die on phones and cameras to doubt that is the only problem.

People probably see more of a problem on raspberry pis because they write more to the memory than other systems.


Reminds me of years ago, I had to send a load of photos to a client of something I'd travelled 5 hours to photograph.

Didn't even check, just took and sent. Then went back.

Only to find that the SD card in the phone was dying, and every image was corrupted from about 1/3 down...


It's actually weird that Raspberry Pis seem so bad at using SD cards.

I've just scrapped a large PBX which had a whopping 8MB SD card in it. Its 486 processor booted Linux off the SD card, it wrote CDRs to the SD card, it served said CDRs up over a web interface off it, it saved its logs to the card, and all the other day-to-day chatter of a running system.

It was installed pretty much 20 years ago. The date sticker on the SD card shows that it was replaced 18 years ago, presumably as part of an upgrade.


I'm going to guess that it was not a normal COTS SD card, and instead designed for continuous writes.

The factors for SD card failures in raspberry pi seem to be mostly SD cards not meant for the purpose and having things like swap, atime or disk logs enabled which leads to unnecessary writes.


It just appears to be a pretty standard Sandisk SD card.

Because of a properly configured, boot disk. Just boots once per power cycle.

Meanwhile something like the pi, on most images, is writing logs to the storage CONSTANTLY. Not to mention is regularly booted.


This was constantly writing to the SD card. It was the only storage on the device.

Rebooting it was a big deal though, so that rarely happened.


you're attributing a lot of intent to the details of a site that is almost certainly 100% AI generated

Still supposedly a human prompting stuff to happen in any case. Not to mention, I think you'd have to prompt against the model to avoid making URL routes in a 100% frontend application, leading me to believe the "missing" page navigation is intentional.

for some of these gpus you can set a very reduced power limit for modest reduction in performance, tdp is not the full story


you can install openwrt on basically any x86 mini-pc, works fine


I know, and I have, however, I would prefer a ultra-low-power arm-based platform.



it's very easy to verge into OLPC type thinking with this, you probably should just give them normal bikes instead of trying to come up with some bespoke DIY-able system


the OLPC was the "send them bikes" disguised as the DIY-able system. They designed what they thought would be the best computer for impoverished children when in reality those children had no real use for a computer


I dunno, saying they don't have a use for a computer seems incorrect to me, even in the poorest parts of the world today, a substantial minority have acquired their own computer that's cheap, small, efficient, has a built in long distance radio, can run arbitrary software, doesn't require reliable electricity (I mean cell phones)

obviously it's a different technological and economic today compared to 2005, but still, "the global poor don't need computers" is questionable just based on the fact that they are spending their own money to get them


Saying they had no use for a computer was a poor choice of words. Rather, OLPC wasn't what those communities needed


It's extremely expensive to ship bikes around - even old unused ones.


isn't this what bike thieves do though?


> bespoke

I see what you did there


if we had a standardized way to deliver packages like we do mail (the heavily regulated mailbox!) this would not be a problem, it's a phenomenal waste of human effort to navigate this uniquely for every location

it's not like a doorman where there's useful social interaction


I have a PO Box (I do not have a mailbox so the city gives me a PO box for free). A kind human gets my packages from the back and hands them to me. It's great and efficient. The only inconvenience is the hours are a little narrow.


I'm not very familiar with nix, how well does it do with cross compiling? is anyone actually using it for yocto sorts of domains?


You can generally remotely build for any supported architecture[1]. The build process will be painfully slow though if you have cache misses.

[1] https://search.nixos.org/options?channel=26.05&query=boot.bi...


Nitpick: binfmt.emulatedSystems is not true cross compilation. It sets up a nixos to enable emulated native compilation. After painful tinkering, i prefer this way too. When i recall correctly, for cross compilation you'd have to use nixpkgs flags.


At some point it's a matter of perspective. If the emulated system is fast, it feels the same as "true" cross compilation. In the end it's a bunch of bits and bytes which produces some other bits and bytes when we poke it. If it goes fast, it's good. If it goes slow, it's painful.


Is this similar to QEMU user mode emulation?


We do Nix-based cross-compiling for non-embedded software (ARM vs x64). I'd say it works mostly transparently.


Nixos can cross build os images for Le Potato with device tree stuff: https://hub.libre.computer/t/nixos-for-aml-s905x-cc-le-potat...


have you tried OpenTabletDriver?


Yes, but if I remember correctly I ran into some issues with it. It's been a long time ago so I don't remember what those issues were.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: