It was nice knowing Raspberry Pi while they lasted. Going to suck losing something that has changed the homegrown embedded system hobby forever.
You must log in or register to comment.
Sooner or later capitalism ruins everything.
Then it’s a good thing that no countries have pure capitalism for their economy.
We need regulation on corporations to keep them in check.
The fact that regulations make capitalism less dangerous doesn’t mean that capitalism is fine as long as its regulated.
Hand grenades have a tonne of safety features, but you wouldn’t let your kid play with one. “Safer” isn’t the same thing as “safe”.
What would you propose as being better than the mix of capitalism and socialism that almost every country already has for their economy?
Both extremes lead to terrible outcomes.
Has any country actually TRIED anarchy? I know it sounds terrible on paper, but like…could it possibly be any worse then whatever the fuck north korea is doing? The dictator and his dogs eat very well. Nice beef meals. Whereas the citizens are more like prisoners within a country. Most never even seeing beef because it’s too expensive.
Would their lives actually be any worse off if there was just no government, no police, no military, no rules, just everybody for themselves?
Because it kind of seems like they got nothing to lose. It be an amazing case study.
There was the Capitol Hill Occupied Protest. It’s not a whole country going anarchist and no doubt the limited amount of people with the nessisary skill sets to have a functioning society (judging from the food garden they set up) held back the viability of the protest, but in general the Capitol Hill Occupied Protest was widly seen as a wild failure.
It’s an interesting thing to look up on, and I’d definitely recommend anyone who is serious about anarchism to study it for the potential pit falls of an anarchist society that they would need to work out first.
“Listen, we already watered down our deadly poison a little and it’s still killing us. Unless you have a better idea we’ll just have to keep drinking deadly poison.”
I’d let my kid play with a grenade. Then again, I don’t have kids by choice, so to imply I had kids would be to imply that at some point something went terribly wrong. But rectified in the most absurd method possible.
Plus, you couldn’t go to jail for child abuse, because what parent is “double checking” that the grenade he’s playing with is in fact a toy? BECAUSE WHERE THE HELL DOES THIS 3 YEAR OLD GET A GRENADE???
That logic would track in court. A very sad, very bizzare set of circumstances. That theres no way you could blame the parent for.
And so begins “Line must go up” and the inevitable enshittification .
I’m willing to bet they’ll start adding telemetry features in RPiOS for “quality purposes” a few years from now.
They already have that proprietary and opaque GPU that has full memory access akin to the Intel ME, and its programming is very difficult to audit. There has been something quite fishy about them ever since they left their educational mission behind after the Pi 1 and went for-profit.
Isn’t the GPU documented now?
https://docs.broadcom.com/doc/12358545
There are reverse engineered docs as well: https://github.com/hermanhermitage/videocoreiv
Yeah, expect nothing more than enshitification. That way, if they don’t enshitify like every company does, then we’ll be pleasantly surprised.
The nice thing about being a pessimist is that you are always either right, or pleasantly surprised.
Not with my luck
That’s the spirit!
The shit winds are coming, Randy
Tech companies as soon as they are publicly traded:
Let the enshitification begin!
The Pi5 is already a shitshow with crazy power usage requiring a special power supply instead of a normal USB C phone charger.
We’re lucky that the SBC space has gotten really solid over the last couple years. ARM-based, X86-based, and even some RISC-V systems.
The PI isn’t the only only game in town now, and actually gets beat in several different applications depending on use case.
As shareholder value and line-must-go-up takes over the company culture, progress and innovation will happen more and more in the hands of companies and orgs that actually care about their product’s quality and features.
Still disappointing though, the Pi was my first introduction to IoT and low power computing.
Yeah I’d take a 3b-ish PI for say 30€ any day (IDK if that’s realistic pricing). If I need beefy hardware I just use a PC?
The Pi4 had a good price on release. Then Covid hit.
With the Pi5 the Pi foundation is just milking it. Overpriced chip on an inefficient outdated 28nm process node.
They’ve crossed the event horizon of enshitification.
N100 mini PCs are where it’s at these days anyways. Unless you need the GPIO pins or are running some weird niche configuration, you’re better off grabbing any N100, they’re cheaper too.
I look for broken but working sff/tiny deals. Scored a sweet i5 7500 /16gb system for $100CAD. Just had a broken audio port I was never going to use.
The fool you will be revealed to be once I complete my Ethernet Over Audio implementation.
Ethernet Over Audio
Isn’t that just a telephone modem?
No, silly, that’s Audio Over Ethernet! /j
It is I! USB-C-MAN! Begone with you foul villain!
Oh boy, you’ve got a lot of protocols you can borrow for your OSI layer 1. ribbitradio and the telephone modem spec.
I just want you to know this is one of my favorite comments of all time.
After some light searching, am I missing something? I don’t see n100 cheaper than rpi 5
Yeah, they’re nearly twice the price.
Far more capable though, and typically specced with 16GB RAM and a 500GB SSD.
You’re forgetting to include the Pi heatsink, the Pi power supply and the Pi enclosure.
Yeah good point, adds $10-$30 on top of rpi
Cheapest I’ve seen was $105
PIs are kind of screwed from N* on the higher power end and ESP32 (or similar high power micro controllers) the lower end.
It’s become an underpowered middle player no one needs.
It was good while it lasted. PI3’s for $30 we’re amazing.
Added benefit of most using low power intel CPUs
They’re not actually lower powered, they just have a TDP limit set.
E.g. A 8500 and 8500T will idle at the same power consumption, but the 8500T has a TDP limit set.
GPIOs are the easy bit. You can get those no issue on x86. It’s I2C and SPI that are the issue with x86. You can get the buses sure, but all the device drivers are Device Tree based. You can’t just throw in Device Tree overlays on x86.
Idk, with I2C if it’s not something that needs a kernel level driver, there usually isn’t a problem with interacting with it from user space, for example basically all RAM RGB controllers are I2C and OpenRGB has no problem with them. I’m pretty sure I’ve only ever used an I2C device tree overlay for an RTC.
Also I2C/SMBus is present everywhere on x86, like some graphics cards expose it through their HDMI ports, even some server motherboards have a header for it; but for GPIO I’m unaware of any motherboards that expose it, so good luck researching the chipset and tracing out the pins.
Only a fraction of it is RTCs. What is in the Pi overlays folder is from everything. Not even all the DT I2C RTCs. There is loads of ADCs, DACs, IO extenders, all sorts.
It’s really annoying you can’t do DT on x86 Linux. It’s a bit of a gap in the platform. It would make Linux ARM based developer’s lives easier.
ADCs, DACs, IO extenders
These should all work without kernel drivers. For example, here’s a user space python library for ADS1*15 ADCs, or Nuvoton MS51 IO Expanders. Unless you need very specific timing or require the kernel to know about it, you shouldn’t need a kernel driver.
You can of course write drivers for them, but then it’s you own abstraction not the standard Linux abstraction. (You can hack something up with IIO for that stuff, but it’s not pretty). There is CUSE (part of FUSE) you can do some character devices with.
Existing drivers in Python are messy to use if you our not developer in Python.
The nice thing about in kernel is:
- it’s done for you already
- the interface is standard and will work with anything that uses that class of device
- it’s langauge agnostic.
The Linux kernel does hardware abstraction. It’s not a microkernel. There is limited support for proper userspace drivers.
If you doing some application specific app, that will only work with those chips, use do it in userspace. But to make a normal system for normal use, you want things in kernel like normal.
I have a pi 4, how would the transfer work? Can you install pios on the n100 and just clone stuff over?
N100 is a standard Intel x86 family chip, so no. Plenty of power though, so you’d be able to install any Linux distro or even Windows if you wanted to disgust Lemmy.
Or get a used Thinkcentre tiny, way cheaper. Some have a serial out too.
Can you swap the hard drive between any generation and still have it boot and work 100%? To me that was the second biggest feature after all the gpio and i2c buses I used to hack all manner of stuff together. Heck I even have a cargo trailer powered by a pi!
Raspberry Pi has been over priced for a long time. I’m not saying they’ve been a net positive or negative, but if you think this will make them a bad company then I think they’ve been pretty bad for a bit.
I picked up a radxa zero last year and have been quite enjoying it. the hardware is better than a pi zero but costs less. same with a lot of other SBCs
but raspberry pi has a lot of inertia behind it, a lot of software and hardware support. people will keep using them, just like they keep using Ubuntu, even though it’s a soulless corporate husk of what it one was
As long as I’m not locked into using their OS on their hardware I’ll still be interested. I have a 3,4&5 doing various tasks around my house and enjoy the little boards.
I had never heard of radxa. Looks awesome!
Orange PI comes to mind, getting better over time too.
Everyone here seems pretty negative on this news. Any particular reason?
Going publicly traded fucks every company up with nextquarter-itis.
Mostly that IPOs put companies into ‘infinite growth mode’ which is obviously impossible, so their product just degrades over time. They can’t just do ‘good enough’ anymore.
Raspberry pi foundation was launched as a charity, and the end goal was to produce a ton of very cheap computers to help children learn about programming. Since then, it has been soo ubiquitous for embedded stuff that for the last couple of years they have basically become unaffordable for the very audience they were intended for. Now they are seeking an ipo because they are used in everything, except as cheap computers for children.
Are they really used in a bunch of stuff? I still onlt see them included in hobby/homelab/maker/education stuff.
Every time a company goes public, they become more and more profitable until the only way to continue on that trajectory is to worsen their own product.
Think they’ll still be selling the Pico for $4 or the Zero for $15 after they’re reporting to shareholders?
Big pharma companies jack up the prices of life saving medicine that’s been affordable for decades and don’t lose a bit of sleep. You bet your ass a hobby electronics company will jack up prices as far as they think they can.
Price is one thing but the push for returns on investments is massive, this means that it’s time to start cutting corners on everything (except maybe marketing! Yea!). Quality, repairability, and innovation all start to crumble.
Don’t call Raspberry a hobbyist electronics company. Their primary consumer has been business and enterprise customers for years now, industrial/controls companies jumped all over the pi as a super easy drop-in board that can be programmed by any code monkey.
The Pi hardware shortage of the last few years has mostly been because of this demand, with Raspberry openly saying they were prioritizing bulk corporate orders foe their production volume over hobby consumers. Fuck the little guy, Pi is dead.
Going public introduces shareholders that prioritizes return on investment as opposed to making technology and knowledge about technology accessible for many.
It doesn’t always end this way but often enough to worry about it…
Because the more commercial they get, the more they stray from their original purpose as a charity to provide low-cost machines for kids to learn about computer science.
First there was the Dynabook, then OLPC, then Raspberry Pi, and now we’ve basically got to start over yet again because enshittification is imminent.
In Tech, an IPO means the business is market ready to be sold off in pieces, ie stocks. The people who buy the product don’t care what it does, they use the product maker as a vehicle to more growth and profit. Typically that means the people who now own the business make poor choices about cost cutting, like off shoring support and removing unuseful documentation while removing people with critical tribal knowledge about processes. Each step the new owner takes will be to make the business more profitable, and in the world of business, the only thing they care about are the numbers and not the environment or people that created those numbers.
Opening up to institutional investment means opening yourself up to ownership by a culture that demands infinite growth. In recent years this has gotten particularly bad; with the rise in interest rates, stocks can no longer deliver moderate growth and still be considered worthwhile investments. Everything is either a rocketship to the moon, or its a sell. Combine that with a string of US court cases that have interpreted tge law in such a way as to foster the belief that its illegal for companies to put anything ahead of shareholder value, and what you get is a top down imperative to squeeze the maximum profit out of everything. When you see Microsoft mulling over ideas like putting ads in your start menu, or EA talking about in-game advertising, this is why. When you see Spotify raising prices multiple times while crowing about how their content production costs are basically non-existent and changing their contracts so that smaller artists literally don’t get paid for their music, this is why.
They think that it’s gonna ruin the company
And they’re right.
There are a high proportion of far-left types on here. I could see them wanting something to be government-owned or something. But wanting a company to be privately-owned rather than publicly-owned seems odd to me.
And the “enshittification” comments seem odd too.
“Enshittification” isn’t some sort of catch-all term for a company doing worse. Doctorow coined it to refer to a point where a company that had been losing money to grow a customer base ends the rapid-growth phase and starts monetizing that base.
That makes business sense for some companies with low marginal costs and high fixed costs, and especially where there is network effect, like social media companies.
But here, the company is profitable, and not unreasonably so. Like, they don’t have a monetization phase that they need to transition to.
In 2023 alone, Raspberry Pi generated $266 million in revenue and $66 million in gross profit.
Raspberry Pi priced its IPO on the London Stock Exchange on Tuesday morning at £2.80 per share, valuing it at £542 million, or $690 million at today’s exchange rate.
We’re mostly negative on publicly traded companies because their ceo is legally obligated to squeeze blood from a stone or they quite literally will get sued by the shareholders, plenty of examples out there. The exceptions are usually there because the previous owners wrote contracts, etc to help keep the company as it was prior but even then it only works for so long. Check out Ben and Jerry’s and their whole debacle on the subject.
There are certain fiduciary obligations that CEOs hold to shareholders. But on the flip side, if someone opposes the transition of privately-owned companies to being publicly-owned, then their position is that only the wealthy, those who can outright own a company rather than only part of it, via shares, may own companies. That seems quite like a policy exceptionally loaded towards the wealthy. It would make capital much harder to get, so it would be harder for someone who wants to start a company to do so. Only very wealthy entities – stuff like very wealthy families – would be able to own companies of any significant size. They would have little competition for their capital, and would be able to demand extremely favorable terms for it. Less-wealthy people would be intrinsically disadvantaged by their inability to must outright buy companies. Less capital availability would tend to impact wages negatively.
It seems to me stupendously at odds with the sort of thing that I would expect someone on the left end of the spectrum to want.
So, what are the alternatives?
https://lemux.minnix.dev/c/sbcs
There are so many!
Hi there! Looks like you linked to a Lemmy community using a URL instead of its name, which doesn’t work well for people on different instances. Try fixing it like this: !sbcs@lemux.minnix.dev
There’s tons of similar SBC’s out there from Chinese manufacturers, like Orange Pi, Banana Pi, etc; usually using mediatek RISC-V or rockchip ARM processors. They’re all poorly supported on the software and documentation side though and take more work to get going, which has always been where Raspberry shined- nobody else has made embedded computing so easily accessible with click and go OS options and continuous kernel maintenance.
Probably the only board closest to software parity is the pine64 boards… but it’s still not quite as good.This is the key point for alternatives. None seem to have the community and support (docs, s/w quality etc) that is remotely close to that of the Raspberry Pi.
Guess the community for some of these is about to get much bigger. I’m not in the market for an SBC but this is a big negative against the Pi.
They have more features though, like extra Ethernet, PCIe brackets and M2 slots on the board
Those features don’t mean shit if you can’t use a modern OS after a couple of years.
Why could you not run a modern OS after a couple of years? Those SBC manufacturers did not invent an entirely new processor architecture for their computers, you can just generically compile the kernel (plus maybe some slight device tree work).
you can just generically compile the kernel
Not always. I have numerous old now useless SBCs that never merged their shit with the mainline linux kernel so my only option is to run something 10+ years old.
Beagle bone
deleted by creator
Orange pi is getting better and better. Far from raspberry though.
I got a ‘LePotato’ a few years back when Pi had stock issues, and it worked quite well as a Pi 4 clone.
Yep, using one to run clipper for my 3d printer with armbian as the OS. It’s been rock solid for me. There obviously some adaptation and discovery when trying to use the io as it’s similar-but-not the same as the raspberry pi io and manipulating it is not the same. But it works, it was available, it was competitively cheap, and it’s been stable
Plus I get to say I’m running my 3d printer on a potato
LePotato is a great budget board for pi-hole.
I think Pine64 is pretty cool.
Radxa as well. I have a Rock Pi 4B running as my home server and it has been a great Pi 4 alternative. I also have an Indiedroid Nova with RK3588S which should be better than the Pi 5 bit the GPU drovers aren’t quite there yet. Once GPU drivers are in it should be an incredible board.
I like your attitude!
Oh jees… welp it was great while it lasted.
Welp. Fuck Raspberry Pi. The entire stock market is one big scam.
Got my last Pi (RBP5) to try to set up a simple TV player under linux… unfortunately the performance was shit… had to go with Android and it’s barely OK (bang for buck)
With the IPO I expect RBP are going to become more expensive and significantly enshitified… so that’s that
? RPi5 is something like 2x faster than RPi4. Are you using some format that RPi doesn’t accelerate? Or are you running something heavy?
I almost picked up an RPi5 to replace my NAS, but the SATA hat was out of stock so I just did a smaller upgrade with stuff laying around my house (Phenom II x4 -> Ryzen 1700, mostly for power savings).
Pi5 doesn’t have h264 hardware. Pi4 is probably better for media centers right now.
That’s encoding, right? It seems to have 4k60 HEVC decoding, which should be plenty for a media center.
That assumes your media collection is all hevc. That’s not the case for most people.
It only needs to be HEVC for 4k content, 1080p works fine in pretty much any format. Most people probably have mostly 1080p or 720p content.
Yes, that’s my point. If you have a library full of 1080 h264 then the pi 4 is a better choice. The Pi5 will struggle with software decoding compared to the 4.
At the end of the day, they’re different boards with different use cases. I think a lot of people don’t appreciate that enough.
What software do you run for that, and is there support for a remote control?
For a standard media center, kodi is pretty great.
I was planning to use it to drive one of my TVs, so basically to be an HDTV player.
The Raspbian OS was fine, the Emby client would not start and the performance on the web client was not great.
Ah, okay. I’m not familiar with Emby, I’ve mostly only used Kodi on my RPi4. I’m guessing there’s a way to get reasonable performance, but you may need to transcode.
What software were you running?
I was planning to use it to drive one of my TVs, so basically to be an HDTV player.
The Raspbian OS was fine, the Emby client would not start (segmentation fault) and the performance on the web client was not great.
Now on Android, Emby client runs pretty well (better than on the FireTV sticks I am trying to replace) but I could not get Google Play working (yet) which left me without F1TV (the only “other” vid app I care about for now to run on the TV)
