Comment by teruakohatu | Hacker News Reader

Comment by teruakohatu | Hacker News Reader

Parent Full thread

> First up, the Raspberry Pi 4 compute module. From a scrappy little “$35 computer” put out by a charity originally for the educational market, Raspberry Pi has taken over the world of single board computers, socket by socket. Thanks to the financial backing it had from government grants and donations as well as tax-free status as a charity, it was able to kickstart an unusually low-margin hardware business model into a profitable and sustainable (and soon to be publicly traded!) organization with economies of scale filling its sails. It also benefits from awesome software support due to the synergy of its charitable activities fostering a cozy relationship with the open source community. Being able to purchase modules like the Raspberry Pi CM with all the hard bits like high-speed DDR memory routing, emissions certification, and a Linux distro frees staff resources in other hardware companies (like Formlabs and my own) to focus on other aspects of products.

A very succinct description of the Raspberry Pi past trajectory.

I find it confounding how 'cozy' RPi's relationship is with the open-source community given how little of their project is open-source. But I infer that open-source hardware and low-level software is much less important, interesting, and malleable than the OS-and-above software.

One often under-appreciated aspect is that they have a tiny set of SKUs, which means opensource contributions/docs/articles/hype are all concentrated on those few SKUs. It's especially true for RP2040, their competitors have massive lineups that help to right-size the MCU, but this also means that any given SKU has less opensource support.

This is definitely true, especially when rpi is probably more popular than all other SBCs combined. I always feel burned about 2 years after buying anything but beaglebone, rpi or Jetson nano and can't find support. With rpi you can always find an up to date OS image (for the boards that need an OS).

If you use complex devices like cameras, you can probably find recent drivers and readme files for rpi. For other SBCs you will probably end up trying to adapt documentation for rpi and getting stuck because your OS is using a weird version of libc, or something like that.

Cheap, open or interesting. If your product is at least two of those things, you should be prepared for the Open Source community to hack it.

Well put. And I agree. I draw similarities between RPi and OtherOS on the PS3. Sony permitted Linux on their consoles and in exchange the hardware was locked. PS3 was not hacked until they removed the OtherOS feature and was the last console of that generation to be hacked. RPi is not Sony and would (probably, hopefully) not start limiting user OS choice, but the moment that happens, it will be completely and thoroughly hacked as there would be an incentive to do so.

> But I infer that open-source hardware and low-level software is much less important, interesting, and malleable than the OS-and-above software.

Nope. It was all about the subsidy.

Because the RPi was subsidized, competitors like the Beaglebone series couldn't achieve a critical mass since they were always being undercut. This set up a decade+ of mindshare.

And then the RPi foundation knifed everbody in the shortage by shunting all the RPis to commercial companies like FormLabs.

Even now that low-end x86 machines are cheaper than the RPi's, everybody still sings the praises of the RPi's.

Marketing always wins.

> And then the RPi foundation knifed everbody in the shortage by shunting all the RPis to commercial companies like FormLabs.

The Foundation did no such thing because it doesn’t manufacture the boards and it’s not their decision.

The Trading company manufactures the boards, and it chose not to destroy its commercial contracts (all of which experienced limited stock too, particularly of CM4s, even though they were prioritising CM4s), which kept the lights on. They were already selling dramatically more units into commercial channels (Citrix for example) than to hobbyists.

The two entities also introduced the Raspberry Pi Pico and Pico W in the pandemic era, which were wildly popular and kept their educational goals on track, as well as giving suppliers like Pimoroni and Adafruit something to design absolutely amazing products around, with a microcontroller that wasn’t supply-constrained. IMO they more than made the case that most people don’t need a whole linux box to run a physical computing project.

I don’t get the whining. Do the little Intel boxes work for you? If so, use them and leave Raspberry Pi in the dust; there’s more than one way to add GPIO, SPI, I2C and PIO to your projects.

(Side note: the Beagleboard is a Texas Instruments product, is it not? I am not convinced its origins are all that different to the Pi, in that regard. If TI couldn’t compete on price and marketing, that is on TI, surely)

Beaglebones were popular for a while for GPU farms for crypto, and the shortages were way worse than raspi. I fried one in a PhD project and had to spend a few hundred dollars to get a new one. I used to swear by beaglebones, and they had some nice features, like one usb cable to provide power, expose a serial console, and bridge your internet connection to the board. And one killer feature: 2 independent microcontroller cores that can access the main CPU's memory directly. You could use them to bitbang almost anything. Over time though, most of the newer beaglebones focused on one specific feature and used several different form factors, and most lost the features I loved. BeagleV fire looks interesting but it's missing wifi.

Back in the day (like 12+ years ago) I used some intel SBCs with GPIO, i2c etc. At that time they were nice because they just worked with your favorite linux distribution and put your binaries on it. At the time it was painful to setup an environment for cross-compilation. Today that's easy.

Raspberry Pis are never the best SBC available, and have some very closed aspects but they are usually available, have good community and corporate support and are supported for a long time. I've bought a lot of the other SBCs out there and after a couple of years you find that you're almost the only user left, and the vendor has 30 new products and don't support yours. Then one day you're trying to update your linux kernel for feature y but it breaks feature x, and searching for help yields workarounds and patches, or even OS images for the same issue... for rasberry pi. After a few times, you learn to stick to raspi, or maybe Jetson if you need more.

Funny you mention Intel. I got burned by Intel Edison getting discontinued. Intel has certainly burned their bridges with the embedded industry and I certainly would never buy from them if they chose to re-enter.

Edison was really enormously promising, but it sort of only existed as a reactionary product in that whole ed-tech/maker bubble that blew up and then shrank a few years back.

That "low-end" x86 machine still has an Intel/AMD chipset in addition to the processor, possibly a BIOS/UEFI chip from yet another vendor - they're substantially more difficult from a hardware standpoint to get going than any ARM MPU. Debug tools are inferior, as well.

ARM MPUs became popular because they are actually easy to deploy.

Their efforts keeping up to date with current Linux kernels is better than most boards of that nature. That deserves some open source credit.

I hope that OSS hardware gets as cheap as these tiny proprietary chips in the future, but I think it'll take a while.

Reply on news.ycombinator.com