• [OT] Raspberry Pi 4

    From Alan Browne@bitbucket@blackhole.com to comp.sys.mac.system on Saturday, February 20, 2021 10:51:41
    From Newsgroup: comp.sys.mac.system


    For those who may be curious...

    I received the Pi last week and have been having some fun with it.

    1) I got the Raspberry Pi Keyboard and mouse as part of a "Kit" deal
    from CanaKit.

    The keyboard is pretty bad, quality wise. It works, but the feeling is
    cheap and cramped.

    The mouse, however, is pretty good.

    I do most of my work from the Mac via VNC. Some keyboard shortcuts
    don't work well however (ie: the F buttons ... can't find a way in VNC
    to assign them).

    2) The Pi 4 B.

    I got the 8 GB version which in hindsight is massive overkill for my
    project. Unloaded, Raspberry Pi OS (Linux) uses less than 200 MB of memory.

    Load Chromium and an IDE it's at 340MB.

    2 GB would have been more than enough for the project I'm doing.

    1.5 GHz 4 core ARM (64 bit), Gb Ethernet, WiFi (incl 5GHz), USB 2.0 and
    3 ports (2 ea), Bluetooth, audio, dual 4K HDMI ... and the more
    important to me GPIO header with configurable discretes (12) and various serial data (I2C, UART, SPI (2)) and PCM.

    I added heat sinks to the CPU, memory and I/O controller and a tiny fan (noiseless) that takes 5V from the GPIO header.

    Raspberry Pi OS is 32 bits, however. There is a Beta release in 64 bits
    and there is a 64 bit Ubuntu. I might load up the latter and see how it
    does. (More than enough SD cards around here...).

    I'm coding (so far) in Pascal (Free Pascal compiler) with crude h/w
    access. Works fine. Hooked up some buttons and lights to get things
    going (polled reads).

    I may be forced to Python 3 or C. Both of which I despise (that said,
    there is a monster library of code for Python 3 devoted to Pi and quite
    a bit of C as well). If forced to do so, I'll write a middle code
    'driver' and pipe the data to the main program in Pascal and v-v. TBD.
    CPU Efficiency isn't important as long as the sampling rate is high
    enough.

    Another option is to make my own kernel variant. That would be both fun
    and a massive time consumer - but I'd have very direct h/w control.
    This would be in C and possibly assembler.

    Waiting on some GPS and accel/gyros to get going. Shopping for a
    display (small touch screen) and Li-ion power control components in the meantime as well as learning about LoRa.

    I purchased a cheap 1080p monitor (BenQ) for this project and I regret
    not getting a 4K display instead. May return it this week and upgrade.
    Coding efficiently needs a lot of screen space...

    --
    "...there are many humorous things in this world; among them the white
    man's notion that he is less savage than the other savages."
    -Samuel Clemens
    --- Synchronet 3.18b-Win32 NewsLink 1.113