Hacker News Reader

Comment by azinman2 | Hacker News Reader

Makes me sad that they’ll now have to increase profits forever, instead of functioning on their mission and doing what’s right. This may mean moving manufacturing to China, using lower cost components, etc etc.

View on HN

The first order bit of better outreach is a low price point, which they've completely lost the plot on.

If going public is what takes them to realize it, I'll cheer it on.

Yeah. I just don't see how they could possibly grow in a way that satisfies the stock market. The only way to create an appearance of such growth is to go Boeing and essentially burn down the house for warmth. For an organization like this, it's basically always step one on an irreversible death spiral.

From my experience doing IT in industrial manufacturing, there is a ton of room for them to replace expensive PLCs where they aren't really needed. The company that makes the big fancy automated machines also make the simpler auxillary ones that supplement their main offerings. Since they know how to build with PLCs, that's what they use instead of sticking a Pi or some small form PC in it instead. Stuff like barcode readers on a manufacturing line or some simple piece of QA equipment at the end of the line measuring the height of a bottle or something. Adds thousands of dollars to the BoM but is complete overkill.

There's strong diminishing returns in this sort of growth. With anything, you can optimize it until you can't. The market requires you to grow forever. How will Raspberry Pi Ltd still be growing in 50 years when the entire orchard of low hanging fruit is long since picked barren?

I don't know, how does any other company do it?

> From my experience doing IT in industrial manufacturing, there is a ton of room for them to replace expensive PLCs where they aren't really needed.

From my experience with PLCs, they are hardened and robust to the extreme degree required by the harsh environments in which they are used. Last time I looked, the R Pi was the complete opposite. It can be crashed by just a nearby camera flash, and then leave its SD card corrupt to boot. Fine for hobbyists (apparently). Not fine for industrial applications.

Right, I'm not suggesting they replace critical PLCs. They are overused in places where those requirements aren't necessary just because PLCs are what their engineers know.

Compute Modules are targeted squarely at industry. Environmental hardening is provided by the cabinet you install them in.

I built an RPi 4 based plasma cutter and I've never had a crash. You don't get more EM interference than that, short of putting it in a microwave.

> From my experience doing IT in industrial manufacturing, there is a ton of room for them to replace expensive PLCs where they aren't really needed

Just for fun try to find a way to get a Pi mounted up to a DIN rail with an enclosure that can resist industrial vibration and has good particulate penetration/filtering and ruggedized connectors. Good luck finding accessories that aren't 3D printed with materials that will break down when coated in machine oil or in direct sunlight.

Now try to figure out how to get whatever monster you built installed in an ISO 9001 or AS 9100 shop.

Again, I'm not talking about machinery that actually needs PLCs, I get that. A lot of stuff doesn't have vibrations, temperature, safety issues etc. but is built around PLCs out of tradition and BoM reasons.

"Here's our end of the production line fan that blows dust off of your product for $4k because we used a PLC." There is lots of stuff like that that just has no business using a PLC.

I (thank god) got out of manufacturing IT around 2000 but I still have friends working in that industry. A guy I know makes an absolute fortune frankensteining old MFM drives together and fixing resistors on 10base-T network cards because someone thought the exact same thing and subbed out a PLC for a computer to save a few bucks.

Trust me, I loathe PLCs, but Allen Bradley will still sell you a drop in replacement for something the now retired previous guy put in service 15 years ago.

Yeah I'm out of it now too. You still have the same issues running PLCs. I've had to find old ass network cards on EBAY. I understand why PLCs are used. I turned down a project building a machine using Arduino's and raspberry Pi's because it was inappropriate for the application. But there is definitely a subset of manufacturing equipment that doesn't require PLCs. I've already seen it happening scanning networks and seeing devices show up with RPi MACs.

I've started seeing PI show up on the DIN rails in some industrial equipment, usually HID related. PLCs are built to safety critical standards, a PI with some error might go executing code off into never-never land, where a PLC will fail to a known output state. Obvious oversimplification. https://en.wikipedia.org/wiki/Safety_integrity_level

It's not an insurmountable problem though. Have a watchdog voltage with a relay or something. Engineering reliable failure modes is trivial.

> Makes me sad that they’ll now have to increase profits forever

Providing a return to shareholders does not require increasing profits. Investing constant or declining profits into acquisitions or share buybacks can provide the same returns to shareholders.

What is "right"? They haven't been a good value for over 2 years.

The goal hasn't ever been about producing a device cheaply. It's about supporting education, encouraging tinkering and changing attitudes [1]

[1] https://www.raspberrypi.org/about/

Every altruistic thing is easier if you're trying to do it with a $10 product rather tham a $100 product, all else being equal

Good thing they still offer a $10 product then - the Raspberry Pi Zero.

and the RP2040! There is not much to differentiate it with ESP32 at the moment, but it's their first shot at a microcontroller.

"...hasn't ever been..." is not true. Someone above posted an earlier statement from the archive.

Good value for what? They have never been good value as desktops because the early ones were too slow. Now there are better little computers.

But as hobby computer, they are still a good value. There are faster boards and cheaper boards, but which ones are good? More importantly, which ones will be supported in five years? How easy it to install software? I read about distro that promises support, but then have to check the list of hardware it supports.

I have six year old Pi 3. I can install latest 64-bit OS on it. For my plans, something small and low power is an advantage.

Hardkernel has supported their ARM boards for 10+ years. The Odroid XU3/XU4 line, for example, was released a decade ago and still has active development from the manufacturer with recent kernels and images.

I was really excited about those hardkernel boards 10 years until I got one and found I had to use their outdated custom version of Ubuntu and still half of it didn't work.

I assume things have gotten better since then but they don't have 10+ years of good support.

Lower cost of components doesn't mean lower quality. What made the cost lower is logistics and the availability of components within the region.

Yup, cue the imminent enshittification. E.g. your Pi6 will require sign-in and come with ads.

Reply on news.ycombinator.com