3 ms·

I just can’t come up with any compelling use for a raspberry pi. The practical purposes for an RPi have diminished. They're still good for when you need somet

I just can’t come up with any compelling use for a raspberry pi. The practical purposes for an RPi have diminished. They're still good for when you need something small, light or low power. But the market for refurb mini-PCs really replaces most of the instances where you're just looking for a cheap, but semi-powerful, computer. And things like the ESP-32 have reduced the Pi's practicality on the other end where you just need something powerful enough to read and transmit data from a sensor.

meatmanek 2y ago
> And things like the ESP-32 have reduced the Pi's practicality on the other end where you just need something powerful enough to read and transmit data from a sensor. Especially with ESPHome making it so easy to integrate with common sensors without writing code. That said, the Pi Zero W is pretty cheap and small (similar in size to many ESP32 development boards) so it provides a nice upgrade path if you want to do more processing on device, exceed a few megabytes of ram, etc
moffkalast 2y ago
And the Pi Foundation's refusal to do literally any power management while increasing TDP into the sky with marginal efficiency improvements has diminished its use for robotics and other battery powered applications too. This has just been one of my largest pet peeves with the entire Pi lineup. The Pi 5 is the worst offender that idles at like 3 watts doing jack shit. That's an entire 3000mAh battery gone in under four hours that would've easily lasted several days on an average smartphone SOC with twice the memory and speed.
- ssl-3 2y ago
  That's valid. My first reaction was "Well, the OS is open -- isn't it? Just set up good power management yourself and publish it for all to use, and it'll probably get accepted upstream eventually." But then I remembered: The OS is open, but the hardware is not (because -- in one word -- Broadcom). But maybe with an IPO, they can generate enough cash to get their own CPU started up. Maybe something based around [checks flip-chart] RISC-V or something, with enough documentation to allow people to fix the [checks excuse card] power management problem. They've certainly shown some willingness to work in that direction with the RP2040.
  - moffkalast 2y ago
    From my understanding the Pi Foundation was founded by ex-Broadcom engineers and some that still work there. The two are so intertwined that it's not unlike it being the Broadcom Foundation. I think it's rather unlikely that they'd ever move away from using their chips since they are theirs.
    - ssl-3 2y ago
      Yes, the whole thing did start with some folks who were under Broadcom employ. Perhaps that will change: After all, how many engineers do you know who feel motivated to hold down regular mid-tier jobs at two publicly-traded companies?

meatmanek 2y ago

> And things like the ESP-32 have reduced the Pi's practicality on the other end where you just need something powerful enough to read and transmit data from a sensor. Especially with ESPHome making it so easy to integrate with common sensors without writing code. That said, the Pi Zero W is pretty cheap and small (similar in size to many ESP32 development boards) so it provides a nice upgrade path if you want to do more processing on device, exceed a few megabytes of ram, etc

moffkalast 2y ago

And the Pi Foundation's refusal to do literally any power management while increasing TDP into the sky with marginal efficiency improvements has diminished its use for robotics and other battery powered applications too. This has just been one of my largest pet peeves with the entire Pi lineup. The Pi 5 is the worst offender that idles at like 3 watts doing jack shit. That's an entire 3000mAh battery gone in under four hours that would've easily lasted several days on an average smartphone SOC with twice the memory and speed.

ssl-3 2y ago
That's valid. My first reaction was "Well, the OS is open -- isn't it? Just set up good power management yourself and publish it for all to use, and it'll probably get accepted upstream eventually." But then I remembered: The OS is open, but the hardware is not (because -- in one word -- Broadcom). But maybe with an IPO, they can generate enough cash to get their own CPU started up. Maybe something based around [checks flip-chart] RISC-V or something, with enough documentation to allow people to fix the [checks excuse card] power management problem. They've certainly shown some willingness to work in that direction with the RP2040.
- moffkalast 2y ago
  From my understanding the Pi Foundation was founded by ex-Broadcom engineers and some that still work there. The two are so intertwined that it's not unlike it being the Broadcom Foundation. I think it's rather unlikely that they'd ever move away from using their chips since they are theirs.
  - ssl-3 2y ago
    Yes, the whole thing did start with some folks who were under Broadcom employ. Perhaps that will change: After all, how many engineers do you know who feel motivated to hold down regular mid-tier jobs at two publicly-traded companies?